Jinfeng Cao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jinfeng Cao has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Jie Lyu
Department of Orthopedics, Jinshan District Central Hospital Affiliated to Shanghai University of Medicine and Health Sciences
2 shared publications Hongtao Cheng
Medical Instrumentation School, Shanghai University of Medicine and Health Sciences, Shanghai, 201318, China
1 shared publication- Jia Han
Department of Orthopedics, Jinshan District Central Hospital Affiliated to Shanghai University of Medicine and Health Sciences
1 shared publication Linlin Zhang
College of Rehabilitation Sciences, Shanghai University of Medicine and Health Sciences
1 shared publicationLinyan Liu
College of Medical Instruments, Shanghai University of Medicine and Health Sciences
1 shared publicationPeiming Zhang
Medical Instrumentation School, Shanghai University of Medicine and Health Sciences, Shanghai, 201318, China
1 shared publication
Latest Publications
- Open Access
- 24 September 2026
- Published by Frontiers Media SA in Frontiers in Bioengineering and Biotechnology
- Vol. 14, 1936471
Automated assessment of balance function based on a lightweight end-to-end multimodal deep learning framework: predicting Mini-BESTest scores from center of pressure signals in a single stance condition
Background: Balance dysfunction is a common health issue among older adults and patients with neurological disorders. The Mini-Balance Evaluation Systems Test (Mini-BESTest) is a widely used clinical tool for balance assessment. However, its full implementation requires a trained clinician to administer 14 tasks, and the scoring is subject to inter-rater variability. Objective: This study aims to develop an end-to-end, lightweight multimodal deep learning framework (with approximately 55 K parameters) to automatically predict the total Mini-BESTest score and its four subscale scores, utilizing only center of pressure (COP) signals from a single stance condition alongside clinical metadata. Methods: COP signals and clinical metadata from 163 subjects in a public dataset (Santos et al., PeerJ, 2016, 4) were utilized. We proposed a lightweight multimodal deep learning model comprising three dilated convolution branches with varying dilation rates, a channel attention mechanism, and a metadata fusion module. Three independent comparative experiments were conducted to investigate the effects of stance condition, trial number, and time window length on predictive performance. Based on these results, a suitable data configuration was selected for subsequent experiments. An adaptive task weighting scheme based on homoscedastic uncertainty was employed to balance the five regression targets with varying score ranges. Five-fold cross-validation was used to evaluate the models, and out-of-fold (OOF) predictions were constructed to provide a full-sample assessment of predictive performance. Results: Under the selected data configuration (foam surface with eyes open, second trial, first 30 s of data), the five-fold cross-validation mean absolute error (MAE) for the predicted Mini-BESTest total score was 1.971 (out of 28). The full-sample OOF prediction MAE was 1.979, with a Pearson correlation coefficient of r = 0.758 and R 2 = 0.574. In ablation experiments, removing the metadata branch led to the largest numerical increase in total-score MAE (to 2.338), while removal of the channel attention module resulted in a smaller increase (to 2.034). A metadata-only model achieved a total-score MAE of 2.105. Paired Wilcoxon signed-rank tests did not show statistically significant differences between the full model and the ablated variants (all p > 0.05). Conclusion: This study demonstrates that an end-to-end deep learning framework can predict the complete Mini-BESTest score from 30 s of quiet standing COP signals under foam surface eyes-open conditions, together with basic clinical metadata. While maintaining predictive accuracy, this lightweight framework (approx. 55 K parameters) may represent a potential direction for future deployment in portable or remote settings, subject to further engineering and clinical validation.
