Pub Date : 2025-11-28DOI: 10.1109/JTEHM.2025.3638856
Jan-Willem Klok;Yannick Smits;Roelf Postema;Asþor T. Steinþorsson;Jenny Dankelman;Tim Horeman
Objective: Grasping force control is crucial for safe laparoscopic surgery. However, force feedback is limited as haptic information on grasping strength and tissue stiffness is mostly lost due to internal instrument backlash and friction. This increases tissue trauma risk as excessive grasping forces can lead to (postoperative) complications. This study aims to develop a grasping force feedback providing add-on for a laparoscopic grasper and to validate its impact on skills acquisition in basic laparoscopic skills training. Method: The ShaftFlex, a shaft-based grasping force measurement system providing feedback was designed as an add-on for standard reusable instruments. It consists of a compliant element deflecting proportionally to the applied grasping force, and a Hall sensor measuring that deflection. Influence on skills acquisition was evaluated in a comparative study where novices were divided into a Feedback and No feedback group, performing five training trials of a silicon torus transfer boxtrainer task. Afterwards, both groups performed a post-training task without feedback. Grasping force, time to completion and number of errors were measured. Results: There was a significant difference in mean grasping force between groups for all training trials and the post-training trial. In the Feedback group, there was no significant increase in grasping force when feedback was removed. Conclusion: The ShaftFlex working principle provided a feasible, sustainable method to measure grasping forces exerted by a laparoscopic grasper, enabling immediate haptic feedback. It potentially enhances objective skill assessment, providing feedback on training performance. In a clinical context, the ShaftFlex might be useful in surgery where delicate tissue is grasped.
{"title":"Design and Validation of a Grasping Force Measuring Vibrotactile Feedback Add-On for Laparoscopic Instruments","authors":"Jan-Willem Klok;Yannick Smits;Roelf Postema;Asþor T. Steinþorsson;Jenny Dankelman;Tim Horeman","doi":"10.1109/JTEHM.2025.3638856","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3638856","url":null,"abstract":"Objective: Grasping force control is crucial for safe laparoscopic surgery. However, force feedback is limited as haptic information on grasping strength and tissue stiffness is mostly lost due to internal instrument backlash and friction. This increases tissue trauma risk as excessive grasping forces can lead to (postoperative) complications. This study aims to develop a grasping force feedback providing add-on for a laparoscopic grasper and to validate its impact on skills acquisition in basic laparoscopic skills training. Method: The ShaftFlex, a shaft-based grasping force measurement system providing feedback was designed as an add-on for standard reusable instruments. It consists of a compliant element deflecting proportionally to the applied grasping force, and a Hall sensor measuring that deflection. Influence on skills acquisition was evaluated in a comparative study where novices were divided into a Feedback and No feedback group, performing five training trials of a silicon torus transfer boxtrainer task. Afterwards, both groups performed a post-training task without feedback. Grasping force, time to completion and number of errors were measured. Results: There was a significant difference in mean grasping force between groups for all training trials and the post-training trial. In the Feedback group, there was no significant increase in grasping force when feedback was removed. Conclusion: The ShaftFlex working principle provided a feasible, sustainable method to measure grasping forces exerted by a laparoscopic grasper, enabling immediate haptic feedback. It potentially enhances objective skill assessment, providing feedback on training performance. In a clinical context, the ShaftFlex might be useful in surgery where delicate tissue is grasped.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"14 ","pages":"1-10"},"PeriodicalIF":4.4,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11271394","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145802348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: To address the limitations of conventional aphasia therapy by developing and clinically evaluating a machine learning based interactive lab for personalized rehabilitation in post-stroke patients. Methods and Procedures: A four week clinical trial was conducted with 27 aphasia patients, randomly assigned to an experimental group ($n=11$ ) using the Language Interactive Lab and a control group ($n=16$ ) receiving conventional therapy. Language performance was assessed using the Chinese Communicative Aphasia Test (CCAT). System interaction data were also used to train classifiers for aphasia severity and recovery tracking. Results: The experimental group showed statistically significant improvements in 7 out of 9 CCAT subtests ($p lt 0.05$ ) and a highly significant total score increase ($p lt 0.001$ ) compared to the control group. Machine learning classifiers achieved up to 91.7% accuracy in predicting aphasia severity and recovery progression. Conclusion: The proposed interactive lab integrates gamified therapy with real time, explainable machine learning assessment, demonstrates clinical efficacy in improving language outcomes, and offers a scalable framework for AI-driven, adaptive neurorehabilitation that has been clinically validated within a hospital setting and designed to align with Taiwan Food and Drug Administration (TFDA) software-as-a-medical-device (SaMD) regulatory principles for translational deployment in clinical environments and hospital investigational use guidelines. Clinical Impact—The integration of gamified digital therapy with machine learning analytics supports personalized, data driven intervention for aphasia rehabilitation in both clinical and home settings, particularly in resource limited environments. Clinical and Translational Impact Statement—This study supports Clinical Research by demonstrating that AI-powered digital therapy significantly improves language outcomes in post-stroke aphasia patients and offers a pathway to scalable, at home neurorehabilitation.
