Christopher K. Rhea
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Christopher K. Rhea has a total of 143 co-authors. Here are the top co-authors by number of shared publications:
- Scott E. Ross
Department of Kinesiology, The University of North Carolina at Greensboro, Greensboro, NC 27402, United States
6 shared publications - Ross Gore
Center for Secure and Intelligent Critical Systems, Old Dominion University, Suffolk, VA 23435, USA
5 shared publications W. Geoffrey Wright
Affiliation Department of Health and Rehabilitation Sciences, Temple University, Philadelphia, Pennsylvania, United States of America
5 shared publicationsAdam W. Kiefer
Division of Sports Medicine, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; Department of Pediatrics, College of Medicine, University of Cincinnati, Cincinnati, OH, USA; Center for Cognition, Action & Perception, Department of Psychology, University of Cincinnati, Cincinnati, OH, USA
4 shared publications- Atmaram Yarlagadda
McDonald Army Health Center, Newport News, VA 23604, USA
4 shared publications - Brittany S. Samulski
School of Rehabilitation Sciences, Old Dominion University, Norfolk, VA 23508, USA
4 shared publications
Latest Publications
- Open Access
- 14 October 2026
- Published by MDPI AG in Technologies
- Vol. 14 (10) , 665
Hierarchical Clustering Reveals Multilevel Structure for Multifactorial Assessment in Community-Dwelling Older Adults
Multifactorial falls assessment is recommended for older adults at elevated risk but can generate numerous potentially overlapping measurements, increasing testing burden despite potentially diminishing informational returns as correlated measures are added. Characterizing how measures covary may inform subsequent efforts to evaluate more efficient assessment procedures. Community-dwelling older adults (n = 298) completed a single-visit assessment including the Physiological Profile Assessment (PPA), preferred- and maximal-speed walking, Montreal Cognitive Assessment (MoCA), and Modified Falls Efficacy Scale (MFES). Hierarchical agglomerative clustering was applied to 51 variables. Dendrogram fidelity, cluster separation, and resampling stability were evaluated using the cophenetic correlation coefficient (CCC), silhouette width, Jaccard coefficients, and adjusted Rand index (ARI). The dendrogram showed strong fidelity to the correlation-distance structure (CCC = 0.937). Among the examined partitions, k = 2 had the highest mean silhouette width (0.333), indicating weak separation, and demonstrated high resampling stability. This partition grouped MoCA and MFES items with visual, strength, and maximal balance-range measures, separately from gait, postural sway, reaction-time, sensory, and coordinated-stability measures. At k = 3, the latter grouping remained unchanged while MoCA items separated from the MFES and physiologic measures; however, this finer, hypothesis-guided partition had lower separation and lower stability for the MoCA-dominant cluster. These findings describe the covariance organization of this multifactorial assessment battery and provide candidate groupings for future research on measurement redundancy, information retention, and reproducibility across datasets.
- 14 October 2026
- Published by MDPI AG in Technologies
- Vol. 14 (10) , 665
Open AccessHierarchical Clustering Reveals Multilevel Structure for Multifactorial Assessment in Community-Dwelling Older Adults
Multifactorial falls assessment is recommended for older adults at elevated risk but can generate numerous potentially overlapping measurements, increasing testing burden despite potentially diminishing informational returns as correlated measures are added. Characterizing how measures covary may inform subsequent efforts to evaluate more efficient assessment procedures. Community-dwelling older adults (n = 298) completed a single-visit assessment including the Physiological Profile Assessment (PPA), preferred- and maximal-speed walking, Montreal Cognitive Assessment (MoCA), and Modified Falls Efficacy Scale (MFES). Hierarchical agglomerative clustering was applied to 51 variables. Dendrogram fidelity, cluster separation, and resampling stability were evaluated using the cophenetic correlation coefficient (CCC), silhouette width, Jaccard coefficients, and adjusted Rand index (ARI). The dendrogram showed strong fidelity to the correlation-distance structure (CCC = 0.937). Among the examined partitions, k = 2 had the highest mean silhouette width (0.333), indicating weak separation, and demonstrated high resampling stability. This partition grouped MoCA and MFES items with visual, strength, and maximal balance-range measures, separately from gait, postural sway, reaction-time, sensory, and coordinated-stability measures. At k = 3, the latter grouping remained unchanged while MoCA items separated from the MFES and physiologic measures; however, this finer, hypothesis-guided partition had lower separation and lower stability for the MoCA-dominant cluster. These findings describe the covariance organization of this multifactorial assessment battery and provide candidate groupings for future research on measurement redundancy, information retention, and reproducibility across datasets.
