SBIR-STTR Award

This Small Business Innovation Research (SBIR) Phase-I project proposes the development of an image- processing-based tool, named PostureCheck, aimed at automatically detecting when patients perform undesirable compensatory movements during robot-assisted upper-limb rehabilitation exercises. The system will be based on a standard video camera (e.g., GoPro) that will be used to capture the movements of the subject. The automatic detection of undesirable compensatory movements is especially important when patients use a rehabilitation robotic system with minimum supervision, i.e. when a single therapist oversees the therapeutic sessions of multiple patients simultaneously. In this context, PostureCheck? may be capable of tracking robot-assisted rehabilitation exercises and enable feedback modalities to discourage the performance of undesirable compensatory movements. Our long-term goal is to integrate PostureCheck? with the Barrett Upper-extremity Robotic Trainer - BURT®, which we developed with special emphasis on stroke rehabilitation. The combination of PostureCheck? with the BURT® device would be ideally suited for deployment in ?Robotic Gyms?, where a single therapist oversees the therapeutic sessions of several patients simultaneously, thus allowing rehabilitation centers to offer high-dosage, high-intensity interventions despite the limited number of therapists currently available in the US. To demonstrate the feasibility of the proposed concept, we will develop PostureCheck? to detect the most common compensatory movements automatically. To achieve this goal, we will rely on recently developed artificial intelligence (AI) methods referred to as Deep Learning. These methods have recently broken records in the human-posture analysis, joint-skeleton detection, and recognition of human activities using a single inexpensive camera. The proposed video-based PostureCheck? tool will be the first system to exploit the capabilities of hybrid Deep Neural Networks, for real-time detection of compensatory movements during robot- assisted rehabilitation. The proposed SBIR Phase-I activities are organized in three aims. In Aim 1, feedback from rehabilitation experts at Spaulding Rehabilitation Hospital will be used to collect video data and to label compensatory movements observed during the performance of robot-assisted rehabilitation exercises by using the BURT® system. In Aim 2, Deep Learning techniques will be used to develop a robust detection of undesirable compensatory movements during the performance of robot-assisted rehabilitation exercises. Finally, in Aim 3, the algorithms developed in Aim 2 will be optimized. Specifically, we will test implementations that are suitable to generate real-time feedback. Computationally efficient implementations of the algorithms will enable - in future studies - the development of new modalities of control of the rehabilitation robot with the objective of discouraging undesirable compensatory movements.

Project Terms:
Algorithms; American; Architecture; Artificial Intelligence; base; Clinical; clinical efficacy; Clinical Research; Data; deep learning; deep neural network; design; Detection; Development; Devices; dosage; efficacy evaluation; Environment; Exercise; exercise rehabilitation; experience; Feedback; Future; Generations; Goals; Hospitals; Human; Human Activities; Hybrids; image processing; Impaired cognition; improved; improved functioning; Individual; interest; Intervention; Intuition; Joints; Label; Laboratories; Lateral; meetings; Methods; Modality; monocular; Motion; Motor; motor impairment; motor recovery; Movement; Names; network architecture; novel; Patients; Pattern; Performance; Phase; phase 1 study; phase 2 study; Physical therapy; Physiological; Posture; Records; recruit; Rehabilitation Centers; Rehabilitation therapy; Research Personnel; restoration; Robot; robot assistance; robot rehabilitation; robotic device; Robotics; Series; Severities; Shoulder; Side; Skeleton; Small Business Innovation Research Grant; Specialist; Stroke; stroke rehabilitation; stroke survivor; stroke therapy; Supervision; System; Techniques; Technology; Testing; Therapeutic; Time; tool; Training; Upper Extremity; usability; Validation; Video Recording; Videotape; Vision;

This Small Business Innovation Research (SBIR) Phase-II project proposes the deployment of an AI-poweredimage-processing tool, named PostureCheck, to automatically detect when patients perform undesirablecompensatory movements during robot-assisted upper-extremity (UE) rehabilitation exercises. The system isbased on a standard video camera and the artificial intelligence (AI) software developed under the NIH Phase-Iproject. PostureCheck™ will be fully integrated with Barrett Technology's existing Burt® UE rehabilitation robot,ensuring that patients are provided appropriate and timely feedback to encourage the correct performance oftherapeutic exercises without requiring constant therapist supervision. Burt® is an FDA-listed UE rehabilitation robot that supports the weight of a patient's arm and hand while thepatient moves his/her arm to interact with on-screen games. Burt® offers guided assistance, visual and hapticfeedback, and activities and assessments to both help train the patient and measure progress. Clinical studieshave shown that Burt® can be used to improve patients' performance in one-on-one sessions with a therapist.However, the system lacks the capability for a single therapist to monitor and work with several patients andBurt® systems at one time. The proposed SBIR Phase-II activities are organized in three aims. In Aim 1, the annotated dataset generatedin Phase-I will be used to develop three separate AI modules. These modules will be integrated into a frameworkthat allows therapists to monitor the outcome of the modules, and, through feedback, enable the system to auto-adapt to improve performance continuously. In Aim 2, stakeholder feedback will be gathered and integrated todesign multiple user interfaces for the PostureCheck™ tool. Specific interfaces will be created for use bytherapists during and after RT sessions. PostureCheck™ will be integrated with the Burt® device to empowertherapists to provide effective feedback to patients and deter the use of compensatory movements. Finally, inAim 3, forty-two stroke survivors will be recruited in an interventional study deploying the PostureCheck™ andBurt® systems. Subjects will undergo eighteen RT sessions in either an individual or group therapy format. Motorperformance between groups will be compared to gather information about the suitability of the combinedsystems for multi-patient RT therapy in future rehabilitation centers. The long-term commercial goal of the project is to provide a practical Burt®-plus-PostureCheck™ system toempower therapists to supervise multiple patients simultaneously through a gamut of useful functionalities. Thesystem will be suitable for deployment in clinics as well as rehabilitation centers such as wellness gyms.

Public Health Relevance Statement:
Project Narrative Each year, more than 650,000 Americans survive a type of stroke that particularly weakens one side of the body in most cases. During upper-limb rehabilitation, appropriate feedback from a therapist to discourage stroke survivors from performing undesirable compensatory movements results in better motor recovery and ultimately improved function. This proposal aims to expand PostureCheck™, to provide timely and targeted feedback and enable therapists to perform robot-assisted and multi-patient therapy in a minimally supervised session.

Project Terms:
<21+ years old>