{"title":"Translational Evaluation of a Machine Learning-Based Interactive Lab for Aphasia Rehabilitation in Post Stroke Patients","authors":"Mukul Kumar;Rei-Zhe Wu;Shih-Ching Yeh;Eric Hsiao-Kuang Wu;Po-Yi Tsai","doi":"10.1109/JTEHM.2025.3638643","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3638643","url":null,"abstract":"Objective: To address the limitations of conventional aphasia therapy by developing and clinically evaluating a machine learning based interactive lab for personalized rehabilitation in post-stroke patients. Methods and Procedures: A four week clinical trial was conducted with 27 aphasia patients, randomly assigned to an experimental group (<inline-formula> <tex-math>$n=11$ </tex-math></inline-formula>) using the Language Interactive Lab and a control group (<inline-formula> <tex-math>$n=16$ </tex-math></inline-formula>) receiving conventional therapy. Language performance was assessed using the Chinese Communicative Aphasia Test (CCAT). System interaction data were also used to train classifiers for aphasia severity and recovery tracking. Results: The experimental group showed statistically significant improvements in 7 out of 9 CCAT subtests (<inline-formula> <tex-math>$p lt 0.05$ </tex-math></inline-formula>) and a highly significant total score increase (<inline-formula> <tex-math>$p lt 0.001$ </tex-math></inline-formula>) compared to the control group. Machine learning classifiers achieved up to 91.7% accuracy in predicting aphasia severity and recovery progression. Conclusion: The proposed interactive lab integrates gamified therapy with real time, explainable machine learning assessment, demonstrates clinical efficacy in improving language outcomes, and offers a scalable framework for AI-driven, adaptive neurorehabilitation that has been clinically validated within a hospital setting and designed to align with Taiwan Food and Drug Administration (TFDA) software-as-a-medical-device (SaMD) regulatory principles for translational deployment in clinical environments and hospital investigational use guidelines. Clinical Impact—The integration of gamified digital therapy with machine learning analytics supports personalized, data driven intervention for aphasia rehabilitation in both clinical and home settings, particularly in resource limited environments. Clinical and Translational Impact Statement—This study supports Clinical Research by demonstrating that AI-powered digital therapy significantly improves language outcomes in post-stroke aphasia patients and offers a pathway to scalable, at home neurorehabilitation.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"561-570"},"PeriodicalIF":4.4,"publicationDate":"2025-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11271240","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145729291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-26DOI: 10.1109/JTEHM.2025.3637293
Yu Meng;Javier Garcia-Casado;Gema Prats-Boluda;Jose Luis Martinez-de-Juan;Carmen Padilla Prieto;Rogelio Monfort-Ortiz;Vicente José Diago-Almela;Dongmei Hao;Guangfei Li;Yiyao Ye-Lin
Objective: Electrohysterography (EHG) has been shown to provide valuable information for assessing preterm birth risk. However, few studies have focused on multiple gestations (MG), a well-known risk factor for preterm birth. This study aimed to comprehensively characterize and compare uterine EHG signals between singleton (SG) and MG pregnancies during the third trimester. Method: This prospective cohort study analyzed 383 EHG recordings from 61 SG and 92 MG women during the third trimester. A whole-window approach was used to extract four key EHG features: peak-to-peak amplitude (PPA), Kurtosis of the Hilbert Envelope (KHE), median frequency (MDF) and sample entropy (SampEn). Generalized additive models (GAM) were applied to evaluate temporal trends across gestational age (GA) and gestation type (SG and MG). Results: In SG pregnancies, PPA and KHE progressively increased, with a significant rise in KHE at labour. MDF remained stable until labour, while SampEn gradually declined, especially at term. MG pregnancies showed similar but less pronounced trends: MG exhibited a notably earlier activation of uterine activity than SG before 32 weeks of gestation (WoG), and a slowing-down electrophysiological progression beyond 32 WoG, resulting in similar characteristics with no significant differences. Conclusion: These findings provide electrophysiological evidence suggesting that MG pregnancies may enter a labour-preparatory state earlier, potentially increasing the PTB risk, while the later convergence of EHG features may indicate compensatory mechanisms to delay labour. This work integrates EHG signal analysis with clinical obstetric care, offering valuable insights for clinical management and early PTB risk assessment in MG pregnancies.