- Open Access
- 8 October 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (19) , 9929
Perceived Benefits of Adapted Fitness Participation Among Adults with Physical Impairments
Background/Objectives: Regular physical activity is recommended to maintain a healthy lifestyle. Individuals with physical impairments have barriers that limit their participation in traditional exercise. Understanding their perceptions of adapted fitness programming can aid in the large-scale development and implementation of such programs. The aim of this study was to identify perceived benefits of adults with physical impairments participating in an adapted fitness program. Methods: Eight adapted fitness program participants (7 males and 1 female) with various neurologic diagnoses or bilateral lower extremity amputation completed the study. A consensual qualitative research (CQR) design with a semi-structured interview guide with evidence of content validity was used. The interview guide explicitly prompted physical, functional, emotional, and social benefits. Interviews were conducted via Zoom. Data were analyzed through three phases of consensus coding by three researchers. Trustworthiness measures included member-checking, multi-analyst triangulation, and an external auditor. Results: Three themes with associated categories were developed through the broader CQR analysis. The present study focuses on the intrinsic benefits theme, which included four categories: (1) general physical benefits, (2) functional movement/activities of daily living benefits, (3) mental/emotional benefits, and (4) social benefits. Participants described perceived improvements in physical capacity and daily function, as well as benefits related to mood, identity, and social connection. These benefits coexisted with continued functional deficits and variation in participants’ experiences of social connection. Conclusions: Participants reported physical, functional, mental/emotional, and social benefits that they associated with adapted fitness participation. These findings describe the perceived value of participation among eight participants in an adapted fitness program. The findings may inform discussions of participants’ goals and experiences but do not establish program effectiveness. Rather, they provide insight into how participants perceived the adapted fitness program and its relevance to their physical function and well-being.
- 8 October 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (19) , 9929
Open AccessPerceived Benefits of Adapted Fitness Participation Among Adults with Physical Impairments
Background/Objectives: Regular physical activity is recommended to maintain a healthy lifestyle. Individuals with physical impairments have barriers that limit their participation in traditional exercise. Understanding their perceptions of adapted fitness programming can aid in the large-scale development and implementation of such programs. The aim of this study was to identify perceived benefits of adults with physical impairments participating in an adapted fitness program. Methods: Eight adapted fitness program participants (7 males and 1 female) with various neurologic diagnoses or bilateral lower extremity amputation completed the study. A consensual qualitative research (CQR) design with a semi-structured interview guide with evidence of content validity was used. The interview guide explicitly prompted physical, functional, emotional, and social benefits. Interviews were conducted via Zoom. Data were analyzed through three phases of consensus coding by three researchers. Trustworthiness measures included member-checking, multi-analyst triangulation, and an external auditor. Results: Three themes with associated categories were developed through the broader CQR analysis. The present study focuses on the intrinsic benefits theme, which included four categories: (1) general physical benefits, (2) functional movement/activities of daily living benefits, (3) mental/emotional benefits, and (4) social benefits. Participants described perceived improvements in physical capacity and daily function, as well as benefits related to mood, identity, and social connection. These benefits coexisted with continued functional deficits and variation in participants’ experiences of social connection. Conclusions: Participants reported physical, functional, mental/emotional, and social benefits that they associated with adapted fitness participation. These findings describe the perceived value of participation among eight participants in an adapted fitness program. The findings may inform discussions of participants’ goals and experiences but do not establish program effectiveness. Rather, they provide insight into how participants perceived the adapted fitness program and its relevance to their physical function and well-being.
