The goal of this basic experimental clinical trial is to understand the effect of multitasking practice on the structure of neural representations of tasks in the human lateral prefrontal cortex and control brain regions. The main question it aims to answer is: What changes in neural representational structure predict improvements in multitasking behavior due to multitasking practice? Healthy human participants will learn two independent tasks, each mapping a set of stimuli to motor responses based on different rules. Participants will be randomized to one of two interventions. Participants assigned to the multitask practice intervention (MPI) will practice multitasking the two tasks over multiple days. Those assigned to the single-task practice intervention (SPI) will instead practice each task separately while controlling for the total number of practice opportunities associated with each task across the interventions. Both before and after the practice, the ability of all participants to perform both tasks simultaneously will be behaviorally measured using a well-established psychological refractory period (PRP) paradigm, and their neural representations will be measured using functional MRI while they perform the two tasks. Researchers will then compare improvements in multitasking behavior across the two groups, as well as changes in neural representational geometry of the tasks in the lateral prefrontal cortex and control brain regions, and test whether multitasking training is associated with specific changes in neural representations in the lateral prefrontal cortex.
Background and Study Rationale: Human performance on a task deteriorates when it is performed concurrently with another task. This multitasking cost has been attributed to competition for shared computational or biological resources and there is considerable debate in the field about the specific nature of these putative resources. Extensive multitasking practice is known to reduce, and sometimes, abolish these costs. These practice-related gains have been attributed to changes in central processing, more efficient task scheduling, or the learning of specialized task representations. The human prefrontal cortex is critical to flexible cognitive control of behavior, and is involved in representing tasks. However, it remains unknown what effect multitasking practice has on prefrontal task representations. In this study, we will directly address a gap in our knowledge by testing the link between multitasking costs and the geometry of neural manifolds representing task inputs in the lateral PFC. Study Design: Overview: This study is a single-center, randomized controlled trial designed to examine the effect of multitask practice on multitasking behavior, the geometry of neural representations of tasks in the lateral prefrontal cortex and the relationship between these effects. A total of 60 healthy subjects are planned, 30 subjects being randomized to each, the multitask practice group (MPG), and the single task practice group (SPG). Baseline and endline assessments of multitasking behavior and neural representations in lateral PFC with fMRI will be conducted in all subjects. Study Treatments: Single-task practice treatment: The single-task practice treatment will involve 3 sessions of practice during which subjects will practice instructed tasks in individual task blocks. Up to 48 blocks of practice will be performed. Subjects will be given feedback at the end of each block on their accuracy and speed of responding and they will be given an improvement target. Multitask practice treatment: The multitask practice treatment will involve 3 sessions of practice during which subjects will practice instructed tasks in the psychological refractory period paradigm. Up to 48 blocks of practice will be performed. Subjects will be given feedback at the end of each block on their accuracy and speed of responding and they will be given an improvement target. Subject Selection: Subjects will be recruited from the Providence community. Subjects will be screened over a phone call. Inclusion criteria include right-handedness, age between 18 and 35, and normal or corrected-to-normal vision. Exclusion criteria include history of psychiatric or neurological deficits and MR contraindications. Consenting procedures will be conducted in person during Study Session 1 and continuing consent will be periodically obtained during all lab Study visits. Random Assignment: Up to 60 eligible patients will be randomly assigned to SPG or MPG treatment groups in a 1:1 ratio by using a random number generator coded in MATLAB. Neither the experimenters nor the study subjects, will be blinded to the assignment. Study Structure: A total of 14 study sessions are planned for each subject. Study session 1 will be employed for informed consent, scheduling and familiarization with study tasks. Baseline evaluations will take place during study sessions 2 through 5. Treatment will be administered during study sessions 6 through 8 during which subjects will receive either 3 sessions of multi-task practice (MPG) or 3 sessions of single-task practice (SPG). Endline