Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Pediatric functional magnetic resonance neuroimaging: tactics for encouraging task compliance

Pediatric functional magnetic resonance neuroimaging: tactics for encouraging task compliance Background: Neuroimaging technology has afforded advances in our understanding of normal and pathological brain function and development in children and adolescents. However, noncompliance involving the inability to remain in the magnetic resonance imaging (MRI) scanner to complete tasks is one common and significant problem. Task noncompliance is an especially significant problem in pediatric functional magnetic resonance imaging (fMRI) research because increases in noncompliance produces a greater risk that a study sample will not be representative of the study population. Method: In this preliminary investigation, we describe the development and application of an approach for increasing the number of fMRI tasks children complete during neuroimaging. Twenty-eight healthy children ages 9- 13 years participated. Generalization of the approach was examined in additional fMRI and event-related potential investigations with children at risk for depression, children with anxiety and children with depression (N = 120). Essential features of the approach include a preference assessment for identifying multiple individualized rewards, increasing reinforcement rates during imaging by pairing tasks with chosen rewards and presenting a visual ‘road map’ listing tasks, rewards and current progress. Results: Our results showing a higher percentage of fMRI task completion by healthy children provides proof of concept data for the recommended tactics. Additional support was provided by results showing our approach generalized to several additional fMRI and event-related potential investigations and clinical populations. Discussion: We proposed that some forms of task noncompliance may emerge from less than optimal reward protocols. While our findings may not directly support the effectiveness of the multiple reward compliance protocol, increased attention to how rewards are selected and delivered may aid cooperation with completing fMRI tasks Conclusion: The proposed approach contributes to the pediatric neuroimaging literature by providing a useful way to conceptualize and measure task noncompliance and by providing simple cost effective tactics for improving the effectiveness of common reward-based protocols. Background involving an inability to remain in a scanner to complete Functional magnetic resonance imaging is increasingly fMRI tasks is one common and significant problem. This being used to advance our understanding of normal and paper describes the development and application of an pathological brain function and development in children approach we believe may improve the effectiveness of con- and adolescents. However, there are a number of challenges ventional reward-based approaches used to encourage task that clinicians and researchers encounter. Noncompliance compliance. We discuss some issues surrounding task non- compliance and offer that integrating tactics derived from learning based behavior therapies into conventional * Correspondence: schlund@kennedykrieger.org 1 reward-based protocols may help encourage compliance. Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh PA, USA Using a case study design, preliminary results show Full list of author information is available at the end of the article © 2011 Schlund et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 2 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 improvements in the number of study tasks completed in Conventional reward-based protocols in pediatric healthy children and those with clinical disorders, providing neuroimaging proof of concept data for the recommended tactics. While In all pediatric functional neuroimaging studies, research- the present discussion focuses primarily on task compliance ers and clinicians use rewards to encourage and maintain during functional neuroimaging investigations, some of the task compliance. This highlights an important tie to rein- forcement learning theories. The widespread application problems discussed and recommendations may generalize of reward protocols also highlights recognition of the rela- to other forms of neuroimaging (e.g., clinical MRI, PET, tionship between task compliance and a sufficiently DTI). rewarding neuroimaging environment. Accordingly, con- Task noncompliance in pediatric functional neuroimaging ventional reward-based protocols often employ multiple Noncompliance has been recognized as a central issue sources and different kinds of rewards to encourage moti- within pediatric fMRI research [1-5], as well as clinical vation and task completion. One source is the monetary MRI procedures [6], for some time. It extends from diffi- compensation provided for participation [e.g., [19]]. Task culties entering a scanner to completing fMRI tasks, per- compliance may also be influenced with monetary rewards forming tasks accurately and remaining motionless. Many earned directly as a result of performance on an fMRI task interventions have been developed to enhance coopera- [e.g., [20]]. Researcher-identified rewards (stickers, glow- tion. These include cognitive-behavior modifications such pens, gift certificates, coloring books or brain pictures) as relaxation [7] as well as play therapy [8], observing a represent yet another major source of reward [e.g., [17]]. role model [9] and providing scanner exposure/simulation Lastly, there are social rewards, which include words of [10-12]. Reinforcement based protocols are commonly encouragement and verbal praise for working hard, good used to help children learn to minimize head motion performance and remaining in the scanner. [13-16]. There are also comprehensive packages that bring Reward-based approaches clearly help to create a posi- together many different basic techniques, providing for a tive, encouraging, and supportive environment necessary more systematic approach [e.g., [17]]. One important gap, for successful pediatric neuroimaging research. It seems however, in the pediatric imaging literature concerns why important to note, however, that for a significant number of children, especially young children and sensitive clinical task noncompliance occurs and what ways are available to populations, reward-based approaches will not be enough intervene. While many recognize task noncompliance as a to promote task compliance. Nevertheless, there still may problem, it has only received a cursory treatment in the be ways of improving or strengthening existing conven- pediatric neuroimaging literature. This is rather surprising tional reward-based approaches. Results of developmental and unfortunate given that increases in task noncompli- ance produce a greater risk that a study sample is not studies on reinforcement processes, tactics used in beha- representative of the study population. vior therapy for children and head motion training pre- Our understanding of the prevalence of task noncom- parations that tap reinforcement as a change agent offer pliance within the pediatric neuroimaging literature is some important insights into why a reward-based protocol another area of weakness. Several functional neuroima- may fail and how to improve its effectiveness. ging studies have reported that task compliance improves One of the ways a researcher-identified reward, such as with age and is higher in typically developing children sticker or trinket, may fail to maintain task compliance is relative to variety of clinical populations [3,5,14]. While that the reward does not have the capability to function this seems reasonable, the picture remains somewhat as a reinforcer, which strengthens or makes a behavior clouded because definitions of task compliance vary [18]. more likely to occur (e.g., completing an fMRI task). For example, one investigation [9] defined compliance Thus, while a subject may report that they ‘like’ or ‘want’ ‘success’ as completing an anatomical scan and at least a preselected reward, it may simply not encourage or one or more of four total scheduled tasks. By compari- maintain a target behavior. In fact, finding appropriate son, another investigation [18] defined compliance as rewards that work as reinforcers is a major component of completing a whole battery of fMRI tasks that produced effective sticker charts [21-23]. A second reason why a interpretable data for inclusion in group statistical ana- researcher-identified reward may fail is that the subject lyses. These differences in definition have several poten- may view the reward as desirable or valuable ‘now,’ but tial negative consequences for pediatric functional because it is not earned until ‘later,’ thesubjectivevalue of the reward may plummet over time, along with its neuroimaging research. The first is that reported success potentially reinforcing function. Loss of value over time rates for a particular age group or clinical population is referred to as temporal discounting and evidence from may vary markedly across investigations. The second is it developmental studies has shown that delayed rewards prevents meaningful evaluation of any intervention for noncompliance and complicates comparisons between are discounted to a greater extent in young children (6- interventions. 11 years) as compared with adolescents (12-17 years) Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 3 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 [24]. This may partly account for reported reductions in serious head injury, (d) having eye problems or difficul- task compliance in younger children. A third related rea- ties in vision not corrected by the use of glasses or con- son why the value of a researcher-identified reward may tact lenses, measured as vision of 30/20 or better with diminish is that subjects encounter mounting demands both eyes open using a hand-held eye-chart or (e) metal associated with participation, such as a remaining or devices contraindicated for MRI. All participants were motionless, remaining in the scanner for a long period recruited from community advertisements. After a and completing multiple, often effortful (and boring) detailed description of the study and before participation, tasks. With these cumulative demands/costs, reward parents gave written informed consent for their child’s value may diminish and performance breaks down. participation in the study. Children gave written Recent evidence from a large scale developmental study informed assent. All studies reported were approved by (N = 849; 4-14 years) shows motivational differences to the University of Pittsburgh Institutional Review Board. monetary reward in children as a function of age and gender, with older children and males more resistant to Primary study groups higher response costs [25]. Finally, competition among The twenty-eight subjects recruited for the study were rewards can also influence task compliance. Participation subdivided into three groups based upon their order of in a research study is essentially the choice of one recruitment. Table 1 provides demographic information rewarding activity over another. Noncompliance can for each group. Groups were constructed based upon emerge when study-rewards cannot compete with more changes made in the reward protocol and fMRI tasks to valued concurrent non-study rewards, such as visiting facilitate remaining in the scanner to complete tasks. friends, or delayed non-study rewards, such as going to Compliance in the first twelve subjects and, two later dinner after the study. additional subjects, was encouraged using a