- Open Access
- 13 March 2026
- Published by Brazilian Journal for Motor Behavior in Brazilian Journal of Motor Behavior
- Vol. 20 (1) , e527
Nonlinear differences in postural sway across attentional focus conditions in individuals with and without anterior cruciate ligament reconstruction
BACKGROUND: The structure of variability in postural sway reflects the adaptability and health of the postural control system. Entropy quantifies this structure by measuring the regularity of repeated fluctuations over time. Although entropy has been used to examine sway in neurological populations, little is known about how sway complexity changes following anterior cruciate ligament reconstruction (ACLR). Attentional focus may also influence postural control, with external focus typically promoting more efficient and automatic regulation than internal focus. AIM: To examine how attentional focus (internal vs. external) influences the structure of postural sway in individuals with and without ACLR. METHODS: Forty-three participants (20 ACLR, 23 controls) completed quiet standing trials on a force plate under three attentional focus conditions: internal focus on the injured or non-injured limb, and external focus. Sample entropy was calculated to assess sway complexity, and traditional linear sway variables (mean velocity, path length) were also examined. RESULTS: Participants with ACLR exhibited lower sample entropy than controls, indicating more regular and less adaptable sway patterns. In the ACLR group, external focus instructions significantly increased sample entropy relative to internal focus conditions. Linear sway measures also reflected improved control under external focus, with lower mean sway velocity and reduced path length. INTERPRETATION: Individuals with ACLR demonstrate reduced complexity in postural control, but external focus instructions can enhance adaptability. These findings suggest that attentional focus modulates sway dynamics after ACLR. Incorporating external focus into rehabilitation may facilitate more automatic and flexible postural control strategies.
- 13 March 2026
- Published by Brazilian Journal for Motor Behavior in Brazilian Journal of Motor Behavior
- Vol. 20 (1) , e527
Open AccessNonlinear differences in postural sway across attentional focus conditions in individuals with and without anterior cruciate ligament reconstruction
BACKGROUND: The structure of variability in postural sway reflects the adaptability and health of the postural control system. Entropy quantifies this structure by measuring the regularity of repeated fluctuations over time. Although entropy has been used to examine sway in neurological populations, little is known about how sway complexity changes following anterior cruciate ligament reconstruction (ACLR). Attentional focus may also influence postural control, with external focus typically promoting more efficient and automatic regulation than internal focus. AIM: To examine how attentional focus (internal vs. external) influences the structure of postural sway in individuals with and without ACLR. METHODS: Forty-three participants (20 ACLR, 23 controls) completed quiet standing trials on a force plate under three attentional focus conditions: internal focus on the injured or non-injured limb, and external focus. Sample entropy was calculated to assess sway complexity, and traditional linear sway variables (mean velocity, path length) were also examined. RESULTS: Participants with ACLR exhibited lower sample entropy than controls, indicating more regular and less adaptable sway patterns. In the ACLR group, external focus instructions significantly increased sample entropy relative to internal focus conditions. Linear sway measures also reflected improved control under external focus, with lower mean sway velocity and reduced path length. INTERPRETATION: Individuals with ACLR demonstrate reduced complexity in postural control, but external focus instructions can enhance adaptability. These findings suggest that attentional focus modulates sway dynamics after ACLR. Incorporating external focus into rehabilitation may facilitate more automatic and flexible postural control strategies.
- Open Access
EEG and IMU Gait Signal Processing: A Comparative Assessment of the “Reza” Exponential Filter and Classical Filters
Noise degrades both EEG and gait signals, and classical IIR filters (Butterworth, Chebyshev, elliptic) involve trade-offs between passband flatness, ripple, and roll-off. This study compared a novel exponential “Reza” filter with these designs for neural and locomotor data. We analyzed an open-source mobile brain–body imaging dataset with EEG and gait data from 49 healthy adults (EEG: -channel, ; IMUs: six APDM Opals, ). EEG channels were grand-averaged and band-pass filtered at , while IMU axes were averaged and band-pass filtered at . The outcomes were signal-to-noise ratio SNR (dB) and band-integrated Welch PSD (). Repeated-measures ANOVAs tested the effect of filter types (Butterworth, Chebyshev I, elliptic, Reza) with Bonferroni-adjusted post hoc tests for the six pairwise filter comparisons (). We reported partial eta-squared () as the ANOVA effect size. For EEG, PSD did not differ among filters , whereas SNR differed