{"title":"A Comprehensive Study of Uterine Muscle Activity During the Third Trimester: Comparison of Singleton and Multiple Gestations","authors":"Yu Meng;Javier Garcia-Casado;Gema Prats-Boluda;Jose Luis Martinez-de-Juan;Carmen Padilla Prieto;Rogelio Monfort-Ortiz;Vicente José Diago-Almela;Dongmei Hao;Guangfei Li;Yiyao Ye-Lin","doi":"10.1109/JTEHM.2025.3637293","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3637293","url":null,"abstract":"Objective: Electrohysterography (EHG) has been shown to provide valuable information for assessing preterm birth risk. However, few studies have focused on multiple gestations (MG), a well-known risk factor for preterm birth. This study aimed to comprehensively characterize and compare uterine EHG signals between singleton (SG) and MG pregnancies during the third trimester. Method: This prospective cohort study analyzed 383 EHG recordings from 61 SG and 92 MG women during the third trimester. A whole-window approach was used to extract four key EHG features: peak-to-peak amplitude (PPA), Kurtosis of the Hilbert Envelope (KHE), median frequency (MDF) and sample entropy (SampEn). Generalized additive models (GAM) were applied to evaluate temporal trends across gestational age (GA) and gestation type (SG and MG). Results: In SG pregnancies, PPA and KHE progressively increased, with a significant rise in KHE at labour. MDF remained stable until labour, while SampEn gradually declined, especially at term. MG pregnancies showed similar but less pronounced trends: MG exhibited a notably earlier activation of uterine activity than SG before 32 weeks of gestation (WoG), and a slowing-down electrophysiological progression beyond 32 WoG, resulting in similar characteristics with no significant differences. Conclusion: These findings provide electrophysiological evidence suggesting that MG pregnancies may enter a labour-preparatory state earlier, potentially increasing the PTB risk, while the later convergence of EHG features may indicate compensatory mechanisms to delay labour. This work integrates EHG signal analysis with clinical obstetric care, offering valuable insights for clinical management and early PTB risk assessment in MG pregnancies.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"14 ","pages":"11-18"},"PeriodicalIF":4.4,"publicationDate":"2025-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11269692","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145802383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cervical intraepithelial neoplasia (CIN) represents a spectrum of premalignant lesions requiring accurate early detection to prevent progression to invasive cervical cancer. Colposcopy with visual inspection using acetic acid (VIA) is the gold standard for CIN assessment but suffers from substantial interobserver variability, limiting diagnostic consistency. We evaluated hyperspectral imaging (HSI) as an objective, non-invasive method for characterizing CIN-related tissue changes. This prospective proof-of-principle clinical study enrolled women with histologically confirmed CIN3 indicated for large-loop excision of the transformation zone (LLETZ). Standardized colposcopic images following VIA were obtained and annotated independently by five certified colposcopists according to IFCPC Rio 2011 classification. These annotations served as pathological tissue region references and were quantitatively assessed using intersection over union metrics to evaluate interobserver agreement. HSI was performed immediately prior to LLETZ using the TIVITA Tissue System, capturing spectral reflectance data across 500–995 nm in 100 wavelength bands. Spatial correspondence between colposcopic and hyperspectral images was achieved through homography transformation based on landmark alignment, allowing expert annotations to be projected into the HSI domain. Reflectance spectra from annotated areas were averaged to calculate four proprietary HSI-derived tissue indices, which revealed significantly higher values in CIN-affected regions compared to healthy tissue (p <0.01, Wilcoxon signed-rank test), suggesting increased vascularization and water content. Our findings highlight conventional colposcopy limitations due to examiner subjectivity and support HSI’s potential to provide reproducible, quantitative biomarkers for CIN. HSI integration into clinical workflows may enhance cervical cancer screening objectivity and enable reliable diagnostics in resource-limited settings. Clinical and Translational Impact Statement— Hyperspectral imaging enables objective detection of cervical intraepithelial neoplasia and could improve diagnostic accuracy while reducing unnecessary biopsies
{"title":"Detection of Cervical Intraepithelial Neoplasia Using Hyperspectral Tissue Signatures","authors":"Ovidiu Jurjuţ;Martin Weiss;Yannick Daniel;Sabine Matovina;Felix Neis;Katharina Rall;Katharina Schöpp;Melanie Henes;Walter Linzenbold;Sara Y. Brucker;Jürgen Andress","doi":"10.1109/JTEHM.2025.3630878","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3630878","url":null,"abstract":"Cervical intraepithelial neoplasia (CIN) represents a spectrum of premalignant lesions requiring accurate early detection to prevent progression to invasive cervical cancer. Colposcopy with visual inspection using acetic acid (VIA) is the gold standard for CIN assessment but suffers from substantial interobserver variability, limiting diagnostic consistency. We evaluated hyperspectral imaging (HSI) as an objective, non-invasive method for characterizing CIN-related tissue changes. This prospective proof-of-principle clinical study enrolled women with histologically confirmed CIN3 indicated for large-loop excision of the