conventional In this paper, we describe how integrating tactics reward-based protocol–seebelow.Thisisthe “Reward” derived from learning based behavior therapies into con- protocol group (N = 14). A second group also received ventional reward-based protocols may help to improve or the conventional reward-based protocol and the multiple maintain task compliance. These tactics include (1) using reward compliance protocol (MRCP)—see below. This preference assessments to identify multiple subject-speci- group was designated the “MRCP” group (N = 5). Finally, fic rewards, (2) increasing reinforcement rates during the third group was designated the “MRCP Plus” group imaging by providing a reward for each task, and (3) pre- (N = 9). This latter group received the conventional senting a visual ‘road map’ during imaging that lists reward-based protocol, the MRCP and three of the seven tasks, associated rewards and progress. For brevity, this fMRI tasks were shortened in duration. This action collection of tactics will be referred to as the multiple decreased the total time needed to complete all fMRI reward compliance protocol. In what follows, we describe tasks from 66.7 min to 55.9 min, working under the idea the development and application of our approach in sev- that reducing task demand may enhance task compliance. eral groups of children that participated in an fMRI investigation. Our results showing increases in the per- Generalization test groups centage of fMRI tasks completed in several groups of Table 1 also provides demographic information and per- children provide proof of concept data for the multiple cent compliance data for several additional groups of reward compliance protocol. Additional support is pro- children from three different investigations. The first two vided by results showing our approach generalized to groups participated in two separate fMRI studies on child several additional fMRI and event-related potential inves- and adolescent depression. Both studies involved com- tigations and clinical populations (children at risk for pletingfourorfivetasks duringa90minute3TfMRI depression, children with anxiety and children with session. The first group included 34 subjects ages 10-15 depression). years with half at high familial risk for depression. The second group included 32 depressed and non-depressed Methods subjects ages 9-17 years. Generalization was further Subjects tested in a large scale NIMH funded investigation on Twenty-eight healthy children ages 9-13 years (mean childhood anxiety. Fifty-four children ages 9-13 years 11.1, SD = 1.83) participated. Participation required com- diagnosed with anxiety completed (a) a 2 hour 3T fMRI pleting a battery of clinical assessments and a 2 hr 3T session that required completing six tasks, and (b) a 2.5 functional neuroimaging session. Exclusion criteria for hour ERP session that required completing three tasks. the study included: (a) symptoms suggestive of an Axis I psychiatric disorder based on parent report on the Child Tasks and compliance measure or Adolescent Symptom Inventory-4 [26,27], (b) the exis- For the primary study groups, seven tasks were pre- tence of a major systemic medical illness, (c) a history of sented during neuroimaging in a randomized order, Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 4 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Table 1 Demographics and compliance results Grouping Groups N Age Range Mean Age (SD) Total Tasks Average Duration Compliance* Primary Group Reward group 14 9-13 yrs. 11.1 (1.78) 120 min. 7 68% MRCP group 5 9-13 yrs. 11.3 (2.15) 120 min. 7 97% MRCP Plus group 9 9-13 yrs. 11.2 (1.55) 120 min. 7 94% MRCP Combined group ^ 14 9-13 yrs. 11.2 (1.54) 120 min. 7 95% Generalization Groups+ Depression Study 1 group 34 10-15 yrs. 12.5 (1.91) 90 min. 4 100% Depression Study 2 group 32 9-17 yrs. 14.1 (1.98) 90 min. 5 100% Anxiety Study: fMRI group 1 21 9-13 yrs. 10.8 (1.32) 120 min. 6 82% fMRI group 2 33 9-13 yrs. 10.3 (1.30) 120 min. 6 86% ERP 54 9-13 yrs. 10.5 (1.31) 150 min. 3 100% ^ Pooling of MRCP and MRCP Plus groups. *Total number of fMRI or ERP tasks completed/total number of tasks. +All groups received a version of the MRCP. each separated by rest periods that lasted up to 5 min- children suggest some tactics for enhancing conventional utes. The tasks used assessed attention, reward sensitiv- reward-based protocols. Below we describe the rationale ity, threat processing and discriminated avoidance and behind the development of a multiple reward compliance approach [28]. Compliance was defined as remaining in protocol. Critical elements of the protocol include (1) the scanner during the time period a particular task was using preference assessments prior to imaging to identify presented. Each task completed was considered one multiple subject-specific rewards, (2) increasing reinfor- “success"——regardless of whether the task was spread cement rates during imaging by providing a reward for over multiple neuroimaging runs. Percent compliance each task, and (3) presenting a visual ‘road map’ during was calculated by dividing the number of tasks com- imaging that lists tasks, associated rewards and current pleted successfully (N of successes) by the total number progress. of programmed tasks (our N = 7; total possible suc- Identifying preferred rewards cesses). Defining task compliance in this way sets task One approach to adverting task noncompliance issues compliance apart from problems related to excessive related to selection of an ineffective reward(s) is to let head movement and performance accuracy—each of subjects identify their own. Clinicians developing beha- which are distinguishable by their own dependent mea- vioral treatments for typically developing children and sures. This approach also differs from other approaches those with cognitive dysfunction commonly employ ‘pre- which include head motion and task performance in ference assessments’ to identify preferred stimuli that defining success [18]. may serve as subsequent reinforcers [29]. Accordingly, the first element of the multiple reward compliance pro- Conventional reward-based protocol tocol involved identifying multiple subject-preferred The general approach we employed to encourage and rewards. The preference assessment implemented prior maintain compliance followed many conventional prac- to neuroimaging required subjects to select seven pre- tices [1-5,17] with most subjects receiving simulator ferred toys (one for each fMRI task run) from a large training (i.e., being placed in a mock scanner and hearing drawer containing small toys, such as stickers, rubber scanner noises for approximately 5 min) and all receiving balls, glowing pens...(items costing ~$1.00). This pretraining on select imaging tasks. Subjects also received approach effectively eliminated any guesswork about as part of the reward-based protocol monetary compen- reward value by using choice as an index of subjective sation for participation, a picture of their brain and verbal value. Incidentally, the order of reward selection provides support and encouragement for working hard, good per- insight into which items are viewed more favorably, formance and remaining in the scanner. potentially highlighting those rewards with greater moti- vational properties. Following selection, a preference Multiple reward compliance protocol hierarchy can be constructed by requiring children to physically order rewards from most to least preferred. As noted above, conventional reward-based protocols can fail to maintain task compliance for a variety of reasons— The resulting preference hierarchy may prove especially researcher-selected rewards may be inappropriate, dis- useful under conditions where some neuroimaging tasks counted, devalued or uncompetitive. Fortunately, basic are markedly more difficult than others and allow the and clinical research studies on reinforcement processes pairing of more difficult tasks with most preferred selec- and tactics used in learning based behavior therapies for tions to enhance motivation. Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 5 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Reducing reward delays and increasing reinforcement rate future events and provide an increased sense of control. One approach to combating decreases in reward effec- At least one published protocol has employed a version of tiveness caused by delays or cost/effort is by reducing this approach with a “virtual sticker chart” that highlights delays and increasing the number of rewards earned per progress [17]. One strength behind the sticker chart unit time. While increasing the frequency of verbal approach is the progress information it provides. However, praise, encouragement and feedback can be effective in a potentially important weakness is that stickers provided facilitating task compliance, a potentially more effective may lack rewarding or reinforcing properties for some approach is to use preference assessment results to pair subjects. selected rewards with fMRI tasks. Accordingly, the sec- ond element of the multiple reward compliance protocol Results involved pairing subject-selected rewards highlighted in Application of the multiple reward compliance protocol the preference assessment with imaging tasks and Table 1 provides demographic information and informa- informing subjects they will earn one of their chosen tion on percent task compliance for each group. In gen- rewards after completing each task. Providing multiple eral, results show a higher percentage of fMRI tasks were rewards during imaging effectively insulates rewards completed in groups that received the multiple reward against the negative effects of temporal delays and cost/ compliance protocol, providing some proof of concept effort by increasing the local rate of reinforcement (i.e., data for the approach. Panel A in Figure 2 shows task number of rewards earned per minute during a session). compliancedatafor consecutivesubjects. Percenttask In effect, the multiple reward compliance protocol compliance in the Reward group averaged 68.4% (SD = approach models “catch them being good” clinical 35.5%), the MRCP group averaged 97.1% (SD = 6.4%) approaches that stress providing high rates of reinforce- and the MRCP Plus group averaged 93.6% (SD = 10.4%). ment for appropriate behavior [21-23]. Panels B and C in Figure 2 provide an additional perspec- Visual progress display tive on our efforts to improve and maintain compliance Considerable developmental research shows age-related using the MRCP. Panel B in Figure 2 shows survival improvements in memory and information processing curves for the Reward group and pooled data from the speed and reductions in susceptibility to interference MRCP group and the MRCP Plus group (hereafter [30,31]. Within any neuroimaging investigation, children termed the ‘MRCP combined’ group). The rationale for are exposed to a wealth of information and demands that pooling groups for this analysis was to match the group can be overwhelming and aversive, prompting noncompli- size of the Reward group (N = 14), which seems appro- ance. Researchers routinely manage task information and priate given that both groups received the MRCP demands using verbal communication during imaging rest whereas the Reward group did not. Results of the analysis periods, informing subjects about current and upcoming provide a rudimentary picture of success when faced with tasks as well as providing reassurance about progress (e.g., multiple tasks. Plotted is the percentage of subjects in “You finished that task. Just three more tasks left. You are each group that completed one to all seven tasks during doing great!”). Accordingly, the third element of the multi- neuroimaging. The function reveals 90% of Reward ple reward compliance protocol involves supplementing group subjects initially completed 1-2 tasks, but as the verbal communications with a visual presentation that number of tasks increased the percentage dropped to highlights information about tasks, demands and progress. 