Jarrad Fisher
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jarrad Fisher has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Craig Stuart Anderson
Institute of Science and Technology for Brain-inspired Intelligence, Fudan University, Shanghai, China (C.S.A.).
3 shared publications - Xiaoying Chen
The George Institute for Global Health, University of New South Wales, Sydney, NSW, Australia
3 shared publications - Alexandra Hurden
Brain Health Program, The George Institute for Global Health, Sydney, NSW Australia
2 shared publications - Christopher Bunn
Advanced Multimedia Research Lab, Centre of Artificial Intelligence, University of Wollongong, Wollongong, NSW, Australia
2 shared publications - Ceyaan Rauniyar
Faculty of Medicine and Health, University of New South Wales, Sydney, NSW, Australia
1 shared publication - Christine T. Shiner
School of Clinical Medicine, Faculty of Medicine and Health, St Vincent’s Healthcare Clinical Campus, University of NSW Sydney, Sydney, Australia
1 shared publication
Latest Publications
- Open Access
- 2 November 2026
- Published by SAGE Publications in Neurorehabilitation and Neural Repair
Development of a Single-Camera Markerless Motion Capture System for Assessing Upper Limb Function Post-Stroke (FULMA)
Background: Upper limb impairment is common after stroke. Standardised measures such as the Action Research Arm Test and Fugl Meyer Assessment are widely used but may show inter-rater variability and limited sensitivity to subtle motor changes. Objective: To develop and evaluate the Functional Upper Limb Movement Assessment (FULMA) system, a single-camera markerless motion capture tool for objective movement analysis of upper limb function with clinician facing outputs after stroke. Methods: A prospective observational study included 39 stroke participants and 19 healthy controls. Participants completed 9 functional tasks recorded with a LiDAR depth camera. Kinematic features were analysed using machine learning models to predict an ordinal task performance score anchored to Action Research Arm Test scoring descriptors. Predictions were compared with independent occupational therapist ratings, and automated kinematic commentary was generated. Results: The multilayer perceptron achieved a mean accuracy of 83.3%, with task level accuracies of 75.0% to 91.7% and weighted kappa values of 0.67 to 0.88. The support vector machine achieved a mean accuracy of 78.5%, with task level accuracies of 72.2% to 89.0% and weighted kappa values of 0.59 to 0.83. Performance varied across tasks, with the strongest results for cup to mouth and hair brushing. Commentary highlighted prolonged movement time, reduced smoothness and increased trunk displacement. Conclusion: This developmental study showed preliminary agreement between automated ordinal score predictions and therapist ratings within a controlled dataset. External validation and evaluation of measurement properties are required before clinical use. Kinematic commentary should be considered exploratory rather than a validated clinical interpretation.
- 2 November 2026
- Published by SAGE Publications in Neurorehabilitation and Neural Repair
Open AccessDevelopment of a Single-Camera Markerless Motion Capture System for Assessing Upper Limb Function Post-Stroke (FULMA)
Background: Upper limb impairment is common after stroke. Standardised measures such as the Action Research Arm Test and Fugl Meyer Assessment are widely used but may show inter-rater variability and limited sensitivity to subtle motor changes. Objective: To develop and evaluate the Functional Upper Limb Movement Assessment (FULMA) system, a single-camera markerless motion capture tool for objective movement analysis of upper limb function with clinician facing outputs after stroke. Methods: A prospective observational study included 39 stroke participants and 19 healthy controls. Participants completed 9 functional tasks recorded with a LiDAR depth camera. Kinematic features were analysed using machine learning models to predict an ordinal task performance score anchored to Action Research Arm Test scoring descriptors. Predictions were compared with independent occupational therapist ratings, and automated kinematic commentary was generated. Results: The multilayer perceptron achieved a mean accuracy of 83.3%, with task level accuracies of 75.0% to 91.7% and weighted kappa values of 0.67 to 0.88. The support vector machine achieved a mean accuracy of 78.5%, with task level accuracies of 72.2% to 89.0% and weighted kappa values of 0.59 to 0.83. Performance varied across tasks, with the strongest results for cup to mouth and hair brushing. Commentary highlighted prolonged movement time, reduced smoothness and increased trunk displacement. Conclusion: This developmental study showed preliminary agreement between automated ordinal score predictions and therapist ratings within a controlled dataset. External validation and evaluation of measurement properties are required before clinical use. Kinematic commentary should be considered exploratory rather than a validated clinical interpretation.
- Open Access
- 31 August 2026
- Published by Elsevier in SSRN Electronic Journal
An Explainable and Automated Framework for Upper-Limb Motion Assessment
Stroke-induced upper-limb impairment limits daily activities and creates a need for accurate, scalable rehabilitation assessment. Existing automated methods using wearable sensors or 3D skeleton data typically map movement signals directly to clinical scores. Although these approaches can predict impairment, they provide limited insight into how patient movement differs from expected functional motion, reducing clinical interpretability. This paper presents an explainable framework for automated upper-limb motion assessment based on reference-driven deviation analysis. For each functional task, healthy skeleton sequences are used to construct task-specific normative statistics that capture expected motion patterns and natural inter-subject variability. Patient movements are represented in a joint feature space combining absolute kinematic metrics with standardized deviation features computed relative to the healthy reference. This representation supports both ARAT-aligned score prediction and instance-level interpretation. Feature attribution identifies the kinematic components that most influence each prediction, while deviation analysis characterizes the direction, magnitude, and temporal extent of the associated movement abnormalities. The framework is evaluated on a clinically collected dataset of 39 stroke patients and 19 healthy controls performing nine functional upper-limb tasks. Results show substantial agreement with therapist-provided ARAT-aligned scores, with an average Cohen’s kappa of 0.75 across tasks. Case studies further demonstrate that the generated feedback corresponds to observable movement deficits across impairment levels. These findings indicate that healthy-reference deviation modeling provides an effective and clinically interpretable basis for skeleton-based assessment of post-stroke upper-limb function and supports scalable assessment in clinical and home rehabilitation settings. The dataset and code will be made publicly available upon acceptance.