transformation zone (LLETZ). Standardized colposcopic images following VIA were obtained and annotated independently by five certified colposcopists according to IFCPC Rio 2011 classification. These annotations served as pathological tissue region references and were quantitatively assessed using intersection over union metrics to evaluate interobserver agreement. HSI was performed immediately prior to LLETZ using the TIVITA Tissue System, capturing spectral reflectance data across 500–995 nm in 100 wavelength bands. Spatial correspondence between colposcopic and hyperspectral images was achieved through homography transformation based on landmark alignment, allowing expert annotations to be projected into the HSI domain. Reflectance spectra from annotated areas were averaged to calculate four proprietary HSI-derived tissue indices, which revealed significantly higher values in CIN-affected regions compared to healthy tissue (p <0.01, Wilcoxon signed-rank test), suggesting increased vascularization and water content. Our findings highlight conventional colposcopy limitations due to examiner subjectivity and support HSI’s potential to provide reproducible, quantitative biomarkers for CIN. HSI integration into clinical workflows may enhance cervical cancer screening objectivity and enable reliable diagnostics in resource-limited settings. Clinical and Translational Impact Statement— Hyperspectral imaging enables objective detection of cervical intraepithelial neoplasia and could improve diagnostic accuracy while reducing unnecessary biopsies","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"532-539"},"PeriodicalIF":4.4,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11236451","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145612106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: Environmental noise poses a major barrier to the accuracy of self-administered hearing tests conducted outside clinical settings. There is a pressing need for effective noise control solutions to enable reliable hearing threshold measurements in everyday environments. This study introduces an optimized active noise cancellation (ANC) technique based on auditory masking characteristics. Method: The method was implemented in a mobile hearing test system using calibrated true wireless Bluetooth earphones. Electroacoustic validation and clinical testing were conducted across four ANC scenarios: normal, generic ANC off, generic ANC on, and optimized ANC on in 65 dB(A) pink noise. Results: A total of 50 participants completed hearing tests at eight frequencies (0.25–8 kHz), and results were compared to standard audiometry. The optimized ANC yielded the highest signal-to-noise ratio in noisy conditions and demonstrated strong agreement with standard hearing thresholds (r = 0.99, p <.01) in normal environments. Under 65 dB(A) noise, the proposed method significantly outperformed generic ANC with smaller hearing measurement error, improving threshold accuracy across most frequencies. Conclusion: The proposed ANC technique enhances hearing test reliability in noisy conditions, supporting accurate, self-administered hearing assessments outside clinical settings. This technology has strong potential for home or community-based hearing healthcare applications.
目的：环境噪声是在临床环境之外进行的自我听力测试准确性的主要障碍。迫切需要有效的噪声控制解决方案，以便在日常环境中实现可靠的听力阈值测量。本文介绍了一种基于听觉掩蔽特性的优化主动降噪技术。方法：采用校准后的真无线蓝牙耳机在移动听力测试系统中实施该方法。电声验证和临床测试在四种情况下进行：正常、普通ANC关闭、普通ANC打开和65 dB(A)粉红噪声下的优化ANC打开。结果：共有50名参与者完成了8个频率（0.25-8 kHz）的听力测试，并将结果与标准听力学进行了比较。优化后的ANC在噪声条件下产生最高的信噪比，与正常环境下的标准听力阈值非常吻合（r = 0.99, p < 0.01）。在65 dB(A)噪声下，该方法显著优于一般的ANC，具有较小的听力测量误差，提高了大多数频率的阈值精度。结论：提出的ANC技术提高了噪声条件下听力测试的可靠性，支持临床之外准确的、自我管理的听力评估。这项技术在家庭或社区听力保健应用方面具有很大的潜力。
{"title":"Optimized Active Noise Cancellation for Hearing Tests Using Auditory Masking Characteristics","authors":"Hsiu-Lien Cheng;Ying-Hui Lai;Po-Hsun Huang;Wen-Huei Liao","doi":"10.1109/JTEHM.2025.3629999","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3629999","url":null,"abstract":"Objective: Environmental noise poses a major barrier to the accuracy of self-administered hearing tests conducted outside clinical settings. There is a pressing need for effective noise control solutions to enable reliable hearing threshold measurements in everyday environments. This study introduces an optimized active noise cancellation (ANC) technique based on auditory masking characteristics. Method: The method was implemented in a mobile hearing test system using calibrated true wireless Bluetooth earphones. Electroacoustic validation and clinical testing were conducted across four ANC scenarios: normal, generic ANC off, generic ANC on, and optimized ANC on in 65 dB(A) pink noise. Results: A total of 50 participants completed hearing tests at eight frequencies (0.25–8 kHz), and results were compared to standard audiometry. The optimized ANC yielded the highest signal-to-noise ratio in noisy conditions and demonstrated strong agreement with standard hearing thresholds (r = 0.99, p <.01) in normal environments. Under 65 dB(A) noise, the proposed method significantly outperformed generic ANC with smaller hearing measurement error, improving threshold accuracy across most frequencies. Conclusion: The proposed ANC technique enhances hearing test reliability in noisy conditions, supporting accurate, self-administered hearing assessments outside clinical settings. This technology has strong potential for home or community-based hearing healthcare applications.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"540-551"},"PeriodicalIF":4.4,"publicationDate":"2025-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11230825","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145674800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-24DOI: 10.1109/JTEHM.2025.3625388