strongly : Chebyshev and elliptic yielded the highest mean SNR and did not differ from each other, while both exceeded Butterworth, Reza was the lowest. For IMU, both SNR ) and PSD differed: Reza produced the highest mean SNR (significantly exceeding elliptic and Chebyshev), while Butterworth exceeded Chebyshev; meanwhile, IMU PSD showed a clear ordering with Reza retaining the most motion-band power, followed by Butterworth, then Chebyshev, with elliptic retaining the least. These results showed that filter choice materially shapes EEG and gait outcomes. For EEG, Chebyshev maximized SNR, while elliptic and Reza maintained comparable fidelity. For IMU gait signals, Reza matched Butterworth for denoising and preserved more signal power. Therefore, filter choice should be guided by the target outcome (SNR vs. band power) rather than a single default design.Open AccessEEG and IMU Gait Signal Processing: A Comparative Assessment of the “Reza” Exponential Filter and Classical Filters
Noise degrades both EEG and gait signals, and classical IIR filters (Butterworth, Chebyshev, elliptic) involve trade-offs between passband flatness, ripple, and roll-off. This study compared a novel exponential “Reza” filter with these designs for neural and locomotor data. We analyzed an open-source mobile brain–body imaging dataset with EEG and gait data from 49 healthy adults (EEG: -channel, ; IMUs: six APDM Opals, ). EEG channels were grand-averaged and band-pass filtered at , while IMU axes were averaged and band-pass filtered at . The outcomes were signal-to-noise ratio SNR (dB) and band-integrated Welch PSD (). Repeated-measures ANOVAs tested the effect of filter types (Butterworth, Chebyshev I, elliptic, Reza) with Bonferroni-adjusted post hoc tests for the six pairwise filter comparisons (). We reported partial eta-squared () as the ANOVA effect size. For EEG, PSD did not differ among filters , whereas SNR differed strongly : Chebyshev and elliptic yielded the highest mean SNR and did not differ from each other, while both exceeded Butterworth, Reza was the lowest. For IMU, both SNR ) and PSD differed: Reza produced the highest mean SNR (significantly exceeding elliptic and Chebyshev), while Butterworth exceeded Chebyshev; meanwhile, IMU PSD showed a clear ordering with Reza retaining the most motion-band power, followed by Butterworth, then Chebyshev, with elliptic retaining the least. These results showed that filter choice materially shapes EEG and gait outcomes. For EEG, Chebyshev maximized SNR, while elliptic and Reza maintained comparable fidelity. For IMU gait signals, Reza matched Butterworth for denoising and preserved more signal power. Therefore, filter choice should be guided by the target outcome (SNR vs. band power) rather than a single default design. - Open Access
- 27 February 2026
- Published by MDPI AG in Biomechanics
- Vol. 6 (1) , 23
Standardization of Neuromuscular Reflex Analysis—Role of Fine-Tuned Vision-Language Model Consortium and OpenAI gpt-oss Reasoning LLM-Enabled Decision Support System
Background/Objectives: Accurate assessment of neuromuscular reflexes, such as the Hoffmann reflex (H-reflex), plays a critical role in sports science, rehabilitation, and clinical neurology. Conventional interpretation of H-reflex electromyography (EMG) waveforms is subject to inter-rater variability and interpretive bias, limiting reliability and standardization. This study aims to develop an automated, interpretable, and robust agentic AI–driven framework for H-reflex waveform analysis. Methods: We propose a fine-tuned Vision–Language Model (VLM) consortium combined with a reasoning Large Language Model (LLM)–enabled decision support system for automated H-reflex interpretation. Multiple VLMs were fine-tuned on curated datasets of H-reflex EMG waveform images annotated with expert clinical observations, recovery timelines, and athlete metadata. The VLM outputs were aggregated using a consensus-based strategy and further refined by a specialized reasoning LLM to ensure coherent, transparent, and explainable diagnostic assessments. Model fine-tuning employed Low-Rank Adaptation (LoRA) and 4-bit quantization to enable efficient deployment on consumer-grade hardware. Results: Experimental evaluation demonstrated that the proposed hybrid system delivers accurate, consistent, and clinically interpretable assessments of neuromuscular states, including fatigue, injury, and recovery, directly from EMG waveform images and contextual metadata. Compared with baseline models, the fine-tuned VLM consortium exhibited substantially improved precision, consistency, and contextual awareness, while the reasoning LLM enhanced diagnostic coherence through cross-model consensus and structured reasoning, thereby supporting responsible and explainable AI-driven decision making. Conclusions: This work presents, to the authors’ knowledge, the first integration of a responsible and explainable AI-driven decision support system for H-reflex analysis. The proposed framework advances the automation and standardization of neuromuscular diagnostics and establishes a foundation for next-generation AI-assisted decision support systems in sports performance monitoring, rehabilitation, and clinical neurophysiology.