35%. By comparison, the MRCP combined group consis- Figure 1 provides an illustration of a visual progress dis- tently completed more tasks overall. play we presented to subjects during imaging rest periods, To explore the level of compliance improvement, we which can be easily created and presented using Microsoft used a censored geometric distribution model to esti- ® ® ® Word , Powerpoint or programmed using Eprime .The mate the probability of task compliance. The analysis format presents information about the pairing of tasks revealed that 11.8% fewer participants were estimated as with subject-identified rewards and current progress. Each likely to complete each consecutive task in the Reward task is listed (under ‘Jobs’) and separated by imaging run, group compared to only 4.1% for the MRCP combined such that tasks requiring multiple imaging runs are listed group. Bootstrap analysis (10,000 iterations) suggested multiple times. In the center of the display is a progress the percentage of subjects that end participation is sig- bar the researcher can advance downward toward a ‘finish nificantly higher in Reward group than the MRCP com- line’ after each task is completed. Presentation of the visual bined group (p < 0.05), and the difference of percentage display during breaks enables subjects to quickly observe has a 95% CI of (1.1%, 15.9%). As a result, about 41.52% their progress, tasks, earned rewards and upcoming of subjects would be predicted to finish all seven tasks rewards. The information provided about upcoming tasks in the Reward group, while 74.6% of subjects would be mayalsoinsulateagainst thedevelopment of negative predicted to finish all seven tasks in the MRCP com- affective responses associated with the uncertainty of bined group. Lastly, panel C in Figure 2 shows ratings Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 6 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Figure 1 Visual progress display presented during neuroimaging rest periods. Informational display highlighting each experimental task, a progress bar and subject-identified rewards earned for completing tasks. The display was presented between imaging tasks and functioned as a supplement to researcher’s instructions and verbal praise/encouragement for participation. Tasks are listed by imaging runs, such that those requiring multiple runs are listed multiple times. Task-paired rewards, listed as ‘prizes,’ were identified prior to neuroimaging using a preference assessment procedure and were earned contingent upon task completion, regardless of performance accuracy. The progress bar descended after a task was completed. assessing how ‘enjoyable’ participation was on a four research personnel, different types of neurophysiological point post-experimental questionnaire (1 = Not at all, 3 techniques and more challenging pediatric populations, = No Feelings, 5 = A Lot). Ratings for subjects in the such as children with anxiety or depression. Results high- MRCP combined group (N= 14;M =4.2, SD= 1.05) lighting generalization of the multiple reward compliance were more favorable, but not significantly (p >0.05), protocol would provide additional proof of concept data than ratings for subjects in the Reward group (N = 12, for the approach. two cases missing; M = 3.5, SD = 1.24). To examine this issue, the multiple reward compliance protocol was applied in a number of additional fMRI and Generalization tests of the multiple reward compliance event-related potential (ERP) investigations that protocol employed different research assistants and pediatric While we did not use a randomized controlled trial, populations (e.g., children at risk for depression, children improvements in the percentage of fMRI tasks completed with depression and children with anxiety). For all stu- across subjects suggests the multiple reward compliance dies, task compliance was again defined as the percentage protocol may have some utility for encouraging children of total scheduled tasks completed. Table 1 provides to complete tasks. The relative ease and low cost of demographic information and percent compliance data employing the multiple reward compliance protocol sug- for each study. The first two fMRI studies focused on gests its major utility may be as a supplemental ‘insur- child and adolescent depression. In these studies, the ance’ policy to existing methods. One of ways to evaluate multiple reward compliance protocol was administered its utility is to determine whether the findings above without the visual display component. Both studies represent an isolated case, resulting from some aspect of involved completing four or five tasks during a 90 minute our procedure, a research assistant or some feature of the fMRI session. Results presented in Table 1 show task imaging environment. The multiple reward compliance compliance was 100%. Generalization was further tested protocol may have limited value if it cannot be shown to in a large scale NIMH funded investigation on childhood generalize to other investigations that employ different anxiety. Fifty-four children with anxiety completed (a) a Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 7 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Figure 2 Illustration of improvements in compliance in youths (9-13 years old). Panel A provides a time line of the number of fMRI tasks completed by consecutive subjects. Subjects in the Reward group received a conventional reward-based protocol to encourage compliance that consisted of monetary compensation for participation, customary verbal feedback praise/encouragement, simulator training and exposure to select tasks prior to imaging. Subjects in the multiple reward compliance protocol or MRCP group also received the reward-based protocol along with earning a subject-selected reward following completion of each fMRI task and presentation of a visual progress display (see Figure 1). The duration of three tasks was later shortened in the MRCP Plus** subjects (6-14), which decreased total time of tasks from 66.7 min to 55.9 min. Panel B shows significant differences in tasks completed between Reward and the MRCP Combined group (pooled MRCP and MRCP Plus groups). Panel C shows a slightly more favorable, but not significant, view of the experiment by the MRCP Combined group. Bars reflect standard deviations. 2 hour fMRI session that required completing six tasks, next group of 33 consecutive subjects that underwent and (b) a 2.5 hour ERP session that required completing fMRI averaged 86% (SD = 31%) task compliance. Within three tasks. Results presented in Table 1 show the first 21 this group, 27 of the 33 subjects (82%) completed five or consecutive subjects that underwent fMRI, with some six tasks. Finally, task compliance with all three ERP portion not receiving the visual display due to a proce- tasks was 100% for all 54 subjects. Overall, the consis- dural error, averaged 82% (SD = 36%) task compliance. tently high level of task compliance attained across three Within this group, 17 of the 21 subjects (81%) completed different investigations, two different methodologies five or six tasks. Results presented in Table 1 show the (fMRI, ERP) and four different pediatric populations Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 8 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 (healthy children, children at risk for depression, children protocol may lose in generality, meaning their ability to with anxiety and children with depression) offers some remediate poor task performance and excessive head preliminary support for the multiple reward compliance motion, they gain in effectiveness. protocol. Another contribution of our approach is that it high- lights the idea that task noncompliance may result from Discussion less optimally designed reward-based protocols. This approach emphasizes the importance of creating an envir- Noncompliance is a common phenomenon in pediatric onment conducive to completing study tasks which differs neuroimaging [1,2]. Recognition of the problem has in focus from other perspectives that might emphasize prompted many discussions on structuring a positive and supportive environment for children [1-5,17] as well as subject-related factors such as fear, anxiety or boredom as generated specific tactics, such as those for reducing contributors to task noncompliance. In our case, task excessive head motion [9-12]. In this paper, we focused compliance was viewed as emerging from an imaging on task noncompliance, defined as the inability to remain environment that provides adequate and sufficient reward. in a scanner to complete fMRI tasks. It is an especially The end result was an enhanced reward protocol that significant problem in pediatric functional magnetic reso- included (1) a preference assessment to identify multiple nance imaging research because increases in task non- subject-specific rewards, (2) increasing reinforcement rates compliance produces a greater risk that a study sample during imaging by providing a reward for each task, and will not be representative of the study population. Conse- (3) presenting a visual ‘road map’ during imaging that lists quently, efforts to develop procedures for improving task tasks, associated rewards and progress. It is equally impor- noncompliance address an important gap in the pediatric tant to recognize that while our findings may not directly neuroimaging literature. The approach we presented pro- support the effectiveness of the multiple reward compli- vides some heuristics for understanding noncompliance ance protocol, our rationale is supported by the behavior and demonstrated tactics for addressing tack noncompli- therapy literature and basic behavioral research on learn- ance in pediatric neuroimaging. ing processes. All of this is not to say that fear or anxiety One contribution of our approach to pediatric neuroi- are irrelevant to our understanding of noncompliance or maging was our focus on task noncompliance as an issue that our tactics should replace using relaxation techniques worth addressing apart from other forms noncompliance. or exposure training. It is merely to point out that improv- Within the pediatric imaging literature, the term non- ing reward protocols can contribute to enhancing task compliance. compliance encompasses a wide range of problems encountered by researchers and clinicians. These may include failures to enter a scanner, remain in the scanner, Limitations complete tasks, follow task instructions accurately and There were several limitations of the present investigation remain motionless. While all forms of noncompliance are that restrict the conclusions that can be drawn about the significant barriers, it seems reasonably well established multiple reward compliance protocol. The principal lim- that maintaining participation involves addressing many itation was our inability to use a randomized control trial problem areas with different types of tactics and doing so during the course of our investigation, which leaves at different times. For example, some approaches encou- results open to biases and experimenter expectancies nor- rage cooperation early in the process by having subjects mally controlled for with randomized control trial. Never- watch a video of a child completing a routine fMRI