Verónica Barroso-García;Fernando Vaquerizo-Villar;Gonzalo C. Gutiérrez-Tobal;Ehab Dayyat;David Gozal;Timo Leppänen;Roberto Hornero
Objective: Approaches based on a single-channel airflow has shown great potential for simplifying pediatric obstructive sleep apnea (OSA) diagnosis. However, analysis has been limited to feature-engineering techniques, restricting identification of complex respiratory patterns, and reducing diagnostic performance in automated models. Here, we propose deep-learning and explainable artificial intelligence (XAI) to estimate the pediatric OSA severity from airflow, while ensuring transparency in automatic decisions. Technology or Method: We used 3,672 overnight airflow recordings from four pediatric datasets. A convolutional neural network (CNN)-based regression model was trained to estimate the apnea-hypopnea index (AHI) and predict OSA severity. We evaluated and compared Gradient-Weighted Class Activation Mapping (Grad-CAM) and SHapley Additive exPlanations (SHAP) to identify the airflow regions where the CNN focuses for predictions. Results: The proposed model demonstrated high concordance between the actual and estimated AHI (intraclass correlation coefficient from 0.69 to 0.87 in the test group), and high diagnostic performance: four-class Cohen’s kappa between 0.37 and 0.43 and accuracies of 82.03%, 97.09%, and 99.03% for three OSA severity cutoffs (i.e. 1, 5, and 10 e/h) in the test group. The interpretability analysis with Grad-CAM and SHAP revealed that the CNN accurately identifies apneic events by focusing on their onset and offset. Both techniques provided complementary information about the model’s decision-making. While Grad-CAM highlighted respiratory events with abrupt signal changes, SHAP captured more subtle patterns with noise included. Conclusions: Accordingly, our model can help automatically detect pediatric OSA and offers clinicians an explainable approach that enhances credibility and usability, thus providing a path toward clinical translation in early diagnosis. Clinical Impact: This study presents an interpretable deep-learning tool using airflow to accurately detect pediatric obstructive sleep apnea, enabling early, objective diagnosis and supporting clinical decision-making through identification of relevant respiratory patterns.
{"title":"An Explainable Deep-Learning Approach to Detect Pediatric Sleep Apnea From Single-Channel Airflow","authors":"Verónica Barroso-García;Fernando Vaquerizo-Villar;Gonzalo C. Gutiérrez-Tobal;Ehab Dayyat;David Gozal;Timo Leppänen;Roberto Hornero","doi":"10.1109/JTEHM.2025.3625388","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3625388","url":null,"abstract":"Objective: Approaches based on a single-channel airflow has shown great potential for simplifying pediatric obstructive sleep apnea (OSA) diagnosis. However, analysis has been limited to feature-engineering techniques, restricting identification of complex respiratory patterns, and reducing diagnostic performance in automated models. Here, we propose deep-learning and explainable artificial intelligence (XAI) to estimate the pediatric OSA severity from airflow, while ensuring transparency in automatic decisions. Technology or Method: We used 3,672 overnight airflow recordings from four pediatric datasets. A convolutional neural network (CNN)-based regression model was trained to estimate the apnea-hypopnea index (AHI) and predict OSA severity. We evaluated and compared Gradient-Weighted Class Activation Mapping (Grad-CAM) and SHapley Additive exPlanations (SHAP) to identify the airflow regions where the CNN focuses for predictions. Results: The proposed model demonstrated high concordance between the actual and estimated AHI (intraclass correlation coefficient from 0.69 to 0.87 in the test group), and high diagnostic performance: four-class Cohen’s kappa between 0.37 and 0.43 and accuracies of 82.03%, 97.09%, and 99.03% for three OSA severity cutoffs (i.e. 1, 5, and 10 e/h) in the test group. The interpretability analysis with Grad-CAM and SHAP revealed that the CNN accurately identifies apneic events by focusing on their onset and offset. Both techniques provided complementary information about the model’s decision-making. While Grad-CAM highlighted respiratory events with abrupt signal changes, SHAP captured more subtle patterns with noise included. Conclusions: Accordingly, our model can help automatically detect pediatric OSA and offers clinicians an explainable approach that enhances credibility and usability, thus providing a path toward clinical translation in early diagnosis. Clinical Impact: This study presents an interpretable deep-learning tool using airflow to accurately detect pediatric obstructive sleep apnea, enabling early, objective diagnosis and supporting clinical decision-making through identification of relevant respiratory patterns.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"517-531"},"PeriodicalIF":4.4,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11216356","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510122","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-24DOI: 10.1109/JTEHM.2025.3625144
Beatriz S. Arruda;Moaad Benjaber;John Fleming;Robert Toth;Colin G. McNamara;Andrew Sharott;Timothy Denison;Hayriye Cagnan