- 27 February 2026
- Published by MDPI AG in Biomechanics
- Vol. 6 (1) , 23
Open AccessStandardization of Neuromuscular Reflex Analysis—Role of Fine-Tuned Vision-Language Model Consortium and OpenAI gpt-oss Reasoning LLM-Enabled Decision Support System
Background/Objectives: Accurate assessment of neuromuscular reflexes, such as the Hoffmann reflex (H-reflex), plays a critical role in sports science, rehabilitation, and clinical neurology. Conventional interpretation of H-reflex electromyography (EMG) waveforms is subject to inter-rater variability and interpretive bias, limiting reliability and standardization. This study aims to develop an automated, interpretable, and robust agentic AI–driven framework for H-reflex waveform analysis. Methods: We propose a fine-tuned Vision–Language Model (VLM) consortium combined with a reasoning Large Language Model (LLM)–enabled decision support system for automated H-reflex interpretation. Multiple VLMs were fine-tuned on curated datasets of H-reflex EMG waveform images annotated with expert clinical observations, recovery timelines, and athlete metadata. The VLM outputs were aggregated using a consensus-based strategy and further refined by a specialized reasoning LLM to ensure coherent, transparent, and explainable diagnostic assessments. Model fine-tuning employed Low-Rank Adaptation (LoRA) and 4-bit quantization to enable efficient deployment on consumer-grade hardware. Results: Experimental evaluation demonstrated that the proposed hybrid system delivers accurate, consistent, and clinically interpretable assessments of neuromuscular states, including fatigue, injury, and recovery, directly from EMG waveform images and contextual metadata. Compared with baseline models, the fine-tuned VLM consortium exhibited substantially improved precision, consistency, and contextual awareness, while the reasoning LLM enhanced diagnostic coherence through cross-model consensus and structured reasoning, thereby supporting responsible and explainable AI-driven decision making. Conclusions: This work presents, to the authors’ knowledge, the first integration of a responsible and explainable AI-driven decision support system for H-reflex analysis. The proposed framework advances the automation and standardization of neuromuscular diagnostics and establishes a foundation for next-generation AI-assisted decision support systems in sports performance monitoring, rehabilitation, and clinical neurophysiology.
- Open Access
- 2 December 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (23) , 12740
How Many Trials Are Needed for Consistent Clinical Gait Assessment?
Background: Clinical assessment of gait typically consists of patients walking a few trials at various speeds while the clinician assesses performance. Unfortunately, there is no clear guidance on how gait changes across trials, leaving clinicians uncertain about the optimal number of trials needed to observe consistent (non-variable) performance. To address this issue, we examined gait performance from a large dataset of older adults who participated in a community-based comprehensive fall risk assessment. Methods: Community-dwelling, older adults (n = 340; 70.8 ± 7.4 years; 120 men, 220 women) performed gait trials under two conditions: preferred and maximum walking speed. Individuals were encouraged to complete five trials for both conditions. Consistency between gait trials within each condition was calculated using intraclass correlation (ICC) and standard error of measurement (SEM) analysis. Results: Our data showed the middle three trials had the most consistency compared to the average of 2–5 trials. Conclusions: When performing a clinical gait analysis, the first trial should be used to acclimate the participant to the protocol and not used for analysis. Data should be recorded from the next three trials, which is when gait appears to stabilize. Data from a fifth trial differs from the second trial, potentially indicating fatigue and/or motivation changes, so it is recommended that the gait analysis conclude after the fourth trial.
- 2 December 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (23) , 12740
Open AccessHow Many Trials Are Needed for Consistent Clinical Gait Assessment?
Background: Clinical assessment of gait typically consists of patients walking a few trials at various speeds while the clinician assesses performance. Unfortunately, there is no clear guidance on how gait changes across trials, leaving clinicians uncertain about the optimal number of trials needed to observe consistent (non-variable) performance. To address this issue, we examined gait performance from a large dataset of older adults who participated in a community-based comprehensive fall risk assessment. Methods: Community-dwelling, older adults (n = 340; 70.8 ± 7.4 years; 120 men, 220 women) performed gait trials under two conditions: preferred and maximum walking speed. Individuals were encouraged to complete five trials for both conditions. Consistency between gait trials within each condition was calculated using intraclass correlation (ICC) and standard error of measurement (SEM) analysis. Results: Our data showed the middle three trials had the most consistency compared to the average of 2–5 trials. Conclusions: When performing a clinical gait analysis, the first trial should be used to acclimate the participant to the protocol and not used for analysis. Data should be recorded from the next three trials, which is when gait appears to stabilize. Data from a fifth trial differs from the second trial, potentially indicating fatigue and/or motivation changes, so it is recommended that the gait analysis conclude after the fourth trial.