- 24 September 2026
- Published by Frontiers Media SA in Frontiers in Bioengineering and Biotechnology
- Vol. 14, 1936471
Open AccessAutomated assessment of balance function based on a lightweight end-to-end multimodal deep learning framework: predicting Mini-BESTest scores from center of pressure signals in a single stance condition
Background: Balance dysfunction is a common health issue among older adults and patients with neurological disorders. The Mini-Balance Evaluation Systems Test (Mini-BESTest) is a widely used clinical tool for balance assessment. However, its full implementation requires a trained clinician to administer 14 tasks, and the scoring is subject to inter-rater variability. Objective: This study aims to develop an end-to-end, lightweight multimodal deep learning framework (with approximately 55 K parameters) to automatically predict the total Mini-BESTest score and its four subscale scores, utilizing only center of pressure (COP) signals from a single stance condition alongside clinical metadata. Methods: COP signals and clinical metadata from 163 subjects in a public dataset (Santos et al., PeerJ, 2016, 4) were utilized. We proposed a lightweight multimodal deep learning model comprising three dilated convolution branches with varying dilation rates, a channel attention mechanism, and a metadata fusion module. Three independent comparative experiments were conducted to investigate the effects of stance condition, trial number, and time window length on predictive performance. Based on these results, a suitable data configuration was selected for subsequent experiments. An adaptive task weighting scheme based on homoscedastic uncertainty was employed to balance the five regression targets with varying score ranges. Five-fold cross-validation was used to evaluate the models, and out-of-fold (OOF) predictions were constructed to provide a full-sample assessment of predictive performance. Results: Under the selected data configuration (foam surface with eyes open, second trial, first 30 s of data), the five-fold cross-validation mean absolute error (MAE) for the predicted Mini-BESTest total score was 1.971 (out of 28). The full-sample OOF prediction MAE was 1.979, with a Pearson correlation coefficient of r = 0.758 and R 2 = 0.574. In ablation experiments, removing the metadata branch led to the largest numerical increase in total-score MAE (to 2.338), while removal of the channel attention module resulted in a smaller increase (to 2.034). A metadata-only model achieved a total-score MAE of 2.105. Paired Wilcoxon signed-rank tests did not show statistically significant differences between the full model and the ablated variants (all p > 0.05). Conclusion: This study demonstrates that an end-to-end deep learning framework can predict the complete Mini-BESTest score from 30 s of quiet standing COP signals under foam surface eyes-open conditions, together with basic clinical metadata. While maintaining predictive accuracy, this lightweight framework (approx. 55 K parameters) may represent a potential direction for future deployment in portable or remote settings, subject to further engineering and clinical validation.
- Open Access
- 4 October 2016
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 11 (10) , e0163173
Coupled Hybrid Continuum-Discrete Model of Tumor Angiogenesis and Growth
The processes governing tumor growth and angiogenesis are codependent. To study the relationship between them, we proposed a coupled hybrid continuum-discrete model. In this model, tumor cells, their microenvironment (extracellular matrixes, matrix-degrading enzymes, and tumor angiogenic factors), and their network of blood vessels, described by a series of discrete points, were considered. The results of numerical simulation reveal the process of tumor growth and the change in microenvironment from avascular to vascular stage, indicating that the network of blood vessels develops gradually as the tumor grows. Our findings also reveal that a tumor is divided into three regions: necrotic, semi-necrotic, and well-vascularized. The results agree well with the previous relevant studies and physiological facts, and this model represents a platform for further investigations of tumor therapy.
- 4 October 2016
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 11 (10) , e0163173
Open AccessCoupled Hybrid Continuum-Discrete Model of Tumor Angiogenesis and Growth
The processes governing tumor growth and angiogenesis are codependent. To study the relationship between them, we proposed a coupled hybrid continuum-discrete model. In this model, tumor cells, their microenvironment (extracellular matrixes, matrix-degrading enzymes, and tumor angiogenic factors), and their network of blood vessels, described by a series of discrete points, were considered. The results of numerical simulation reveal the process of tumor growth and the change in microenvironment from avascular to vascular stage, indicating that the network of blood vessels develops gradually as the tumor grows. Our findings also reveal that a tumor is divided into three regions: necrotic, semi-necrotic, and well-vascularized. The results agree well with the previous relevant studies and physiological facts, and this model represents a platform for further investigations of tumor therapy.