evaluations will take place during study sessions 9 through 12. Detailed descriptions of each Study session and all paradigms used for measurements are given below. Behavioral paradigms: All behavioral task paradigms will be implemented using Psychtoolbox 3 in MATLAB or in JSPsych. All tasks involve responding to visual objects presented on a computer screen with keypresses either on a keyboard or on a response box. Visual objects will vary along 4 dimensions - shape (square or circle), circle (orange or blue), pattern (dotted or striped) or size (small or large). Along with the visual object, a response panel consisting of the symbols @ and #. Subjects will be introduced to three different tasks: the shape task, color task and the pattern task. In each task, they will make a binary classification of the visual object relying on the relevant task feature (e.g. color in the color task). Each binary class will be associated with one of two categories associated with the @ and # symbols. Subjects will then select the response key associated with the relevant symbol. The left-right position of the two symbols will be varied randomly from trial to trial such that the position would be the same as the previous trial on approximately 50% of the trials, and each trial-type was associated with an equal number of left and right responses. Single task paradigm: In the single-task paradigm, visual objects will be presented for up to 2.2 s. Responses deadline is 2.2s. Inter-trial interval will be 1 s. Trials are presented in pure single-task blocks with 64 trials per block. Mixed tasks paradigm: In the mixed tasks paradigm, visual objects will be presented for up to 2.2 s. Response deadline is 2.2s. Inter-trial interval will be jittered between 2 and 10 seconds. Trials of multiple task types will be presented in the same block, with a word cue identifying the task to be performed on every trial. Psychological Refractory Period paradigm: In the PRP paradigm, two visual objects will be presented on every trial, one in the upper half of the screen, and the other in the bottom half of the screen. Visual objects will be presented with a stimulus onset synchrony (SOA) varying between 50 ms and 1.5 s. Trial length will be up to 3 s. Response deadline will be 3 s for responding to both visual objects. Inter-trial interval will be either 1 s (behavioral blocks) or jittered between 2 and 10 seconds (MRI blocks). Subjects will be instructed to perform the tasks in a specific order (e.g. Color, then Shape), responding to the first stimulus that appears with the rule associated with the first task in the order and to the object that appears second. A total of 3 different task orders will be presented to subjects (e.g. Color -\> Shape, Shape -\> Color and Color -\> Pattern) and these orders will be counterbalanced across subjects such that each task has placed every role equally in each treatment group. All task orders will be administered during both baseline and endline. MRI Protocol: Each subject will under MRI procedures during Study Visits 2, 3, 9 and 10. Each visit will last 2 hrs during which subjects will first be provided refresher instruction on the behavioral tasks before undergoing MRI for 1 hr. During each 1 hr MRI scanning session, subjects will first be screened for metal, briefed on safety procedures and motion minimization requirements, and then situated on the scanning table in a supine position. Their head will be stabilized with soft padding, and their ears will be protected from MR noise with ear plugs. Subjects will view a computer screen through a mirror installed on a headstage and will be provided with two response boxes for performing the tasks and an emergency buzzer for communicating with the experimenters and radiographers on duty. They will then undergo a multiple echo MPRAGE scan (repetition time (TR) = 2530 ms, echo times (TE) = 1.69, 3.55, 5.41, and 7.27 ms, flip angle = 7 degrees, 176 sagittal slices, 1 × 1 × 1 mm voxels) for 6 minutes during which they may close their eyes. Following this, subjects will undergo 5-10 runs of Echo Planar Imaging (TR = 1000 ms, TE = 32.6 ms, flip angle = 64, SMS = 5, Echo spacing 0.59, 65 interleaved slices with voxel size of 2.4mm x 2.4mm x 2.4mm), each lasting between 4 to 9 minutes, during which they will perform multiple blocks of either the mixed tasks paradigm (Days 2 and 9) or the PRP paradigm (Days 3 and 10). Study Session Descriptions: Study Session 1: During Study session 1, subjects will go through the consenting process with a trained research assistant. After informed consent is provided, subjects will be given an overview of the study and they will be introduced to the basic task procedure and the color, shape and pattern rules. They will practice each task in the single task paradigm for 2 blocks each. Following this they will be introduced to the PRP paradigm with their assigned task orders, for 6 blocks. At the end of the session, subjects will be debriefed and remunerated for their participation. Study Session 2: During Study session 2, subjects will undergo baseline measurements with the PRP paradigm with their assigned task