- Open Access
Standardization of Psychiatric Diagnoses -- Role of Fine-tuned LLM Consortium and OpenAI-gpt-oss Reasoning LLM Enabled Decision Support System
The diagnosis of most mental disorders, including psychiatric evaluations, primarily depends on dialogues between psychiatrists and patients. This subjective process can lead to variability in diagnoses across clinicians and patients, resulting in inconsistencies and challenges in achieving reliable outcomes. To address these issues and standardize psychiatric diagnoses, we propose a Fine-Tuned Large Language Model (LLM) Consortium and OpenAI-gpt-oss Reasoning LLM-enabled Decision Support System for the clinical diagnosis of mental disorders. Our approach leverages fine-tuned LLMs trained on conversational datasets involving psychiatrist-patient interactions focused on mental health conditions (e.g., depression). The diagnostic predictions from individual models are aggregated through a consensus-based decision-making process, refined by the OpenAI-gpt-oss reasoning LLM. We propose a novel method for deploying LLM agents that orchestrate communication between the LLM consortium and the reasoning LLM, ensuring transparency, reliability, and responsible AI across the entire diagnostic workflow. Experimental results demonstrate the transformative potential of combining fine-tuned LLMs with a reasoning model to create a robust and highly accurate diagnostic system for mental health assessment. A prototype of the proposed platform, integrating three fine-tuned LLMs with the OpenAI-gpt-oss reasoning LLM, was developed in collaboration with the U.S. Army Medical Research Team in Norfolk, Virginia, USA. To the best of our knowledge, this work represents the first application of a fine-tuned LLM consortium integrated with a reasoning LLM for clinical mental health diagnosis paving the way for next-generation AI-powered eHealth systems aimed at standardizing psychiatric diagnoses.
Open AccessStandardization of Psychiatric Diagnoses -- Role of Fine-tuned LLM Consortium and OpenAI-gpt-oss Reasoning LLM Enabled Decision Support System
The diagnosis of most mental disorders, including psychiatric evaluations, primarily depends on dialogues between psychiatrists and patients. This subjective process can lead to variability in diagnoses across clinicians and patients, resulting in inconsistencies and challenges in achieving reliable outcomes. To address these issues and standardize psychiatric diagnoses, we propose a Fine-Tuned Large Language Model (LLM) Consortium and OpenAI-gpt-oss Reasoning LLM-enabled Decision Support System for the clinical diagnosis of mental disorders. Our approach leverages fine-tuned LLMs trained on conversational datasets involving psychiatrist-patient interactions focused on mental health conditions (e.g., depression). The diagnostic predictions from individual models are aggregated through a consensus-based decision-making process, refined by the OpenAI-gpt-oss reasoning LLM. We propose a novel method for deploying LLM agents that orchestrate communication between the LLM consortium and the reasoning LLM, ensuring transparency, reliability, and responsible AI across the entire diagnostic workflow. Experimental results demonstrate the transformative potential of combining fine-tuned LLMs with a reasoning model to create a robust and highly accurate diagnostic system for mental health assessment. A prototype of the proposed platform, integrating three fine-tuned LLMs with the OpenAI-gpt-oss reasoning LLM, was developed in collaboration with the U.S. Army Medical Research Team in Norfolk, Virginia, USA. To the best of our knowledge, this work represents the first application of a fine-tuned LLM consortium integrated with a reasoning LLM for clinical mental health diagnosis paving the way for next-generation AI-powered eHealth systems aimed at standardizing psychiatric diagnoses.
