Pub Date : 2024-08-23DOI: 10.1109/JTEHM.2024.3448457
Chayut Bunterngchit;Jiaxing Wang;Zeng-Guang Hou
The integration of electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) can facilitate the advancement of brain-computer interfaces (BCIs). However, existing research in this domain has grappled with the challenge of the efficient selection of features, resulting in the underutilization of the temporal richness of EEG and the spatial specificity of fNIRS data.To effectively address this challenge, this study proposed a deep learning architecture called the multimodal DenseNet fusion (MDNF) model that was trained on two-dimensional (2D) EEG data images, leveraging advanced feature extraction techniques. The model transformed EEG data into 2D images using a short-time Fourier transform, applied transfer learning to extract discriminative features, and consequently integrated them with fNIRS-derived spectral entropy features. This approach aimed to bridge existing gaps in EEG-fNIRS-based BCI research by enhancing classification accuracy and versatility across various cognitive and motor imagery tasks.Experimental results on two public datasets demonstrated the superiority of our model over existing state-of-the-art methods.Thus, the high accuracy and precise feature utilization of the MDNF model demonstrates the potential in clinical applications for neurodiagnostics and rehabilitation, thereby paving the method for patient-specific therapeutic strategies.
{"title":"Simultaneous EEG-fNIRS Data Classification Through Selective Channel Representation and Spectrogram Imaging","authors":"Chayut Bunterngchit;Jiaxing Wang;Zeng-Guang Hou","doi":"10.1109/JTEHM.2024.3448457","DOIUrl":"https://doi.org/10.1109/JTEHM.2024.3448457","url":null,"abstract":"The integration of electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) can facilitate the advancement of brain-computer interfaces (BCIs). However, existing research in this domain has grappled with the challenge of the efficient selection of features, resulting in the underutilization of the temporal richness of EEG and the spatial specificity of fNIRS data.To effectively address this challenge, this study proposed a deep learning architecture called the multimodal DenseNet fusion (MDNF) model that was trained on two-dimensional (2D) EEG data images, leveraging advanced feature extraction techniques. The model transformed EEG data into 2D images using a short-time Fourier transform, applied transfer learning to extract discriminative features, and consequently integrated them with fNIRS-derived spectral entropy features. This approach aimed to bridge existing gaps in EEG-fNIRS-based BCI research by enhancing classification accuracy and versatility across various cognitive and motor imagery tasks.Experimental results on two public datasets demonstrated the superiority of our model over existing state-of-the-art methods.Thus, the high accuracy and precise feature utilization of the MDNF model demonstrates the potential in clinical applications for neurodiagnostics and rehabilitation, thereby paving the method for patient-specific therapeutic strategies.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"600-612"},"PeriodicalIF":3.7,"publicationDate":"2024-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10643872","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142137537","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: Optimized deep brain stimulation (DBS) is fast becoming a therapy of choice for the treatment of Parkinson’s disease (PD). However, the post-operative optimization (aimed at maximizing patient clinical benefits and minimizing adverse effects) of all possible DBS parameter settings using the standard-of-care clinical protocol requires numerous clinical visits, which substantially increases the time to optimization per patient (TPP), patient cost burden and limit the number of patients who can undergo DBS treatment. The TPP is further elongated in electrodes with stimulation directionality or in diseases with latency in clinical feedback. In this work, we proposed a deep learning and fMRI-based pipeline for DBS optimization that can potentially reduce the TPP from ~1 year to a few hours during a single clinical visit.Methods and procedures: We developed an unsupervised autoencoder (AE)-based model to extract meaningful features from 122 previously acquired blood oxygenated level dependent (BOLD) fMRI datasets from 39 a priori clinically optimized PD patients undergoing DBS therapy. The extracted features are then fed into multilayer perceptron (MLP)-based parameter classification and prediction models for rapid DBS parameter optimization.Results: The AE-extracted features of optimal and non-optimal DBS were disentangled. The AE-MLP classification model yielded accuracy, precision, recall, F1 score, and combined AUC of 0.96 ± 0.04, 0.95 ± 0.07, 0.92 ± 0.07, 0.93 ± 0.06, and 0.98 respectively. Accuracies of 0.79 ± 0.04, 0.85 ± 0.04, 0.82 ± 0.05, 0.83 ± 0.05, and 0.70 ± 0.07 were obtained in the prediction of voltage, frequency, and x-y-z contact locations, respectively.Conclusion: The proposed AE-MLP models yielded promising results for fMRI-based DBS parameter classification and prediction, potentially facilitating rapid semi-automated DBS parameter optimization. Clinical and Translational Impact Statement—A deep learning-based pipeline for semi-automated DBS parameter optimization is presented, with the potential to significantly decrease the optimization duration per patient and patients' financial burden while increasing patient throughput.
