Advances in Rehabilitation Technology to Transform Health

Abstract

The use of technology in rehabilitation is not new. As PTJ’s latest featured collection illuminates, recent innovative technology-supported rehabilitation solutions have tremendous potential to augment and elevate clinical practice further. Rehabilitative technology has traditionally been defined as aids that help people recover their functioning after injury or illness. Newer perspectives broaden this definition to include technology that enables more refined measurement of movement (eg, wearable sensors), augments treatment (eg, robotic devices, virtual reality), monitors and promotes exercise participation, or provides real-time monitoring and feedback (eg, digital health platforms, mobile health applications).1–3 The goal of rehabilitation is to help patients improve movement and physical function for everyday living. Advances in technology can help to bridge the gap between the clinic and real-world settings, with the potential to improve patient outcomes. Many of the articles in this featured collection highlight the power of movement-sensing technologies to augment clinic-based assessments by capturing movement autonomously, continuously, at a more granular level, and in a naturalistic environment. For example, advances in wearable sensing,4–6 inertial measurement unit (IMU)–vision systems,7 and video-based analysis of movement using pose estimation8 can help rehabilitation professionals gain greater insight into the nature of movement dysfunction, identify phenotypes,9 enhance classification, and determine prognosis—resulting in the implementation of a more targeted, personalized treatment approach. Advances in artificial intelligence (AI), such as machine learning applications, can enhance technology-supported rehabilitation.8,10 With greater automation potential and faster processing of large amounts of complex data, relevant data-driven metrics can be harnessed to enhance clinical decision making.11 More advances are needed to ensure that these technologies are both clinician-friendly and patient-friendly enough to foster adoption in the clinic; however, some technologies already offer platforms that can be more easily integrated into clinical practice, including a smartphone app to monitor falls,12 wrist-worn accelerometers to capture upper extremity movement,4 wearable devices,13 video-based gait analysis,8 and a home-based commercial system to monitor exercise adherence and intensity.14 The articles in this featured collection highlight the power, the opportunities, and the limitations associated with several innovative rehabilitation technologies. Four original research papers present evidence on the clinical utility of wearable sensors to both assess movement and identify individuals at high risk for poor outcomes: Gombatto et al9 demonstrate that a portable motion-tracking sensor (dorsaVi) assessing frequency and duration of low back postures and movement can be used to identify 2 distinct clinical phenotypes in Hispanic/Latino people with chronic back pain. This study provides novel evidence on the use of wearable technology for ecological monitoring to inform more personalized plans of care. Demers and colleagues4 show the feasibility and usability of a wrist-worn wearable sensor (MiGo activity watch) to assess arm and hand performance in daily activities in people with chronic stroke. Results support the use of wearable technology as a potential tool for transferring skills learned in the clinic to the home environment. O’Brien et al5 provide evidence that using sensor data from IMUs can improve prediction of ambulation and risk of falling in patients undergoing inpatient stroke rehabilitation. Findings highlight the potential of wearable technology to improve discharge planning in the inpatient setting. Weston and colleagues6 established the reliability of turning tasks in healthy adults using IMU data on head, trunk, and lumbar spine movement. Instrumented measures of turning may help identify individuals, especially those with Parkinson disease (PD), at high risk for falls and has the potential to inform task-specific training in the clinical setting. Three original research articles highlight how AI technology and video-based systems can be an innovative approach to functional movement and gait assessment in a clinical setting with potential for use in the home environment: Spangler et al7 found that a multimodal sensor system that includes IMUs combined with a computer visions system (2 RGB-D cameras) was a feasible and accurate approach for assessing functional movement in community-dwelling adults discharged from the hospital. This innovative technology has potential for broader future applications, such as monitoring activity in community and home settings. Work by Lin et al10 in infants born at term and preterm shows how AI technology can be an innovative approach for movement tracking and classification in a laboratory setting. Results will inform the use of human pose estimation AI algorithms for early motor assessment in the clinic and home environment to help identify developmental disorders. John and colleagues8 provide evidence that video-based gait analysis using pose estimation AI algorithms can accurately measure gait parameters and hip and knee kinematics in people after stroke. This type of approach may be a useful gait analysis tool to help identify gait deviations in the clinic and home settings. A feasibility trial illustrates how a wearable device and telehealth technology can be used as an interventional tool to promote physical