- Open Access
Standardization of Neuromuscular Reflex Analysis -- Role of Fine-Tuned Vision-Language Model Consortium and OpenAI gpt-oss Reasoning LLM Enabled Decision Support System
Accurate assessment of neuromuscular reflexes, such as the H-reflex, plays a critical role in sports science, rehabilitation, and clinical neurology. Traditional analysis of H-reflex EMG waveforms is subject to variability and interpretation bias among clinicians and researchers, limiting reliability and standardization. To address these challenges, we propose a Fine-Tuned Vision-Language Model (VLM) Consortium and a reasoning Large-Language Model (LLM)-enabled Decision Support System for automated H-reflex waveform interpretation and diagnosis. Our approach leverages multiple VLMs, each fine-tuned on curated datasets of H-reflex EMG waveform images annotated with clinical observations, recovery timelines, and athlete metadata. These models are capable of extracting key electrophysiological features and predicting neuromuscular states, including fatigue, injury, and recovery, directly from EMG images and contextual metadata. Diagnostic outputs from the VLM consortium are aggregated using a consensus-based method and refined by a specialized reasoning LLM, which ensures robust, transparent, and explainable decision support for clinicians and sports scientists. The end-to-end platform orchestrates seamless communication between the VLM ensemble and the reasoning LLM, integrating prompt engineering strategies and automated reasoning workflows using LLM Agents. Experimental results demonstrate that this hybrid system delivers highly accurate, consistent, and interpretable H-reflex assessments, significantly advancing the automation and standardization of neuromuscular diagnostics. To our knowledge, this work represents the first integration of a fine-tuned VLM consortium with a reasoning LLM for image-based H-reflex analysis, laying the foundation for next-generation AI-assisted neuromuscular assessment and athlete monitoring platforms.
Open AccessStandardization of Neuromuscular Reflex Analysis -- Role of Fine-Tuned Vision-Language Model Consortium and OpenAI gpt-oss Reasoning LLM Enabled Decision Support System
Accurate assessment of neuromuscular reflexes, such as the H-reflex, plays a critical role in sports science, rehabilitation, and clinical neurology. Traditional analysis of H-reflex EMG waveforms is subject to variability and interpretation bias among clinicians and researchers, limiting reliability and standardization. To address these challenges, we propose a Fine-Tuned Vision-Language Model (VLM) Consortium and a reasoning Large-Language Model (LLM)-enabled Decision Support System for automated H-reflex waveform interpretation and diagnosis. Our approach leverages multiple VLMs, each fine-tuned on curated datasets of H-reflex EMG waveform images annotated with clinical observations, recovery timelines, and athlete metadata. These models are capable of extracting key electrophysiological features and predicting neuromuscular states, including fatigue, injury, and recovery, directly from EMG images and contextual metadata. Diagnostic outputs from the VLM consortium are aggregated using a consensus-based method and refined by a specialized reasoning LLM, which ensures robust, transparent, and explainable decision support for clinicians and sports scientists. The end-to-end platform orchestrates seamless communication between the VLM ensemble and the reasoning LLM, integrating prompt engineering strategies and automated reasoning workflows using LLM Agents. Experimental results demonstrate that this hybrid system delivers highly accurate, consistent, and interpretable H-reflex assessments, significantly advancing the automation and standardization of neuromuscular diagnostics. To our knowledge, this work represents the first integration of a fine-tuned VLM consortium with a reasoning LLM for image-based H-reflex analysis, laying the foundation for next-generation AI-assisted neuromuscular assessment and athlete monitoring platforms.
- Open Access
Proof-of-TBI -- Fine-Tuned Vision Language Model Consortium and OpenAI-o3 Reasoning LLM-Based Medical Diagnosis Support System for Mild Traumatic Brain Injury (TBI) Prediction
Mild Traumatic Brain Injury (TBI) detection presents significant challenges due to the subtle and often ambiguous presentation of symptoms in medical imaging, making accurate diagnosis a complex task. To address these challenges, we propose Proof-of-TBI, a medical diagnosis support system that integrates multiple fine-tuned vision-language models with the OpenAI-o3 reasoning large language model (LLM). Our approach fine-tunes multiple vision-language models using a labeled dataset of TBI MRI scans, training them to diagnose TBI symptoms effectively. The predictions from these models are aggregated through a consensus-based decision-making process. The system evaluates the predictions from all fine-tuned vision language models using the OpenAI-o3 reasoning LLM, a model that has demonstrated remarkable reasoning performance, to produce the most accurate final diagnosis. The LLM Agents orchestrates interactions between the vision-language models and the reasoning LLM, managing the final decision-making process with transparency, reliability, and automation. This end-to-end decision-making workflow combines the vision-language model consortium with the OpenAI-o3 reasoning LLM, enabled by custom prompt engineering by the LLM agents. The prototype for the proposed platform was developed in collaboration with the U.S. Army Medical Research team in Newport News, Virginia, incorporating five fine-tuned vision-language models. The results demonstrate the transformative potential of combining fine-tuned vision-language model inputs with the OpenAI-o3 reasoning LLM to create a robust, secure, and highly accurate diagnostic system for mild TBI prediction. To the best of our knowledge, this research represents the first application of fine-tuned vision-language models integrated with a reasoning LLM for TBI prediction tasks.