study theless, we submit that our results showing a higher in a scanner [5]. Another recommended tactic to pro- percentage of fMRI task completion by healthy children mote adherence during fMRI involves presenting a vir- provides proof of concept data for the recommended tac- tual sticker chart during rest periods to highlight current tics. Additional support was also provided by results progress [17]. Just as these examples focused on a specific showing our approach generalized to several additional aspect of participation, we focused on encouraging sub- fMRI and event-related potential investigations and clini- jects to complete fMRI tasks—measured as the percen- cal populations (children at risk for depression, children tage of total fMRI tasks completed—by using more with anxiety and children with depression). This level of preferred and more task-specific rewards. An important generalization suggests that the multiple reward compli- shared feature of these approaches and our multiple ance protocol may extend to populations with significant reward compliance protocol is that each is not designed cognitive impairments, such as children with develop- to address all forms of noncompliance, which would mental disabilities who are at increased risk of hypoxia include task performance and head motion. Such an when sedation approaches are used [32]. Another limita- expectation is too demanding. By targeting only one form tion of the present investigation concerns the relative of [potential] noncompliance for improvement, what contributions of the various components of the multiple these approaches and our multiple reward compliance reward compliance protocol (preference assessment, Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 9 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 School of Medicine, Baltimore MD, USA. Department of Behavior Analysis, increase in reward, visual display). Future research is University of North Texas, Denton TX, USA. Department of Psychology, needed that examines which component or components University of Pittsburgh, Pittsburgh PA, USA. Department of Statistics, may contribute the most to improving task compliance. University of Pittsburgh, Pittsburgh PA, USA. During our generalization tests, there was a rather sub- Authors’ contributions stantial sample of subjects that did not receive the visual Design: MS, MC, GS. Data collection: MS, AM, JS, CL, GS. Analysis: MS, GS, IS. display component but nonetheless showed high task Writing: MS, GS, MC, CL, JS, EF, RD and NR All authors read and approved the final manuscript. compliance. This was limited to the two fMRI studies on child and adolescent depression (N = 34 and N = 32) Competing interests and a portion of the first group of children participating The authors declare that they have no competing interests. in the childhood anxiety study (N = 21). Findings show- Received: 9 July 2010 Accepted: 6 May 2011 Published: 6 May 2011 ing high task compliance without the visual display com- ponent suggest identifying and providing rewards References contingent upon task completion may be relatively more 1. Church JA, Petersen SE, Schlaggar BL: The “Task B problem” and other considerations in developmental functional neuroimaging. Hum Brain important than the information supplied by the road Mapp 2010, 31:852-62. map. Another limitation of the present investigation is 2. Kotsoni E, Byrd D, Casey BJ: Special considerations for functional that it remains unclear what subject variables or aspects magnetic resonance imaging of pediatric populations. J Magn Reson Imaging 2006, 23:877-886. of participation contributed to task noncompliance. 3. O’Shaughnessy ES, Berl MM, Moore EN, Gaillard WD: Pediatric Functional Researchers and clinicians might benefit from gathering Magnetic Resonance Imaging (fMRI): Issues and Applications. J Child information during debriefing about the reason(s) for Neurology 2008, 23:791-801. 4. Poldrack RA, Paré-Blagoev EJ, Grant PE: Pediatric functional magnetic early termination. Children’s reports may contain valuable resonance imaging: progress and challenges. Top Magn Reson Imaging information about how to structure the environment or 2002, 13:61-70. tasks in ways that encourage cooperation. 5. Wilke M, Holland SK, Myseros JS, Schmithorst VJ, Ball WS Jr: Functional magnetic resonance imaging in pediatrics. Neuropediatrics 2003, 34:225-33. Conclusions 6. Tyc VL, Fairclough D, Fletcher B, Leigh L, Mulhern RK: Children’s distress Functional magnetic resonance imaging is increasingly during magnetic resonance imaging procedures. Child Health Care 1995, 24:5-19. being used to advance our understanding of normal and 7. Lukins R, Davan IG, Drummond PD: A cognitive behavioural approach to pathological brain function and development in children preventing anxiety during magnetic resonance imaging. J Behav Ther Exp and adolescents. Noncompliance involving an inability to Psychiatry 1997, 28:97-104. 8. Pressdee D, May L, Eastman E, Grier D: The use of play therapy in the remain in a scanner to complete fMRI tasks is one com- preparation of children undergoing MR imaging. Clin Radiol 1997, mon and significant problem. Consequently, researchers 52:945-947. and clinicians devote considerable effort to developing a 9. Byars AW, Holland SK, Strawsburg RH, Bommer W, Dunn RS, Schmithorst VJ, Plante E: Practical aspects of conducting large-scale functional magnetic supportive, positive environment. We proposed that some resonance imaging studies in children. J Child Neurol 2002, 17:885-890. forms of task noncompliance may emerge from less than 10. Rosenberg DR, Sweeney JA, Gillen JS, Kim J, Varanelli MJ, O’Hearn KM, optimal reward protocols. Our findings suggest that Erb PA, Davis D, Thulborn KR: Magnetic resonance imaging of children without sedation: preparation with simulation. J Am Acad Child Adolesc increasing our attention to how rewards are selected and Psychiatry 1997, 36:853-859. delivered may aid cooperation with completing fMRI tasks. 11. de Amorim e Silva CJ, Mackenzie A, Hallowell LM, Stewart SE, Ditchfield MR: The approach presented and preliminary findings contri- Practice MRI: reducing the need for sedation and general anaesthesia in children undergoing MRI. Australas Radiol 2006, 50:319-323. bute to the pediatric neuroimaging literature by providing 12. Epstein JN, Casey BJ, Tonev ST, Davidson M, Reiss AL, Garrett A, a useful way to conceptualize and measure task noncom- Hinshaw SP, Greenhill LL, Vitolo A, Kotler LA, Jarrett MA, Spicer J: pliance and a set of cost effective tactics for improving the Assessment and prevention of head motion during imaging of patients with attention deficit hyperactivity disorder. Psychiatry Res Neuroimaging effectiveness of common reward-based protocols. 2007, 155:75-82. 13. Slifer KJ, Cataldo MF, Cataldo MD, Llorente AM, Gerson AC: Behavior Appl Behav Anal analysis of motion control for pediatric neuroimaging. J List of abbreviations used 1993, 26:469-470. MRCP: Multiple reward compliance protocol 14. Slifer KJ, Bucholtz JD, Cataldo MD: Behavioral training of motion control in young children undergoing radiation treatment without sedation. J Acknowledgements Pediatr Oncol Nurs 1994, 11:55-63. This paper was made possible with the support of NIH grant funding 15. Slifer KJ: A video system to help children cooperate with motion control MH41712 (R.D.), P50 MH080215 (N.R.), K02 MH082998 (G.S.), K01 MH083001 for radiation treatment without sedation. J Pediatr Oncol Nurs 1996, (C.L.), KO1 MHO73077 (J.S.), DA024144-01 (J.S./R.D.), a NARSAD Young 13:91-97. Investigator Award (J.S.), the Beatrice H. Barrett Research Endowment (M.S.) 16. Slifer KJ, Koontz KL, Cataldo MF: Operant-contingency-based preparation and the efforts of Lindsay Proud, Rachel Kolko, Erika Joyce, and Emily of children for functional magnetic resonance imaging. J Appl Behav Anal Yarrison. 2002, 35:191-194. 17. Raschle NM, Lee M, Buechler R, Christodoulou JA, Chang M, Vakil M, Author details Stering PL, Gaab N: Making MR Imaging Child’s Play - Pediatric Department of Psychiatry, University of Pittsburgh School of Medicine, Neuroimaging Protocol, Guidelines and Procedure. JoVE 2009, 29: [http:// Pittsburgh PA, USA. Department of Behavioral Psychology, Kennedy Krieger www.jove.com/index/details.stp?id=1309]. Institute, Baltimore MD, USA. Department of Psychiatry, Johns Hopkins Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 10 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 18. Yerys BE, Jankowski KF, Shook D, Rosenberger LR, Barnes KA, Berl MM, Ritzl EK, Vanmeter J, Vaidya CJ, Gaillard WD: The fMRI success rate of children and adolescents: Typical development, epilepsy, attention deficit/hyperactivity disorder, and autism spectrum disorders. Hum Brain Mapp 2009, 30:3426-3435. 19. Marsh R, Zhu H, Schultz RT, Quackenbush G, Royal J, Skudlarski P, Peterson BS: A developmental fMRI study of self-regulatory control. Hum Brain Mapp 2006, 27:848-863. 20. Bjork JM, Smith AR, Chen G, Hommer DW: Adolescents, adults and rewards: comparing motivational neurocircuitry recruitment using fMRI. PLoS One 2010, 5:e11440. 21. Christophersen ER, Mortweet SL: Parenting that Works: Building Skills that Last a Lifetime. American Psychological Association Books, Washington DC; 22. Christophersen ER: Little People: Guidelines for Common Sense Child Rearing. Westport Publishers, Inc, Kansas City MO, 3 1998. 23. Kazdin AE, Rotella C: The Kazdin Method for Parenting the Defiant Child. Mariner Books, Houghton Mifflen Harccourt Publishers, New York; 2008. 24. Scheres A, Dijkstra M, Ainslie E, Balkan J, Reynolds B, Sonuga-Barke E, Castellanos FX: Temporal and probabilistic discounting of rewards in children and adolescents: effects of age and ADHD symptoms. Neuropsychologia 2006, 44:2092-2103. 25. Chelonis JJ, Osborn SA, Gravelin CR, Paule MG: Assessing motivation in children using a progressive ratio task. Proceedings of the Assoc for Behavior Analysis 36th Annual Conference 2010. 26. Gadow KD, Sprafkin J: Adolescent Symptom Inventory – 4: Norms Manual. Checkmate Plus, New York; 1998. 27. Gadow KD, Sprafkin J: Child Symptom Inventory – 4: Screening Manual. Checkmate Plus, New York; 1998. 28. Schlund MW, Siegle GJ, Ladouceur CD, Silk JS, Cataldo MF, Forbes EE, Dahl RE, Ryan ND: Nothing to fear? Neural systems supporting avoidance behavior in healthy youths. Neuroimage 2010, 52:710-719. 29. Hagopian LP, Long ES, Rush KS: Preference assessment procedures for individuals with developmental disabilities. Behav Modif 2004, 28:668-677. 30. Jenkins L, Myerson J, Hale S, Fry AF: Individual and developmental differences in working memory across the life span. Psychon Bull Rev 1999, 6:28-40. 31. Fry AF, Hale S: Relationships among processing speed, working memory, and fluid intelligence in children. Biol Psychol 2000, 54:1-34. 32. Kannikeswaran N, Mahajan PV, Sethuraman U, Groebe A, Chen X: Sedation medication received and adverse events related to sedation for brain MRI in children with and without developmental disabilities. Paediatr Anaesth 2009, 19:250-256. doi:10.1186/1744-9081-7-10 Cite this article as: Schlund et al.: Pediatric functional magnetic resonance neuroimaging: tactics for encouraging task compliance. Behavioral and Brain Functions 2011 7:10. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral and Brain Functions Springer Journals

Pediatric functional magnetic resonance neuroimaging: tactics for encouraging task compliance

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/pediatric-functional-magnetic-resonance-neuroimaging-tactics-for-WW4YTNwHOa

References (71)

Publisher
Springer Journals
Copyright
Copyright © 2011 by Schlund et al; licensee BioMed Central Ltd.