Background: Tremor is the most common movement disorder and a prevalent symptom of neurodegenerative conditions such as Parkinson’s disease (PD). Given the limitations of medication, which may not effectively treat tremor, and the limited availability of surgical treatments such as deep brain stimulation, there is a pressing clinical need for non-invasive therapeutic alternatives, including peripheral electrical stimulation. The high variability of PD tremor poses a challenge to such therapies and calls for person-specific stimulation parameters. Methods: We developed a wrist-worn system incorporating an adaptable phase-tracking algorithm designed for real-time estimation of Parkinsonian rest tremor phase. The algorithm dynamically adapts to tremor variability, including changes in the axis of maximum excursion and center frequency. The system was first validated offline, followed by in-clinic feasibility testing in three individuals with PD. The system triggered the delivery of both phasic and open-loop electrical stimulation to the participant’s wrist. Results: Robust phase estimation was achieved both offline and in all participants. The system adapted to changes in tremor dominant axis and center frequency. Modest tremor modulation was observed at select person-specific settings. Conclusion: This work provides a novel platform for research involving tremor phase tracking, accounting for PD tremor variability, and a foundation for developing personalized, non-invasive tremor management strategies. Clinical and Translational Impact Statement—This study presents a wearable system for adaptive tremor phase tracking validated in individuals with Parkinson’s disease and establishes a foundation for further development of personalized non-invasive tremor management strategies. Category: Clinical Research
{"title":"An Adaptable Phase-Tracking System for Parkinsonian Rest Tremor: Design and In-Clinic Feasibility","authors":"Beatriz S. Arruda;Moaad Benjaber;John Fleming;Robert Toth;Colin G. McNamara;Andrew Sharott;Timothy Denison;Hayriye Cagnan","doi":"10.1109/JTEHM.2025.3625144","DOIUrl":"10.1109/JTEHM.2025.3625144","url":null,"abstract":"Background: Tremor is the most common movement disorder and a prevalent symptom of neurodegenerative conditions such as Parkinson’s disease (PD). Given the limitations of medication, which may not effectively treat tremor, and the limited availability of surgical treatments such as deep brain stimulation, there is a pressing clinical need for non-invasive therapeutic alternatives, including peripheral electrical stimulation. The high variability of PD tremor poses a challenge to such therapies and calls for person-specific stimulation parameters. Methods: We developed a wrist-worn system incorporating an adaptable phase-tracking algorithm designed for real-time estimation of Parkinsonian rest tremor phase. The algorithm dynamically adapts to tremor variability, including changes in the axis of maximum excursion and center frequency. The system was first validated offline, followed by in-clinic feasibility testing in three individuals with PD. The system triggered the delivery of both phasic and open-loop electrical stimulation to the participant’s wrist. Results: Robust phase estimation was achieved both offline and in all participants. The system adapted to changes in tremor dominant axis and center frequency. Modest tremor modulation was observed at select person-specific settings. Conclusion: This work provides a novel platform for research involving tremor phase tracking, accounting for PD tremor variability, and a foundation for developing personalized, non-invasive tremor management strategies. Clinical and Translational Impact Statement—This study presents a wearable system for adaptive tremor phase tracking validated in individuals with Parkinson’s disease and establishes a foundation for further development of personalized non-invasive tremor management strategies. Category: Clinical Research","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"507-516"},"PeriodicalIF":4.4,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12599896/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145497553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-22DOI: 10.1109/JTEHM.2025.3624469
Aaesha Alzaabi;Imran Saied;Tughrul Arslan
Objective: This study describes the design and evaluation of volunteer user trials of an unobtrusive Wi-Fi Channel State Information (CSI) vital sign sensing system in older participants aged 60 years and older in different home environments. Methods and procedures: In terms of experiment design, the implementation of user-centric sensor placement and integration informed consent with various experimental elements in the design of experiments of older people. The implemented signal processing algorithm, which extracts vital signs from the Wi-Fi CSI signal to obtain respiration and heart rate measurements, employs wavelet filtering techniques. For selecting of vital sign signals from the 52 CSI subcarriers, the Principal Component Sample Entropy (PC-SampEn) was implemented to capture the information most relevant to vital signs.Results: Two cardiorespiratory vital sign measurements were validated against wearable ground-truth devices, a respiratory belt and a photoplethysmogram (PPG). The results demonstrated an expected decrease in accuracy and measurement agreement in uncontrolled home environments.Conclusion: Although respiratory rate measurements have demonstrated promising accuracy and agreement in uncontrolled environments, heart rate measurements observed high variability in these scenarios due to challenging signal extraction. Further experiments must be conducted to address the limitation in sample size and the technical challenges in heart rate signal extraction to improve accuracy. Clinical and Translational Impact: This study provides a design of unobtrusive care technology for vital sign sensing for older adults, demonstrated and evaluated in the context of in-home monitoring for healthcare.