orders, for 12 blocks. At the end of the session, subjects will be debriefed and remunerated for their participation. Study Session 3: During Study session 3, subjects will undergo baseline measurements with the Mixed tasks paradigm for 12 blocks. At the end of the session, subjects will be debriefed and remunerated for their participation. Study Session 4: During Study session 4, subjects will undergo baseline measurements of the neural representations of tasks. The MRI protocol will be administered with the Mixed tasks paradigm. Study Session 5: During Study session 5, subjects will undergo baseline measurements of the neural representations of tasks. The MRI protocol will be administered with the PRP paradigm. Study Sessions 6-8: Subjects will be administered the treatment during these sessions. The single-task practice treatment group will receive 12-24 blocks of practice with the single task paradigm and they will be provided feedback on their performance after every block along with targets for improvement. The multi-task practice treatment will practice one of the three orders in the PRP paradigm across 12-24 blocks Subjects will be given feedback at the end of each block on their accuracy and speed of responding and they will be given an improvement target. Study Session 9: During Study session 9, subjects will undergo endline measurements with the PRP paradigm with their assigned task orders, for 12 blocks. At the end of the session, subjects will be debriefed and remunerated for their participation. Study Session 10: During Study session 10, subjects will undergo endline measurements with the Mixed tasks paradigm for 12 blocks. At the end of the session, subjects will be debriefed and remunerated for their participation. Study Session 11: During Study session 11, subjects will undergo endline measurements of the neural representations of tasks. The MRI protocol will be administered with the Mixed tasks paradigm. Study Session 12: During Study session 12, subjects will undergo endline measurements of the neural representations of tasks. The MRI protocol will be administered with the PRP paradigm.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
BASIC_SCIENCE
Masking
NONE
Enrollment
60
Multitask practice intervention (MPI) includes multiple behavioral testing sessions during which participants receive practice with multitasking two tasks using the psychological refractory period procedure.
Single task practice intervention (SPI) includes multiple behavioral testing sessions during which participants receive separate practice on two tasks.
Metcalf Research Building, Brown University & MRI Research Facility, Brown University
Providence, Rhode Island, United States
RECRUITINGEffect of the practice intervention on change in overall and task-relevant neural separability in lateral prefrontal cortex
Overall neural separability (also referred to as shattering dimensionality) will be measured by employing cross-validated neural decoding of lateral PFC fMRI BOLD multi-voxel patterns, averaging decoding accuracies across all balanced dichotomies of the task inputs. Task-relevant neural separability will be measured by averaging decoding accuracies across all task-relevant variables.
Time frame: through study completion, an average of 1.5 years
Effect of multitasking practice on behavioral multitasking cost
Behavioral multitasking cost will be estimated by computing the difference in mean response time on task 2 on long SOAs and short SOAs in the Psychological Refractory Period paradigm. The effect of multitasking practice on behavioral multitasking will be estimated using linear mixed-effects regression, and is defined as the regression weight associated with the interaction effect of time point and intervention on multitasking cost.
Time frame: through study completion, an average of 1.5 years
Contribution of overall and task-relevant neural separability in lateral prefrontal cortex to the effect of multitasking practice on behavioral multitasking cost
The contribution of prefrontal neural separability change to the effect of multitasking practice on behavioral multitasking will be estimated using linear mixed-effects regression, and is defined as the regression weight associated with the interaction effect of time point, intervention and separability on multitasking cost.
Time frame: through study completion, an average of 1.5 years
Cross-classification generalization performance of task-relevant variables in lateral prefrontal cortex
Cross-classification generalization performance will be estimated by averaging decoding accuracies from cross-classification of individual task variables across subsets of task inputs in the lateral prefrontal cortex.
Time frame: through study completion, an average of 1.5 years
Coding axis angles between stimulus feature representations in two tasks in lateral prefrontal cortex
Coding axis angles will be estimated by computing the angle between the weight vectors of linear classifiers trained to decode stimulus features in each task context.
Time frame: through study completion, an average of 1.5 years
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.