Subject
Biomedicine; Neurosciences; Neurology; Behavioral Therapy; Psychiatry
eISSN
1744-9081
DOI
10.1186/1744-9081-7-10
pmid
21548928
Publisher site
See Article on Publisher Site

Abstract

Background: Neuroimaging technology has afforded advances in our understanding of normal and pathological brain function and development in children and adolescents. However, noncompliance involving the inability to remain in the magnetic resonance imaging (MRI) scanner to complete tasks is one common and significant problem. Task noncompliance is an especially significant problem in pediatric functional magnetic resonance imaging (fMRI) research because increases in noncompliance produces a greater risk that a study sample will not be representative of the study population. Method: In this preliminary investigation, we describe the development and application of an approach for increasing the number of fMRI tasks children complete during neuroimaging. Twenty-eight healthy children ages 9- 13 years participated. Generalization of the approach was examined in additional fMRI and event-related potential investigations with children at risk for depression, children with anxiety and children with depression (N = 120). Essential features of the approach include a preference assessment for identifying multiple individualized rewards, increasing reinforcement rates during imaging by pairing tasks with chosen rewards and presenting a visual ‘road map’ listing tasks, rewards and current progress. Results: Our results showing a higher percentage of fMRI task completion by healthy children provides proof of concept data for the recommended tactics. Additional support was provided by results showing our approach generalized to several additional fMRI and event-related potential investigations and clinical populations. Discussion: We proposed that some forms of task noncompliance may emerge from less than optimal reward protocols. While our findings may not directly support the effectiveness of the multiple reward compliance protocol, increased attention to how rewards are selected and delivered may aid cooperation with completing fMRI tasks Conclusion: The proposed approach contributes to the pediatric neuroimaging literature by providing a useful way to conceptualize and measure task noncompliance and by providing simple cost effective tactics for improving the effectiveness of common reward-based protocols. Background involving an inability to remain in a scanner to complete Functional magnetic resonance imaging is increasingly fMRI tasks is one common and significant problem. This being used to advance our understanding of normal and paper describes the development and application of an pathological brain function and development in children approach we believe may improve the effectiveness of con- and adolescents. However, there are a number of challenges ventional reward-based approaches used to encourage task that clinicians and researchers encounter. Noncompliance compliance. We discuss some issues surrounding task non- compliance and offer that integrating tactics derived from learning based behavior therapies into conventional * Correspondence: schlund@kennedykrieger.org 1 reward-based protocols may help encourage compliance. Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh PA, USA Using a case study design, preliminary results show Full list of author information is available at the end of the article © 2011 Schlund et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 2 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 improvements in the number of study tasks completed in Conventional reward-based protocols in pediatric healthy children and those with clinical disorders, providing neuroimaging proof of concept data for the recommended tactics. While In all pediatric functional neuroimaging studies, research- the present discussion focuses primarily on task compliance ers and clinicians use rewards to encourage and maintain during functional neuroimaging investigations, some of the task compliance. This highlights an important tie to rein- forcement learning theories. The widespread application problems discussed and recommendations may generalize of reward protocols also highlights recognition of the rela- to other forms of neuroimaging (e.g., clinical MRI, PET, tionship between task compliance and a sufficiently DTI). rewarding neuroimaging environment. Accordingly, con- Task noncompliance in pediatric functional neuroimaging ventional reward-based protocols often employ multiple Noncompliance has been recognized as a central issue sources and different kinds of rewards to encourage moti- within pediatric fMRI research [1-5], as well as clinical vation and task completion. One source is the monetary MRI procedures [6], for some time. It extends from diffi- compensation provided for participation [e.g., [19]]. Task culties entering a scanner to completing fMRI tasks, per- compliance may also be influenced with monetary rewards forming tasks accurately and remaining motionless. Many earned directly as a result of performance on an fMRI task interventions have been developed to enhance coopera- [e.g., [20]]. Researcher-identified rewards (stickers, glow- tion. These include cognitive-behavior modifications such pens, gift certificates, coloring books or brain pictures) as relaxation [7] as well as play therapy [8], observing a represent yet another major source of reward [e.g., [17]]. role model [9] and providing scanner exposure/simulation Lastly, there are social rewards, which include words of [10-12]. Reinforcement based protocols are commonly encouragement and verbal praise for working hard, good used to help children learn to minimize head motion performance and remaining in the scanner. [13-16]. There are also comprehensive packages that bring Reward-based approaches clearly help to create a posi- together many different basic techniques, providing for a tive, encouraging, and supportive environment necessary more systematic approach [e.g., [17]]. One important gap, for successful pediatric neuroimaging research. It seems however, in the pediatric imaging literature concerns why important to note, however, that for a significant number of children, especially young children and sensitive clinical task noncompliance occurs and what ways are available to populations, reward-based approaches will not be enough intervene. While many recognize task noncompliance as a to promote task compliance. Nevertheless, there still may problem, it has only received a cursory treatment in the be ways of improving or strengthening existing conven- pediatric neuroimaging literature. This is rather surprising tional reward-based approaches. Results of developmental and unfortunate given that increases in task noncompli- ance produce a greater risk that a study sample is not studies on reinforcement processes, tactics used in beha- representative of the study population. vior therapy for children and head motion training pre- Our understanding of the prevalence of task noncom- parations that tap reinforcement as a change agent offer pliance within the pediatric neuroimaging literature is some important insights into why a reward-based protocol another area of weakness. Several functional neuroima- may fail and how to improve its effectiveness. ging studies have reported that task compliance improves One of the ways a researcher-identified reward, such as with age and is higher in typically developing children sticker or trinket, may fail to maintain task compliance is relative to variety of clinical populations [3,5,14]. While that the reward does not have the capability to function this seems reasonable, the picture remains somewhat as a reinforcer, which strengthens or makes a behavior clouded because definitions of task compliance vary [18]. more likely to occur (e.g., completing an fMRI task). For example, one investigation [9] defined compliance Thus, while a subject may report that they ‘like’ or ‘want’ ‘success’ as completing an anatomical scan and at least a preselected reward, it may simply not encourage or one or more of four total scheduled tasks. By compari- maintain a target behavior. In fact, finding appropriate son, another investigation [18] defined compliance as rewards that work as reinforcers is a major component of completing a whole battery of fMRI tasks that produced effective sticker charts [21-23]. A second reason why a interpretable data for inclusion in group statistical ana- researcher-identified reward may fail is that the subject lyses. These differences in definition have several poten- may view the reward as desirable or valuable ‘now,’ but tial negative consequences for pediatric functional because it is not earned until ‘later,’ thesubjectivevalue of the reward may plummet over time, along with its neuroimaging research. The first is that reported success potentially reinforcing function. Loss of value over time rates for a particular age group or clinical population is referred to as temporal discounting and evidence from may vary markedly across investigations. The second is it developmental studies has shown that delayed rewards prevents meaningful evaluation of any intervention for noncompliance and complicates comparisons between are discounted to a greater extent in young children (6- interventions. 11 years) as compared with adolescents (12-17 years) Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 3 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 [24]. This may partly account for reported reductions in serious head injury, (d) having eye problems or difficul- task compliance in younger children. A third related rea- ties in vision not corrected by the use of glasses or con- son why the value of a researcher-identified reward may tact lenses, measured as vision of 30/20 or better with diminish is that subjects encounter mounting demands both eyes open using a hand-held eye-chart or (e) metal associated with participation, such as a remaining or devices contraindicated for MRI. All participants were motionless, remaining in the scanner for a long period recruited from community advertisements. After a and completing multiple, often effortful (and boring) detailed description of the study and before participation, tasks. With these cumulative demands/costs, reward parents gave written informed consent for their child’s value may diminish and performance breaks down. participation in the study. Children gave written Recent evidence from a large scale developmental study informed assent. All studies reported were approved by (N = 849; 4-14 years) shows motivational differences to the University of Pittsburgh Institutional Review Board. monetary reward in children as a function of age and gender, with older children and males more resistant to Primary study groups higher response costs [25]. Finally, competition among The twenty-eight subjects recruited for the study were rewards can also influence task compliance. Participation subdivided into three groups based upon their order of in a research study is essentially the choice of one recruitment. Table 1 provides demographic information rewarding activity over another. Noncompliance can for each group. Groups were constructed based upon emerge when study-rewards cannot compete with more changes made in the reward protocol and fMRI tasks to valued concurrent non-study rewards, such as visiting facilitate remaining in the scanner to complete tasks. friends, or delayed non-study rewards, such as going to Compliance in the first twelve subjects and, two later dinner after the study. additional subjects, was