{"title":"Design and Evaluation of Volunteer User Trials of Unobtrusive Vital Signs Monitoring for Older People in Care Using Wi-Fi CSI Sensing","authors":"Aaesha Alzaabi;Imran Saied;Tughrul Arslan","doi":"10.1109/JTEHM.2025.3624469","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3624469","url":null,"abstract":"Objective: This study describes the design and evaluation of volunteer user trials of an unobtrusive Wi-Fi Channel State Information (CSI) vital sign sensing system in older participants aged 60 years and older in different home environments. Methods and procedures: In terms of experiment design, the implementation of user-centric sensor placement and integration informed consent with various experimental elements in the design of experiments of older people. The implemented signal processing algorithm, which extracts vital signs from the Wi-Fi CSI signal to obtain respiration and heart rate measurements, employs wavelet filtering techniques. For selecting of vital sign signals from the 52 CSI subcarriers, the Principal Component Sample Entropy (PC-SampEn) was implemented to capture the information most relevant to vital signs.Results: Two cardiorespiratory vital sign measurements were validated against wearable ground-truth devices, a respiratory belt and a photoplethysmogram (PPG). The results demonstrated an expected decrease in accuracy and measurement agreement in uncontrolled home environments.Conclusion: Although respiratory rate measurements have demonstrated promising accuracy and agreement in uncontrolled environments, heart rate measurements observed high variability in these scenarios due to challenging signal extraction. Further experiments must be conducted to address the limitation in sample size and the technical challenges in heart rate signal extraction to improve accuracy. Clinical and Translational Impact: This study provides a design of unobtrusive care technology for vital sign sensing for older adults, demonstrated and evaluated in the context of in-home monitoring for healthcare.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"480-492"},"PeriodicalIF":4.4,"publicationDate":"2025-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11214363","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145455865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective: Therapeutic drug monitoring (TDM) is essential for managing medication dosages in critically ill patients, particularly for antibiotics such as vancomycin. The dynamic physiological conditions of critically ill patients require frequent monitoring of vancomycin levels to ensure therapeutic therapeutic efficacy while minimizing toxicity. Traditional Bayesian methods and pharmacokinetic (PK) models often fail because of the complex and unpredictable nature of these patients’ conditions, as well as the limitations of standard PK modeling.Methods and procedures: This study aimed to establish a gated recurrent unit (GRU)-integrated joint multilayer perceptron network (GointMLP) model to predict sequential vancomycin TDM levels in patients in the intensive care unit. The proposed model consists of three modules to maintain consistent therapeutic vancomycin concentrations while accommodating individual patient differences. By integrating regression and classification predictions, GointMLP provides a dual mechanism for clinicians to verify the reliability of predicted values for informed decision-making. Additionally, we have developed DeepTDM, a comprehensive decision support system designed for real-time vancomycin dose optimization to enhance clinical outcomes.Results: The GointMLP provides more accurate predictions compared to traditional PK models and other machine learning/deep learning approaches. This superior performance is demonstrated not only in local validation cohorts but also in the ethnically diverse MIMIC-IV dataset, validating the model’s robust generalizability.Conclusion: This work addresses the limitations of current methodologies while leveraging advancements in deep learning techniques, particularly demonstrating the effectiveness of GointMLP in enhancing patient outcomes through precise TDM. Efforts are underway to integrate DeepTDM into clinical practice, with the anticipation that it will not only support clinicians in decision-making but also substantially improve therapeutic outcomes for patients undergoing vancomycin therapy. Clinical and Translational Impact Statement: The proposed model and software enable individualized vancomycin dosing for critically ill patients, improving precision dosing and supporting seamless integration into clinical workflows
{"title":"DeepTDM: Deep Learning-Based Prediction of Sequential Therapeutic Drug Monitoring Levels of Vancomycin","authors":"Jinkyeong Park;Dohyun Kim;Donghoon Lee;Minkyu Kim;Yoon Kim;Seon-Sook Han;Yeonjeong Heo;Hyun-Soo Choi","doi":"10.1109/JTEHM.2025.3623605","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3623605","url":null,"abstract":"Objective: Therapeutic drug monitoring (TDM) is essential for managing medication dosages in critically ill patients, particularly for antibiotics such as vancomycin. The dynamic physiological conditions of critically ill patients require frequent monitoring of vancomycin levels to ensure therapeutic therapeutic efficacy while minimizing toxicity. Traditional Bayesian methods and pharmacokinetic (PK) models often fail because of the complex and unpredictable nature of these patients’ conditions, as well as the limitations of standard PK modeling.Methods and procedures: This study aimed to establish a gated recurrent unit (GRU)-integrated joint multilayer perceptron network (GointMLP) model to predict sequential vancomycin TDM levels in patients in