{"title":"Deep Learning and fMRI-Based Pipeline for Optimization of Deep Brain Stimulation During Parkinson’s Disease Treatment: Toward Rapid Semi-Automated Stimulation Optimization","authors":"Jianwei Qiu;Afis Ajala;John Karigiannis;Jürgen Germann;Brendan Santyr;Aaron Loh;Luca Marinelli;Thomas Foo;Radhika Madhavan;Desmond Yeo;Alexandre Boutet;Andres Lozano","doi":"10.1109/JTEHM.2024.3448392","DOIUrl":"https://doi.org/10.1109/JTEHM.2024.3448392","url":null,"abstract":"Objective: Optimized deep brain stimulation (DBS) is fast becoming a therapy of choice for the treatment of Parkinson’s disease (PD). However, the post-operative optimization (aimed at maximizing patient clinical benefits and minimizing adverse effects) of all possible DBS parameter settings using the standard-of-care clinical protocol requires numerous clinical visits, which substantially increases the time to optimization per patient (TPP), patient cost burden and limit the number of patients who can undergo DBS treatment. The TPP is further elongated in electrodes with stimulation directionality or in diseases with latency in clinical feedback. In this work, we proposed a deep learning and fMRI-based pipeline for DBS optimization that can potentially reduce the TPP from ~1 year to a few hours during a single clinical visit.Methods and procedures: We developed an unsupervised autoencoder (AE)-based model to extract meaningful features from 122 previously acquired blood oxygenated level dependent (BOLD) fMRI datasets from 39 a priori clinically optimized PD patients undergoing DBS therapy. The extracted features are then fed into multilayer perceptron (MLP)-based parameter classification and prediction models for rapid DBS parameter optimization.Results: The AE-extracted features of optimal and non-optimal DBS were disentangled. The AE-MLP classification model yielded accuracy, precision, recall, F1 score, and combined AUC of 0.96 ± 0.04, 0.95 ± 0.07, 0.92 ± 0.07, 0.93 ± 0.06, and 0.98 respectively. Accuracies of 0.79 ± 0.04, 0.85 ± 0.04, 0.82 ± 0.05, 0.83 ± 0.05, and 0.70 ± 0.07 were obtained in the prediction of voltage, frequency, and x-y-z contact locations, respectively.Conclusion: The proposed AE-MLP models yielded promising results for fMRI-based DBS parameter classification and prediction, potentially facilitating rapid semi-automated DBS parameter optimization. Clinical and Translational Impact Statement—A deep learning-based pipeline for semi-automated DBS parameter optimization is presented, with the potential to significantly decrease the optimization duration per patient and patients' financial burden while increasing patient throughput.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"589-599"},"PeriodicalIF":3.7,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10643605","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142123042","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 : 2024-08-20DOI: 10.1109/JTEHM.2024.3446511
Rafael Morales Mayoral;Ameer Helmi;Samuel W. Logan;Naomi T. Fitter
Children worldwide are becoming increasingly inactive, leading to significant wellness challenges. Initial findings from our research team indicate that robots could potentially provide a more effective approach (compared to other age-appropriate toys) for encouraging physical activity in children. However, the basis of this past work relied on either interactions with groups of children (making it challenging to isolate specific factors that influenced activity levels) or a preliminary version of results of the present study (which centered on just a single more exploratory method for assessing child movement). This paper delves into more controlled interactions involving a single robot and a child participant, while also considering observations over an extended period to mitigate the influence of novelty on the study outcomes. We discuss the outcomes of a two-month-long deployment, during which $N=8$ participants engaged with our custom robot, GoBot, in weekly sessions. During each session, the children experienced three different conditions: a teleoperated robot mode, a semi-autonomous robot mode, and a control condition in which the robot was present but inactive. Compared to our past related work, the results expanded our findings by confirming with greater clout (based on multiple data streams, including one more robust measure compared to the past related work) that children tended to be more physically active when the robot was active, and interestingly, there were no significant differences between the teleoperated and semi-autonomous modes in terms of our study measures. These insights can inform future applications of assistive robots in child motor interventions, including the guiding of appropriate levels of autonomy for these systems. This study demonstrates that incorporating robotic systems into play environments can boost physical activity in young children, indicating potential implementation in settings crafted to enhance children’s physical movement.