encouraged using a conventional In this paper, we describe how integrating tactics reward-based protocol–seebelow.Thisisthe “Reward” derived from learning based behavior therapies into con- protocol group (N = 14). A second group also received ventional reward-based protocols may help to improve or the conventional reward-based protocol and the multiple maintain task compliance. These tactics include (1) using reward compliance protocol (MRCP)—see below. This preference assessments to identify multiple subject-speci- group was designated the “MRCP” group (N = 5). Finally, fic rewards, (2) increasing reinforcement rates during the third group was designated the “MRCP Plus” group imaging by providing a reward for each task, and (3) pre- (N = 9). This latter group received the conventional senting a visual ‘road map’ during imaging that lists reward-based protocol, the MRCP and three of the seven tasks, associated rewards and progress. For brevity, this fMRI tasks were shortened in duration. This action collection of tactics will be referred to as the multiple decreased the total time needed to complete all fMRI reward compliance protocol. In what follows, we describe tasks from 66.7 min to 55.9 min, working under the idea the development and application of our approach in sev- that reducing task demand may enhance task compliance. eral groups of children that participated in an fMRI investigation. Our results showing increases in the per- Generalization test groups centage of fMRI tasks completed in several groups of Table 1 also provides demographic information and per- children provide proof of concept data for the multiple cent compliance data for several additional groups of reward compliance protocol. Additional support is pro- children from three different investigations. The first two vided by results showing our approach generalized to groups participated in two separate fMRI studies on child several additional fMRI and event-related potential inves- and adolescent depression. Both studies involved com- tigations and clinical populations (children at risk for pletingfourorfivetasks duringa90minute3TfMRI depression, children with anxiety and children with session. The first group included 34 subjects ages 10-15 depression). years with half at high familial risk for depression. The second group included 32 depressed and non-depressed Methods subjects ages 9-17 years. Generalization was further Subjects tested in a large scale NIMH funded investigation on Twenty-eight healthy children ages 9-13 years (mean childhood anxiety. Fifty-four children ages 9-13 years 11.1, SD = 1.83) participated. Participation required com- diagnosed with anxiety completed (a) a 2 hour 3T fMRI pleting a battery of clinical assessments and a 2 hr 3T session that required completing six tasks, and (b) a 2.5 functional neuroimaging session. Exclusion criteria for hour ERP session that required completing three tasks. the study included: (a) symptoms suggestive of an Axis I psychiatric disorder based on parent report on the Child Tasks and compliance measure or Adolescent Symptom Inventory-4 [26,27], (b) the exis- For the primary study groups, seven tasks were pre- tence of a major systemic medical illness, (c) a history of sented during neuroimaging in a randomized order, Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 4 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Table 1 Demographics and compliance results Grouping Groups N Age Range Mean Age (SD) Total Tasks Average Duration Compliance* Primary Group Reward group 14 9-13 yrs. 11.1 (1.78) 120 min. 7 68% MRCP group 5 9-13 yrs. 11.3 (2.15) 120 min. 7 97% MRCP Plus group 9 9-13 yrs. 11.2 (1.55) 120 min. 7 94% MRCP Combined group ^ 14 9-13 yrs. 11.2 (1.54) 120 min. 7 95% Generalization Groups+ Depression Study 1 group 34 10-15 yrs. 12.5 (1.91) 90 min. 4 100% Depression Study 2 group 32 9-17 yrs. 14.1 (1.98) 90 min. 5 100% Anxiety Study: fMRI group 1 21 9-13 yrs. 10.8 (1.32) 120 min. 6 82% fMRI group 2 33 9-13 yrs. 10.3 (1.30) 120 min. 6 86% ERP 54 9-13 yrs. 10.5 (1.31) 150 min. 3 100% ^ Pooling of MRCP and MRCP Plus groups. *Total number of fMRI or ERP tasks completed/total number of tasks. +All groups received a version of the MRCP. each separated by rest periods that lasted up to 5 min- children suggest some tactics for enhancing conventional utes. The tasks used assessed attention, reward sensitiv- reward-based protocols. Below we describe the rationale ity, threat processing and discriminated avoidance and behind the development of a multiple reward compliance approach [28]. Compliance was defined as remaining in protocol. Critical elements of the protocol include (1) the scanner during the time period a particular task was using preference assessments prior to imaging to identify presented. Each task completed was considered one multiple subject-specific rewards, (2) increasing reinfor- “success"——regardless of whether the task was spread cement rates during imaging by providing a reward for over multiple neuroimaging runs. Percent compliance each task, and (3) presenting a visual ‘road map’ during was calculated by dividing the number of tasks com- imaging that lists tasks, associated rewards and current pleted successfully (N of successes) by the total number progress. of programmed tasks (our N = 7; total possible suc- Identifying preferred rewards cesses). Defining task compliance in this way sets task One approach to adverting task noncompliance issues compliance apart from problems related to excessive related to selection of an ineffective reward(s) is to let head movement and performance accuracy—each of subjects identify their own. Clinicians developing beha- which are distinguishable by their own dependent mea- vioral treatments for typically developing children and sures. This approach also differs from other approaches those with cognitive dysfunction commonly employ ‘pre- which include head motion and task performance in ference assessments’ to identify preferred stimuli that defining success [18]. may serve as subsequent reinforcers [29]. Accordingly, the first element of the multiple reward compliance pro- Conventional reward-based protocol tocol involved identifying multiple subject-preferred The general approach we employed to encourage and rewards. The preference assessment implemented prior maintain compliance followed many conventional prac- to neuroimaging required subjects to select seven pre- tices [1-5,17] with most subjects receiving simulator ferred toys (one for each fMRI task run) from a large training (i.e., being placed in a mock scanner and hearing drawer containing small toys, such as stickers, rubber scanner noises for approximately 5 min) and all receiving balls, glowing pens...(items costing ~$1.00). This pretraining on select imaging tasks. Subjects also received approach effectively eliminated any guesswork about as part of the reward-based protocol monetary compen- reward value by using choice as an index of subjective sation for participation, a picture of their brain and verbal value. Incidentally, the order of reward selection provides support and encouragement for working hard, good per- insight into which items are viewed more favorably, formance and remaining in the scanner. potentially highlighting those rewards with greater moti- vational properties. Following selection, a preference Multiple reward compliance protocol hierarchy can be constructed by requiring children to physically order rewards from most to least preferred. As noted above, conventional reward-based protocols can fail to maintain task compliance for a variety of reasons— The resulting preference hierarchy may prove especially researcher-selected rewards may be inappropriate, dis- useful under conditions where some neuroimaging tasks counted, devalued or uncompetitive. Fortunately, basic are markedly more difficult than others and allow the and clinical research studies on reinforcement processes pairing of more difficult tasks with most preferred selec- and tactics used in learning based behavior therapies for tions to enhance motivation. Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 5 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Reducing reward delays and increasing reinforcement rate future events and provide an increased sense of control. One approach to combating decreases in reward effec- At least one published protocol has employed a version of tiveness caused by delays or cost/effort is by reducing this approach with a “virtual sticker chart” that highlights delays and increasing the number of rewards earned per progress [17]. One strength behind the sticker chart unit time. While increasing the frequency of verbal approach is the progress information it provides. However, praise, encouragement and feedback can be effective in a potentially important weakness is that stickers provided facilitating task compliance, a potentially more effective may lack rewarding or reinforcing properties for some approach is to use preference assessment results to pair subjects. selected rewards with fMRI tasks. Accordingly, the sec- ond element of the multiple reward compliance protocol Results involved pairing subject-selected rewards highlighted in Application of the multiple reward compliance protocol the preference assessment with imaging tasks and Table 1 provides demographic information and informa- informing subjects they will earn one of their chosen tion on percent task compliance for each group. In gen- rewards after completing each task. Providing multiple eral, results show a higher percentage of fMRI tasks were rewards during imaging effectively insulates rewards completed in groups that received the multiple reward against the negative effects of temporal delays and cost/ compliance protocol, providing some proof of concept effort by increasing the local rate of reinforcement (i.e., data for the approach. Panel A in Figure 2 shows task number of rewards earned per minute during a session). compliancedatafor consecutivesubjects. Percenttask In effect, the multiple reward compliance protocol compliance in the Reward group averaged 68.4% (SD = approach models “catch them being good” clinical 35.5%), the MRCP group averaged 97.1% (SD = 6.4%) approaches that stress providing high rates of reinforce- and the MRCP Plus group averaged 93.6% (SD = 10.4%). ment for appropriate behavior [21-23]. Panels B and C in Figure 2 provide an additional perspec- Visual progress display tive on our efforts to improve and maintain compliance Considerable developmental research shows age-related using the MRCP. Panel B in Figure 2 shows survival improvements in memory and information processing curves for the Reward group and pooled data from the speed and reductions in susceptibility to interference MRCP group and the MRCP Plus group (hereafter [30,31]. Within any neuroimaging investigation, children termed the ‘MRCP combined’ group). The rationale for are exposed to a wealth of information and demands that pooling groups for this analysis was to match the group can be overwhelming and aversive, prompting noncompli- size of the Reward group (N = 14), which seems appro- ance. Researchers routinely manage task information and priate given that both groups received the MRCP demands using verbal communication during imaging rest whereas the Reward group did not. Results of the analysis periods, informing subjects about current and upcoming provide a rudimentary picture of success when faced with tasks as well as providing reassurance about progress (e.g., multiple tasks. Plotted is the percentage of subjects in “You finished that task. Just three more tasks left. You are each group that completed one to all seven tasks during doing great!”). Accordingly, the third element of the multi- neuroimaging. The function reveals 90% of Reward ple reward compliance protocol involves supplementing group subjects initially completed 1-2 tasks, but as the verbal communications with a visual presentation that number of tasks increased the percentage dropped to highlights information about tasks, demands and progress. 