the intensive care unit. The proposed model consists of three modules to maintain consistent therapeutic vancomycin concentrations while accommodating individual patient differences. By integrating regression and classification predictions, GointMLP provides a dual mechanism for clinicians to verify the reliability of predicted values for informed decision-making. Additionally, we have developed DeepTDM, a comprehensive decision support system designed for real-time vancomycin dose optimization to enhance clinical outcomes.Results: The GointMLP provides more accurate predictions compared to traditional PK models and other machine learning/deep learning approaches. This superior performance is demonstrated not only in local validation cohorts but also in the ethnically diverse MIMIC-IV dataset, validating the model’s robust generalizability.Conclusion: This work addresses the limitations of current methodologies while leveraging advancements in deep learning techniques, particularly demonstrating the effectiveness of GointMLP in enhancing patient outcomes through precise TDM. Efforts are underway to integrate DeepTDM into clinical practice, with the anticipation that it will not only support clinicians in decision-making but also substantially improve therapeutic outcomes for patients undergoing vancomycin therapy. Clinical and Translational Impact Statement: The proposed model and software enable individualized vancomycin dosing for critically ill patients, improving precision dosing and supporting seamless integration into clinical workflows","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"493-506"},"PeriodicalIF":4.4,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11208607","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145455835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-09DOI: 10.1109/JTEHM.2025.3619802
Kelly Long;Ganesh M. Babulal;Sayeh Bayat
Objective: To examine how early pathophysiological changes in Alzheimer’s disease (AD) affect navigational decision-making by analyzing the complexity of driving routes in older adults with and without preclinical AD. Methods: We developed a novel route complexity metric based on the number of left and right turns and the deviation from the most direct path, accounting for cognitive load during navigation. Naturalistic GPS driving data were collected for a year from 111 older adults aged 65–85, with preclinical AD status determined via cerebrospinal fluid amyloid biomarkers. A multiple linear regression model was used to assess the relationship between age, preclinical AD status, and route complexity. Results: The findings of this study indicate that preclinical AD may influence the navigational abilities of older adults. After controlling for age, participants with preclinical AD chose routes with higher baseline complexity than the control group. It further revealed that participants with preclinical AD selected routes with lower complexity as they aged—a trend not observed in healthy controls. Conclusion: Preclinical AD is associated with changes in spatial decision-making that are observable in real-world driving behaviours. The age-related decline in route complexity among those with preclinical AD may reflect compensatory strategies or progressive cognitive changes. Clinical Impact: This study presents a non-invasive, behaviour-based metric that could support early detection of cognitive decline. It may also inform the design of personalized mobility interventions and dementia-friendly mobility systems.
{"title":"Characterizing Navigational Changes in Preclinical Alzheimer’s Disease: A Route Complexity Metric Derived From Naturalistic Driving Data","authors":"Kelly Long;Ganesh M. Babulal;Sayeh Bayat","doi":"10.1109/JTEHM.2025.3619802","DOIUrl":"https://doi.org/10.1109/JTEHM.2025.3619802","url":null,"abstract":"Objective: To examine how early pathophysiological changes in Alzheimer’s disease (AD) affect navigational decision-making by analyzing the complexity of driving routes in older adults with and without preclinical AD. Methods: We developed a novel route complexity metric based on the number of left and right turns and the deviation from the most direct path, accounting for cognitive load during navigation. Naturalistic GPS driving data were collected for a year from 111 older adults aged 65–85, with preclinical AD status determined via cerebrospinal fluid amyloid biomarkers. A multiple linear regression model was used to assess the relationship between age, preclinical AD status, and route complexity. Results: The findings of this study indicate that preclinical AD may influence the navigational abilities of older adults. After controlling for age, participants with preclinical AD chose routes with higher baseline complexity than the control group. It further revealed that participants with preclinical AD selected routes with lower complexity as they aged—a trend not observed in healthy controls. Conclusion: Preclinical AD is associated with changes in spatial decision-making that are observable in real-world driving behaviours. The age-related decline in route complexity among those with preclinical AD may reflect compensatory strategies or progressive cognitive changes. Clinical Impact: This study presents a non-invasive, behaviour-based metric that could support early detection of cognitive decline. It may also inform the design of personalized mobility interventions and dementia-friendly mobility systems.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"13 ","pages":"471-479"},"PeriodicalIF":4.4,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11197551","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145352142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：