activity. Master and colleagues13 show that a wrist-worn wearable device (Fitbit) and telehealth physical activity counseling by a physical therapist is a feasible and acceptable approach to promote physical activity early after lumbar spine surgery. The authors’ discussion of feasibility issues relating to remote data monitoring and delivery of the intervention is applicable for both clinical researchers and health care providers focused on improving physical activity. A multisite randomized controlled trial and a mixed-methods article highlight the use of a web-based platform and mHealth as measurement and monitoring tools: Rosenfeldt and colleagues14 share the benefits of using a home-based exercise platform (Peloton Interactive) to monitor exercise behavior and promote exercise adherence in people with PD. Objective exercise data from outside the clinical setting can help identify adherence subtypes, which subsequently informs shared decision making in rehabilitation. Wales et al12 share the development process for a smartphone app (iFall) that can help with accurate falls reporting in people with PD. The authors provide a roadmap for coproduction of a digital measurement tool that has the potential to support self-management, clinical assessment, and reliable measurement for future research. An embedded pragmatic clinical trial reports on the implementation of a hybrid model with telehealth visits. Lentz and colleagues15 describe a hybrid care pathway in the Veterans Health Administration (VHA) Care System that includes telehealth and in-person care for low back pain. Authors discuss challenges and potential opportunities for how integrated hybrid care models that include physical therapy can function in a real-world health care setting. A scoping review discusses the use of rehabilitation technology in clinical practice. Murphy et al16 focus on the uptake of technology-based interventions in neurorehabilitation clinical practice for people with stroke and other neurological conditions. This review brings attention to important barriers and facilitators for adoption that can help translate rehabilitation technology to the clinical setting. These 12 papers highlight the diversity of current rehabilitation technology research and the opportunities that exist for development and testing of technology-supported assessment tools and management strategies. However, many rehabilitation technologies have not yet been incorporated into routine clinical care; continued work is needed to increase the uptake of technology into clinical practice. The articles in this featured collection provide the basis for future research focused on training and validating sensor-prediction models and AI algorithms, determining whether personalized interventions using wearable technology enhance patient outcomes, examining implementation of innovative care models that provide psychologically informed physical therapy care, testing remote monitoring devices to inform risk stratification and shared decision making, and developing feasible and cost-effective digital tools that improve self-monitoring and self-management approaches. As guest editors of this PTJ featured collection, we hope that the lessons learned from these articles can be considered in relation to the Research Agenda for Physical Therapy from the American Physical Therapy Association (APTA).17 Several of the research priorities directly address rehabilitation technology (“develop and/or determine the effects, usability, and affordability of novel assistive technologies on physical function and other patient outcomes,” “examine the effectiveness of digital health for physical therapist evaluation and intervention,” “explore the role and value of artificial intelligence in improving and customizing patient care”). The featured collection serves to highlight the value that rehabilitation technology can add to population health research (“develop strategies to promote adherence, sustainability, and successful outcomes for population-level mobility and physical activity interventions”), mechanistic research (“explore treatment modifiers and/or biomarkers to guide personalized rehabilitation approaches for optimizing patient outcomes”), and health services research (“evaluate physical therapy delivered via different care models on clinical outcomes and cost-effectiveness”). Only through the integration and adoption of wearable sensing technologies, digital health platforms, and telehealth delivery models—in real-world clinical and home settings—can advances in rehabilitation technology meaningfully transform health. The guest coeditors gratefully acknowledge the manuscript reviewers who contributed their time, expertise, and constructive comments to this featured collection: Natalie Allen, PT, PhD Michael Bade, PT, DPT, PhD Carla Batista, PT, PhD Mark Bowden, PT, PhD Roy Coronado, PT, PhD Keith Cole, PT, DPT, PhD Leo Costa, PT, PhD Maria Earl, PT, PhD Margaret French, PT, DPT, PhD Sujay Galen, PT, PhD Simone Gill, OT, PhD Sara Gombatto, PT, PhD Jill Heathcock, PT, PhD Anne Kloos, PT, PhD Belinda Lawford, PT, PhD Alan Chong Lee, PT, DPT, PhD Michael Lewek, PT, PhD Linda Li, PT, PhD Daniel Miner, PT, DPT, PhD Mark Paterno, PT, MBA Daniel Peterson, PhD Lori Quinn, PT, EdD Ryan Roemmich, PhD Michael Rowe, PT, PhD Veronica Rowe, OT, PhD Sean Rundell, PT, DPT, PhD Michael Schneider, DC, PhD Lisa Simpson, PT, PhD Jo Armour Smith, PT, PhD Jennifer Stevens-Lapsley, PT, PhD Sonu Subudhi, PhD Lynn Snyder-Mackler, PT, PhD Elizabeth Thompson, PT, PhD Melissa Tovin, PT, PhD Antonia Zaferiou, PhD

Topics & Concepts