35%. By comparison, the MRCP combined group consis- Figure 1 provides an illustration of a visual progress dis- tently completed more tasks overall. play we presented to subjects during imaging rest periods, To explore the level of compliance improvement, we which can be easily created and presented using Microsoft used a censored geometric distribution model to esti- ® ® ® Word , Powerpoint or programmed using Eprime .The mate the probability of task compliance. The analysis format presents information about the pairing of tasks revealed that 11.8% fewer participants were estimated as with subject-identified rewards and current progress. Each likely to complete each consecutive task in the Reward task is listed (under ‘Jobs’) and separated by imaging run, group compared to only 4.1% for the MRCP combined such that tasks requiring multiple imaging runs are listed group. Bootstrap analysis (10,000 iterations) suggested multiple times. In the center of the display is a progress the percentage of subjects that end participation is sig- bar the researcher can advance downward toward a ‘finish nificantly higher in Reward group than the MRCP com- line’ after each task is completed. Presentation of the visual bined group (p < 0.05), and the difference of percentage display during breaks enables subjects to quickly observe has a 95% CI of (1.1%, 15.9%). As a result, about 41.52% their progress, tasks, earned rewards and upcoming of subjects would be predicted to finish all seven tasks rewards. The information provided about upcoming tasks in the Reward group, while 74.6% of subjects would be mayalsoinsulateagainst thedevelopment of negative predicted to finish all seven tasks in the MRCP com- affective responses associated with the uncertainty of bined group. Lastly, panel C in Figure 2 shows ratings Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 6 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Figure 1 Visual progress display presented during neuroimaging rest periods. Informational display highlighting each experimental task, a progress bar and subject-identified rewards earned for completing tasks. The display was presented between imaging tasks and functioned as a supplement to researcher’s instructions and verbal praise/encouragement for participation. Tasks are listed by imaging runs, such that those requiring multiple runs are listed multiple times. Task-paired rewards, listed as ‘prizes,’ were identified prior to neuroimaging using a preference assessment procedure and were earned contingent upon task completion, regardless of performance accuracy. The progress bar descended after a task was completed. assessing how ‘enjoyable’ participation was on a four research personnel, different types of neurophysiological point post-experimental questionnaire (1 = Not at all, 3 techniques and more challenging pediatric populations, = No Feelings, 5 = A Lot). Ratings for subjects in the such as children with anxiety or depression. Results high- MRCP combined group (N= 14;M =4.2, SD= 1.05) lighting generalization of the multiple reward compliance were more favorable, but not significantly (p >0.05), protocol would provide additional proof of concept data than ratings for subjects in the Reward group (N = 12, for the approach. two cases missing; M = 3.5, SD = 1.24). To examine this issue, the multiple reward compliance protocol was applied in a number of additional fMRI and Generalization tests of the multiple reward compliance event-related potential (ERP) investigations that protocol employed different research assistants and pediatric While we did not use a randomized controlled trial, populations (e.g., children at risk for depression, children improvements in the percentage of fMRI tasks completed with depression and children with anxiety). For all stu- across subjects suggests the multiple reward compliance dies, task compliance was again defined as the percentage protocol may have some utility for encouraging children of total scheduled tasks completed. Table 1 provides to complete tasks. The relative ease and low cost of demographic information and percent compliance data employing the multiple reward compliance protocol sug- for each study. The first two fMRI studies focused on gests its major utility may be as a supplemental ‘insur- child and adolescent depression. In these studies, the ance’ policy to existing methods. One of ways to evaluate multiple reward compliance protocol was administered its utility is to determine whether the findings above without the visual display component. Both studies represent an isolated case, resulting from some aspect of involved completing four or five tasks during a 90 minute our procedure, a research assistant or some feature of the fMRI session. Results presented in Table 1 show task imaging environment. The multiple reward compliance compliance was 100%. Generalization was further tested protocol may have limited value if it cannot be shown to in a large scale NIMH funded investigation on childhood generalize to other investigations that employ different anxiety. Fifty-four children with anxiety completed (a) a Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 7 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 Figure 2 Illustration of improvements in compliance in youths (9-13 years old). Panel A provides a time line of the number of fMRI tasks completed by consecutive subjects. Subjects in the Reward group received a conventional reward-based protocol to encourage compliance that consisted of monetary compensation for participation, customary verbal feedback praise/encouragement, simulator training and exposure to select tasks prior to imaging. Subjects in the multiple reward compliance protocol or MRCP group also received the reward-based protocol along with earning a subject-selected reward following completion of each fMRI task and presentation of a visual progress display (see Figure 1). The duration of three tasks was later shortened in the MRCP Plus** subjects (6-14), which decreased total time of tasks from 66.7 min to 55.9 min. Panel B shows significant differences in tasks completed between Reward and the MRCP Combined group (pooled MRCP and MRCP Plus groups). Panel C shows a slightly more favorable, but not significant, view of the experiment by the MRCP Combined group. Bars reflect standard deviations. 2 hour fMRI session that required completing six tasks, next group of 33 consecutive subjects that underwent and (b) a 2.5 hour ERP session that required completing fMRI averaged 86% (SD = 31%) task compliance. Within three tasks. Results presented in Table 1 show the first 21 this group, 27 of the 33 subjects (82%) completed five or consecutive subjects that underwent fMRI, with some six tasks. Finally, task compliance with all three ERP portion not receiving the visual display due to a proce- tasks was 100% for all 54 subjects. Overall, the consis- dural error, averaged 82% (SD = 36%) task compliance. tently high level of task compliance attained across three Within this group, 17 of the 21 subjects (81%) completed different investigations, two different methodologies five or six tasks. Results presented in Table 1 show the (fMRI, ERP) and four different pediatric populations Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 8 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 (healthy children, children at risk for depression, children protocol may lose in generality, meaning their ability to with anxiety and children with depression) offers some remediate poor task performance and excessive head preliminary support for the multiple reward compliance motion, they gain in effectiveness. protocol. Another contribution of our approach is that it high- lights the idea that task noncompliance may result from Discussion less optimally designed reward-based protocols. This approach emphasizes the importance of creating an envir- Noncompliance is a common phenomenon in pediatric onment conducive to completing study tasks which differs neuroimaging [1,2]. Recognition of the problem has in focus from other perspectives that might emphasize prompted many discussions on structuring a positive and supportive environment for children [1-5,17] as well as subject-related factors such as fear, anxiety or boredom as generated specific tactics, such as those for reducing contributors to task noncompliance. In our case, task excessive head motion [9-12]. In this paper, we focused compliance was viewed as emerging from an imaging on task noncompliance, defined as the inability to remain environment that provides adequate and sufficient reward. in a scanner to complete fMRI tasks. It is an especially The end result was an enhanced reward protocol that significant problem in pediatric functional magnetic reso- included (1) a preference assessment to identify multiple nance imaging research because increases in task non- subject-specific rewards, (2) increasing reinforcement rates compliance produces a greater risk that a study sample during imaging by providing a reward for each task, and will not be representative of the study population. Conse- (3) presenting a visual ‘road map’ during imaging that lists quently, efforts to develop procedures for improving task tasks, associated rewards and progress. It is equally impor- noncompliance address an important gap in the pediatric tant to recognize that while our findings may not directly neuroimaging literature. The approach we presented pro- support the effectiveness of the multiple reward compli- vides some heuristics for understanding noncompliance ance protocol, our rationale is supported by the behavior and demonstrated tactics for addressing tack noncompli- therapy literature and basic behavioral research on learn- ance in pediatric neuroimaging. ing processes. All of this is not to say that fear or anxiety One contribution of our approach to pediatric neuroi- are irrelevant to our understanding of noncompliance or maging was our focus on task noncompliance as an issue that our tactics should replace using relaxation techniques worth addressing apart from other forms noncompliance. or exposure training. It is merely to point out that improv- Within the pediatric imaging literature, the term non- ing reward protocols can contribute to enhancing task compliance. compliance encompasses a wide range of problems encountered by researchers and clinicians. These may include failures to enter a scanner, remain in the scanner, Limitations complete tasks, follow task instructions accurately and There were several limitations of the present investigation remain motionless. While all forms of noncompliance are that restrict the conclusions that can be drawn about the significant barriers, it seems reasonably well established multiple reward compliance protocol. The principal lim- that maintaining participation involves addressing many itation was our inability to use a randomized control trial problem areas with different types of tactics and doing so during the course of our investigation, which leaves at different times. For example, some approaches encou- results open to biases and experimenter expectancies nor- rage