{"title":"GoBot Go! Using a Custom Assistive Robot to Promote Physical Activity in Children","authors":"Rafael Morales Mayoral;Ameer Helmi;Samuel W. Logan;Naomi T. Fitter","doi":"10.1109/JTEHM.2024.3446511","DOIUrl":"https://doi.org/10.1109/JTEHM.2024.3446511","url":null,"abstract":"Children worldwide are becoming increasingly inactive, leading to significant wellness challenges. Initial findings from our research team indicate that robots could potentially provide a more effective approach (compared to other age-appropriate toys) for encouraging physical activity in children. However, the basis of this past work relied on either interactions with groups of children (making it challenging to isolate specific factors that influenced activity levels) or a preliminary version of results of the present study (which centered on just a single more exploratory method for assessing child movement). This paper delves into more controlled interactions involving a single robot and a child participant, while also considering observations over an extended period to mitigate the influence of novelty on the study outcomes. We discuss the outcomes of a two-month-long deployment, during which \u0000<inline-formula> <tex-math>$N=8$ </tex-math></inline-formula>\u0000 participants engaged with our custom robot, GoBot, in weekly sessions. During each session, the children experienced three different conditions: a teleoperated robot mode, a semi-autonomous robot mode, and a control condition in which the robot was present but inactive. Compared to our past related work, the results expanded our findings by confirming with greater clout (based on multiple data streams, including one more robust measure compared to the past related work) that children tended to be more physically active when the robot was active, and interestingly, there were no significant differences between the teleoperated and semi-autonomous modes in terms of our study measures. These insights can inform future applications of assistive robots in child motor interventions, including the guiding of appropriate levels of autonomy for these systems. This study demonstrates that incorporating robotic systems into play environments can boost physical activity in young children, indicating potential implementation in settings crafted to enhance children’s physical movement.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"613-621"},"PeriodicalIF":3.7,"publicationDate":"2024-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10640114","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142143707","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 : 2024-07-29DOI: 10.1109/JTEHM.2024.3435334
Nikolas Hesse;Sandra Baumgartner;Anja Gut;Hubertus J. A. Van Hedel
Objective: Low-cost, portable RGB-D cameras with integrated motion tracking functionality enable easy-to-use 3D motion analysis without requiring expensive facilities and specialized personnel. However, the accuracy of existing systems is insufficient for most clinical applications, particularly when applied to children. In previous work, we developed an RGB-D camera-based motion tracking method and showed that it accurately captures body joint positions of children and young adults in 3D. In this study, the validity and accuracy of clinically relevant motion parameters that were computed from kinematics of our motion tracking method are evaluated in children and young adults. Methods: Twenty-three typically developing children and healthy young adults (5-29 years, 110–189 cm) performed five movement tasks while being recorded simultaneously with a marker-based Vicon system and an Azure Kinect RGB-D camera. Motion parameters were computed from the extracted kinematics of both methods: time series measurements, i.e., measurements over time, peak measurements, i.e., measurements at a single time instant, and movement smoothness. The agreement of these parameter values was evaluated using Pearson’s correlation coefficients r for time series data, and mean absolute error (MAE) and Bland-Altman plots with limits of agreement for peak measurements and smoothness. Results: Time series measurements showed strong to excellent correlations (r-values between 0.8 and 1.0), MAE for angles ranged from 1.5 to 5 degrees and for smoothness parameters (SPARC) from 0.02-0.09, while MAE for distance-related parameters ranged from 9 to 15 mm. Conclusion: Extracted motion parameters are valid and accurate for various movement tasks in children and young adults, demonstrating the suitability of our tracking method for clinical motion analysis. Clinical Impact: The low-cost portable hardware in combination with our tracking method enables motion analysis outside of specialized facilities while providing measurements that are close to those of the clinical gold-standard.
