On performance and behavioral data collection from online neurorehabilitation tool
Presenting author:
BrainIn (https://brainin.kiv.zcu.cz/) is an online software tool designed to facilitate the neurorehabilitation of people with acquired brain damage. It has been developing as a modular and highly parameterized system that 1) enables online communication between therapists and their patients/families of these patients, 2) offers a wide range of cognitive, attention, memory and motor exercises in varying degrees of difficulty, 3) allows therapists to create, modify and personalize the neurorehabilitation tasks given to their patients, 4) collects data and metadata from the performed tasks to semi-automatically adjust the type and complexity of future tasks.
The BrainIn architecture is based on the set of 1) templates that define the input and output variables for each type of task, 2) tasks that set the input variables defined in templates and form specific exercises for patients, 3) packages that form task units given by therapists to their patients. The user roles include patients, therapists (they create/modify tasks and packages based on the existing templates and define therapy), super-therapists (in addition, they create/modify templates) and administrators.
The data collected includes all input variables' values, detailed results achieved by patients and their behaviour (such as individual clicks, mistakes, and corrections) while performing tasks. These data can be utilized by therapists and by machine learning methods to modify the therapy.
The BrainIn architecture is based on the set of 1) templates that define the input and output variables for each type of task, 2) tasks that set the input variables defined in templates and form specific exercises for patients, 3) packages that form task units given by therapists to their patients. The user roles include patients, therapists (they create/modify tasks and packages based on the existing templates and define therapy), super-therapists (in addition, they create/modify templates) and administrators.
The data collected includes all input variables' values, detailed results achieved by patients and their behaviour (such as individual clicks, mistakes, and corrections) while performing tasks. These data can be utilized by therapists and by machine learning methods to modify the therapy.