cooperation early in the process by having subjects mally controlled for with randomized control trial. Never- watch a video of a child completing a routine fMRI study theless, we submit that our results showing a higher in a scanner [5]. Another recommended tactic to pro- percentage of fMRI task completion by healthy children mote adherence during fMRI involves presenting a vir- provides proof of concept data for the recommended tac- tual sticker chart during rest periods to highlight current tics. Additional support was also provided by results progress [17]. Just as these examples focused on a specific showing our approach generalized to several additional aspect of participation, we focused on encouraging sub- fMRI and event-related potential investigations and clini- jects to complete fMRI tasks—measured as the percen- cal populations (children at risk for depression, children tage of total fMRI tasks completed—by using more with anxiety and children with depression). This level of preferred and more task-specific rewards. An important generalization suggests that the multiple reward compli- shared feature of these approaches and our multiple ance protocol may extend to populations with significant reward compliance protocol is that each is not designed cognitive impairments, such as children with develop- to address all forms of noncompliance, which would mental disabilities who are at increased risk of hypoxia include task performance and head motion. Such an when sedation approaches are used [32]. Another limita- expectation is too demanding. By targeting only one form tion of the present investigation concerns the relative of [potential] noncompliance for improvement, what contributions of the various components of the multiple these approaches and our multiple reward compliance reward compliance protocol (preference assessment, Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 9 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 School of Medicine, Baltimore MD, USA. Department of Behavior Analysis, increase in reward, visual display). Future research is University of North Texas, Denton TX, USA. Department of Psychology, needed that examines which component or components University of Pittsburgh, Pittsburgh PA, USA. Department of Statistics, may contribute the most to improving task compliance. University of Pittsburgh, Pittsburgh PA, USA. During our generalization tests, there was a rather sub- Authors’ contributions stantial sample of subjects that did not receive the visual Design: MS, MC, GS. Data collection: MS, AM, JS, CL, GS. Analysis: MS, GS, IS. display component but nonetheless showed high task Writing: MS, GS, MC, CL, JS, EF, RD and NR All authors read and approved the final manuscript. compliance. This was limited to the two fMRI studies on child and adolescent depression (N = 34 and N = 32) Competing interests and a portion of the first group of children participating The authors declare that they have no competing interests. in the childhood anxiety study (N = 21). Findings show- Received: 9 July 2010 Accepted: 6 May 2011 Published: 6 May 2011 ing high task compliance without the visual display com- ponent suggest identifying and providing rewards References contingent upon task completion may be relatively more 1. Church JA, Petersen SE, Schlaggar BL: The “Task B problem” and other considerations in developmental functional neuroimaging. Hum Brain important than the information supplied by the road Mapp 2010, 31:852-62. map. Another limitation of the present investigation is 2. Kotsoni E, Byrd D, Casey BJ: Special considerations for functional that it remains unclear what subject variables or aspects magnetic resonance imaging of pediatric populations. J Magn Reson Imaging 2006, 23:877-886. of participation contributed to task noncompliance. 3. O’Shaughnessy ES, Berl MM, Moore EN, Gaillard WD: Pediatric Functional Researchers and clinicians might benefit from gathering Magnetic Resonance Imaging (fMRI): Issues and Applications. J Child information during debriefing about the reason(s) for Neurology 2008, 23:791-801. 4. Poldrack RA, Paré-Blagoev EJ, Grant PE: Pediatric functional magnetic early termination. Children’s reports may contain valuable resonance imaging: progress and challenges. Top Magn Reson Imaging information about how to structure the environment or 2002, 13:61-70. tasks in ways that encourage cooperation. 5. Wilke M, Holland SK, Myseros JS, Schmithorst VJ, Ball WS Jr: Functional magnetic resonance imaging in pediatrics. Neuropediatrics 2003, 34:225-33. Conclusions 6. Tyc VL, Fairclough D, Fletcher B, Leigh L, Mulhern RK: Children’s distress Functional magnetic resonance imaging is increasingly during magnetic resonance imaging procedures. Child Health Care 1995, 24:5-19. being used to advance our understanding of normal and 7. Lukins R, Davan IG, Drummond PD: A cognitive behavioural approach to pathological brain function and development in children preventing anxiety during magnetic resonance imaging. J Behav Ther Exp and adolescents. Noncompliance involving an inability to Psychiatry 1997, 28:97-104. 8. Pressdee D, May L, Eastman E, Grier D: The use of play therapy in the remain in a scanner to complete fMRI tasks is one com- preparation of children undergoing MR imaging. Clin Radiol 1997, mon and significant problem. Consequently, researchers 52:945-947. and clinicians devote considerable effort to developing a 9. Byars AW, Holland SK, Strawsburg RH, Bommer W, Dunn RS, Schmithorst VJ, Plante E: Practical aspects of conducting large-scale functional magnetic supportive, positive environment. We proposed that some resonance imaging studies in children. J Child Neurol 2002, 17:885-890. forms of task noncompliance may emerge from less than 10. Rosenberg DR, Sweeney JA, Gillen JS, Kim J, Varanelli MJ, O’Hearn KM, optimal reward protocols. Our findings suggest that Erb PA, Davis D, Thulborn KR: Magnetic resonance imaging of children without sedation: preparation with simulation. J Am Acad Child Adolesc increasing our attention to how rewards are selected and Psychiatry 1997, 36:853-859. delivered may aid cooperation with completing fMRI tasks. 11. de Amorim e Silva CJ, Mackenzie A, Hallowell LM, Stewart SE, Ditchfield MR: The approach presented and preliminary findings contri- Practice MRI: reducing the need for sedation and general anaesthesia in children undergoing MRI. Australas Radiol 2006, 50:319-323. bute to the pediatric neuroimaging literature by providing 12. Epstein JN, Casey BJ, Tonev ST, Davidson M, Reiss AL, Garrett A, a useful way to conceptualize and measure task noncom- Hinshaw SP, Greenhill LL, Vitolo A, Kotler LA, Jarrett MA, Spicer J: pliance and a set of cost effective tactics for improving the Assessment and prevention of head motion during imaging of patients with attention deficit hyperactivity disorder. Psychiatry Res Neuroimaging effectiveness of common reward-based protocols. 2007, 155:75-82. 13. Slifer KJ, Cataldo MF, Cataldo MD, Llorente AM, Gerson AC: Behavior Appl Behav Anal analysis of motion control for pediatric neuroimaging. J List of abbreviations used 1993, 26:469-470. MRCP: Multiple reward compliance protocol 14. Slifer KJ, Bucholtz JD, Cataldo MD: Behavioral training of motion control in young children undergoing radiation treatment without sedation. J Acknowledgements Pediatr Oncol Nurs 1994, 11:55-63. This paper was made possible with the support of NIH grant funding 15. Slifer KJ: A video system to help children cooperate with motion control MH41712 (R.D.), P50 MH080215 (N.R.), K02 MH082998 (G.S.), K01 MH083001 for radiation treatment without sedation. J Pediatr Oncol Nurs 1996, (C.L.), KO1 MHO73077 (J.S.), DA024144-01 (J.S./R.D.), a NARSAD Young 13:91-97. Investigator Award (J.S.), the Beatrice H. Barrett Research Endowment (M.S.) 16. Slifer KJ, Koontz KL, Cataldo MF: Operant-contingency-based preparation and the efforts of Lindsay Proud, Rachel Kolko, Erika Joyce, and Emily of children for functional magnetic resonance imaging. J Appl Behav Anal Yarrison. 2002, 35:191-194. 17. Raschle NM, Lee M, Buechler R, Christodoulou JA, Chang M, Vakil M, Author details Stering PL, Gaab N: Making MR Imaging Child’s Play - Pediatric Department of Psychiatry, University of Pittsburgh School of Medicine, Neuroimaging Protocol, Guidelines and Procedure. JoVE 2009, 29: [http:// Pittsburgh PA, USA. Department of Behavioral Psychology, Kennedy Krieger www.jove.com/index/details.stp?id=1309]. Institute, Baltimore MD, USA. Department of Psychiatry, Johns Hopkins Schlund et al. Behavioral and Brain Functions 2011, 7:10 Page 10 of 10 http://www.behavioralandbrainfunctions.com/content/7/1/10 18. Yerys BE, Jankowski KF, Shook D, Rosenberger LR, Barnes KA, Berl MM, Ritzl EK, Vanmeter J, Vaidya CJ, Gaillard WD: The fMRI success rate of children and adolescents: Typical development, epilepsy, attention deficit/hyperactivity disorder, and autism spectrum disorders. Hum Brain Mapp 2009, 30:3426-3435. 19. Marsh R, Zhu H, Schultz RT, Quackenbush G, Royal J, Skudlarski P, Peterson BS: A developmental fMRI study of self-regulatory control. Hum Brain Mapp 2006, 27:848-863. 20. Bjork JM, Smith AR, Chen G, Hommer DW: Adolescents, adults and rewards: comparing motivational neurocircuitry recruitment using fMRI. PLoS One 2010, 5:e11440. 21. Christophersen ER, Mortweet SL: Parenting that Works: Building Skills that Last a Lifetime. American Psychological Association Books, Washington DC; 22. Christophersen ER: Little People: Guidelines for Common Sense Child Rearing. Westport Publishers, Inc, Kansas City MO, 3 1998. 23. Kazdin AE, Rotella C: The Kazdin Method for Parenting the Defiant Child. Mariner Books, Houghton Mifflen Harccourt Publishers, New York; 2008. 24. Scheres A, Dijkstra M, Ainslie E, Balkan J, Reynolds B, Sonuga-Barke E, Castellanos FX: Temporal and probabilistic discounting of rewards in children and adolescents: effects of age and ADHD symptoms. Neuropsychologia 2006, 44:2092-2103. 25. Chelonis JJ, Osborn SA, Gravelin CR, Paule MG: Assessing motivation in children using a progressive ratio task. Proceedings of the Assoc for Behavior Analysis 36th Annual Conference 2010. 26. Gadow KD, Sprafkin J: Adolescent Symptom Inventory – 4: Norms Manual. Checkmate Plus, New York; 1998. 27. Gadow KD, Sprafkin J: Child Symptom Inventory – 4: Screening Manual. Checkmate Plus, New York; 1998. 28. Schlund MW, Siegle GJ, Ladouceur CD, Silk JS, Cataldo MF, Forbes EE, Dahl RE, Ryan ND: Nothing to fear? Neural systems supporting avoidance behavior in healthy youths. Neuroimage 2010, 52:710-719. 29. Hagopian LP, Long ES, Rush KS: Preference assessment procedures for individuals with developmental disabilities. Behav Modif 2004, 28:668-677. 30. Jenkins L, Myerson J, Hale S, Fry AF: Individual and developmental differences in working memory across the life span. Psychon Bull Rev 1999, 6:28-40. 31. Fry AF, Hale S: Relationships among processing speed, working memory, and fluid intelligence in children. Biol Psychol 2000, 54:1-34. 32. Kannikeswaran N, Mahajan PV, Sethuraman U, Groebe A, Chen X: Sedation medication received and adverse events related to sedation for brain MRI in children with and without developmental disabilities. Paediatr Anaesth 2009, 19:250-256. doi:10.1186/1744-9081-7-10 Cite this article as: Schlund et al.: Pediatric functional magnetic resonance neuroimaging: tactics for encouraging task compliance. Behavioral and Brain Functions 2011 7:10. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit

Journal

Behavioral and Brain Functions – Springer Journals

Published: May 6, 2011

There are no references for this article.