- 31 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessAn Explainable and Automated Framework for Upper-Limb Motion Assessment
Stroke-induced upper-limb impairment limits daily activities and creates a need for accurate, scalable rehabilitation assessment. Existing automated methods using wearable sensors or 3D skeleton data typically map movement signals directly to clinical scores. Although these approaches can predict impairment, they provide limited insight into how patient movement differs from expected functional motion, reducing clinical interpretability. This paper presents an explainable framework for automated upper-limb motion assessment based on reference-driven deviation analysis. For each functional task, healthy skeleton sequences are used to construct task-specific normative statistics that capture expected motion patterns and natural inter-subject variability. Patient movements are represented in a joint feature space combining absolute kinematic metrics with standardized deviation features computed relative to the healthy reference. This representation supports both ARAT-aligned score prediction and instance-level interpretation. Feature attribution identifies the kinematic components that most influence each prediction, while deviation analysis characterizes the direction, magnitude, and temporal extent of the associated movement abnormalities. The framework is evaluated on a clinically collected dataset of 39 stroke patients and 19 healthy controls performing nine functional upper-limb tasks. Results show substantial agreement with therapist-provided ARAT-aligned scores, with an average Cohen’s kappa of 0.75 across tasks. Case studies further demonstrate that the generated feedback corresponds to observable movement deficits across impairment levels. These findings indicate that healthy-reference deviation modeling provides an effective and clinically interpretable basis for skeleton-based assessment of post-stroke upper-limb function and supports scalable assessment in clinical and home rehabilitation settings. The dataset and code will be made publicly available upon acceptance.
- 25 February 2026
- Published by SAGE Publications in Clinical Rehabilitation
- Vol. 40 (9) , 1195-1215
Emergent motion capture technology for upper extremity assessment in stroke: A systematic scoping review
Objective: To systematically appraise motion capture technologies used for clinical assessment of upper limb function in adults post stroke, focusing on applicability, strengths, limitations, and research gaps. Data Sources: A systematic scoping review was conducted in accordance with PRISMA-ScR guidelines. PubMed, MEDLINE, CINAHL, CENTRAL, and IEEE Xplore were searched for studies published between January 2014 and December 2025. Review Methods: Eligible studies quantitatively evaluated motion capture systems in adult stroke populations. Data extracted included technology type, technical specifications, outcomes, and reported strengths and limitations. Results: From 3217 screened studies, 204 were included. Inertial measurement units were most common (71) followed by markerless optical motion capture (43), electromyography (39), marker-based optical motion capture (33), fusion systems (10), and smart devices (8). Marker-based systems provided high precision but were costly and complex. Markerless systems reduced setup requirements and supported clinical use but were prone to tracking errors. Inertial measurement units enabled portable monitoring, though accuracy decreased for slow or fine movements. Electromyography offered insights into neuromuscular activity but was expensive and sensitive to placement. Fusion systems allowed multidimensional assessment but were resource intensive. Smart devices provided accessible options but lacked precision. Most studies focused on chronic stroke, with limited evaluation in acute or home contexts. Reporting of stroke severity was inconsistent, and protocols were rarely standardised. Conclusion: Motion capture technologies show promise for upper limb assessment after stroke but face challenges of cost, accessibility, and standardisation. Research in acute and home settings, along with predictive modelling, is needed to support clinical translation.
- 25 February 2026
- Published by SAGE Publications in Clinical Rehabilitation
- Vol. 40 (9) , 1195-1215
Emergent motion capture technology for upper extremity assessment in stroke: A systematic scoping review
Objective: To systematically appraise motion capture technologies used for clinical assessment of upper limb function in adults post stroke, focusing on applicability, strengths, limitations, and research gaps. Data Sources: A systematic scoping review was conducted in accordance with PRISMA-ScR guidelines. PubMed, MEDLINE, CINAHL, CENTRAL, and IEEE Xplore were searched for studies published between January 2014 and December 2025. Review Methods: Eligible studies quantitatively evaluated motion capture systems in adult stroke populations. Data extracted included technology type, technical specifications, outcomes, and reported strengths and limitations. Results: From 3217 screened studies, 204 were included. Inertial measurement units were most common (71) followed by markerless optical motion capture (43), electromyography (39), marker-based optical motion capture (33), fusion systems (10), and smart devices (8). Marker-based systems provided high precision but were costly and complex. Markerless systems reduced setup requirements and supported clinical use but were prone to tracking errors. Inertial measurement units enabled portable monitoring, though accuracy decreased for slow or fine movements. Electromyography offered insights into neuromuscular activity but was expensive and sensitive to placement. Fusion systems allowed multidimensional assessment but were resource intensive. Smart devices provided accessible options but lacked precision. Most studies focused on chronic stroke, with limited evaluation in acute or home contexts. Reporting of stroke severity was inconsistent, and protocols were rarely standardised. Conclusion: Motion capture technologies show promise for upper limb assessment after stroke but face challenges of cost, accessibility, and standardisation. Research in acute and home settings, along with predictive modelling, is needed to support clinical translation.