Open AccessProof-of-TBI -- Fine-Tuned Vision Language Model Consortium and OpenAI-o3 Reasoning LLM-Based Medical Diagnosis Support System for Mild Traumatic Brain Injury (TBI) Prediction
Mild Traumatic Brain Injury (TBI) detection presents significant challenges due to the subtle and often ambiguous presentation of symptoms in medical imaging, making accurate diagnosis a complex task. To address these challenges, we propose Proof-of-TBI, a medical diagnosis support system that integrates multiple fine-tuned vision-language models with the OpenAI-o3 reasoning large language model (LLM). Our approach fine-tunes multiple vision-language models using a labeled dataset of TBI MRI scans, training them to diagnose TBI symptoms effectively. The predictions from these models are aggregated through a consensus-based decision-making process. The system evaluates the predictions from all fine-tuned vision language models using the OpenAI-o3 reasoning LLM, a model that has demonstrated remarkable reasoning performance, to produce the most accurate final diagnosis. The LLM Agents orchestrates interactions between the vision-language models and the reasoning LLM, managing the final decision-making process with transparency, reliability, and automation. This end-to-end decision-making workflow combines the vision-language model consortium with the OpenAI-o3 reasoning LLM, enabled by custom prompt engineering by the LLM agents. The prototype for the proposed platform was developed in collaboration with the U.S. Army Medical Research team in Newport News, Virginia, incorporating five fine-tuned vision-language models. The results demonstrate the transformative potential of combining fine-tuned vision-language model inputs with the OpenAI-o3 reasoning LLM to create a robust, secure, and highly accurate diagnostic system for mild TBI prediction. To the best of our knowledge, this research represents the first application of fine-tuned vision-language models integrated with a reasoning LLM for TBI prediction tasks.
- Open Access
- 22 April 2025
- Published by MDPI AG in Technologies
- Vol. 13 (5) , 167
The Landscape of Virtual Reality Use in Mobility Rehabilitation from 2010–2023: A Scoping Review
Significant advancements in virtual reality (VR) technology have occurred in the past decade, allowing clinical researchers to take advantage of these reduced barriers to explore the use of VR in patient populations. This scoping review on VR interventions to improve mobility in adults and children focuses on the literature from 2010–2023. A total of 2736 articles were screened and 126 articles met the inclusion criteria. Most of the studies were conducted in inpatient clinical settings (n = 41) and investigated VR interventions to improve balance (n = 118). Less immersive (n = 108) products such as Nintendo Wii or Xbox Kinect were primarily used. Additionally, 37.0% of studies (n = 47) used off-the-shelf programs like Wii Fit Plus and 73.2% of studies (n = 93) found statistically significant improvements in motor outcomes following VR intervention. The articles included in this review suggest that the majority of VR research for physical rehabilitation is being performed in clinical settings. Most studies reported statistically significant improvements in their outcome variables following VR intervention. These observations demonstrate that research in this area is moving beyond proof-of-concept and toward translation to clinical applications.
- 22 April 2025
- Published by MDPI AG in Technologies
- Vol. 13 (5) , 167
Open AccessThe Landscape of Virtual Reality Use in Mobility Rehabilitation from 2010–2023: A Scoping Review
Significant advancements in virtual reality (VR) technology have occurred in the past decade, allowing clinical researchers to take advantage of these reduced barriers to explore the use of VR in patient populations. This scoping review on VR interventions to improve mobility in adults and children focuses on the literature from 2010–2023. A total of 2736 articles were screened and 126 articles met the inclusion criteria. Most of the studies were conducted in inpatient clinical settings (n = 41) and investigated VR interventions to improve balance (n = 118). Less immersive (n = 108) products such as Nintendo Wii or Xbox Kinect were primarily used. Additionally, 37.0% of studies (n = 47) used off-the-shelf programs like Wii Fit Plus and 73.2% of studies (n = 93) found statistically significant improvements in motor outcomes following VR intervention. The articles included in this review suggest that the majority of VR research for physical rehabilitation is being performed in clinical settings. Most studies reported statistically significant improvements in their outcome variables following VR intervention. These observations demonstrate that research in this area is moving beyond proof-of-concept and toward translation to clinical applications.