{"title":"Concurrent Validity of Motion Parameters Measured With an RGB-D Camera-Based Markerless 3D Motion Tracking Method in Children and Young Adults","authors":"Nikolas Hesse;Sandra Baumgartner;Anja Gut;Hubertus J. A. Van Hedel","doi":"10.1109/JTEHM.2024.3435334","DOIUrl":"10.1109/JTEHM.2024.3435334","url":null,"abstract":"Objective: Low-cost, portable RGB-D cameras with integrated motion tracking functionality enable easy-to-use 3D motion analysis without requiring expensive facilities and specialized personnel. However, the accuracy of existing systems is insufficient for most clinical applications, particularly when applied to children. In previous work, we developed an RGB-D camera-based motion tracking method and showed that it accurately captures body joint positions of children and young adults in 3D. In this study, the validity and accuracy of clinically relevant motion parameters that were computed from kinematics of our motion tracking method are evaluated in children and young adults. Methods: Twenty-three typically developing children and healthy young adults (5-29 years, 110–189 cm) performed five movement tasks while being recorded simultaneously with a marker-based Vicon system and an Azure Kinect RGB-D camera. Motion parameters were computed from the extracted kinematics of both methods: time series measurements, i.e., measurements over time, peak measurements, i.e., measurements at a single time instant, and movement smoothness. The agreement of these parameter values was evaluated using Pearson’s correlation coefficients r for time series data, and mean absolute error (MAE) and Bland-Altman plots with limits of agreement for peak measurements and smoothness. Results: Time series measurements showed strong to excellent correlations (r-values between 0.8 and 1.0), MAE for angles ranged from 1.5 to 5 degrees and for smoothness parameters (SPARC) from 0.02-0.09, while MAE for distance-related parameters ranged from 9 to 15 mm. Conclusion: Extracted motion parameters are valid and accurate for various movement tasks in children and young adults, demonstrating the suitability of our tracking method for clinical motion analysis. Clinical Impact: The low-cost portable hardware in combination with our tracking method enables motion analysis outside of specialized facilities while providing measurements that are close to those of the clinical gold-standard.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"580-588"},"PeriodicalIF":3.7,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10614255","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141873112","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: Attention-deficit/hyperactivity disorder (ADHD) is a childhood-onset neurodevelopmental disorder with a prevalence ranging from 6.1 to 9.4%. The main symptoms of ADHD are inattention, hyperactivity, impulsivity, and even destructive behaviors that may have a long-term negative influence on learning performance or social relationships. Early diagnosis and treatment provide the best chance of reducing and managing symptoms. Currently, ADHD diagnosis relies on behavioral observations and ratings by clinicians and parents. Medical diagnosis of ADHD was reported to be delayed because of a global shortage of well-trained clinicians, the heterogeneous nature of ADHD, and combined comorbidities. Therefore, alternative ways to increase the efficiency of early diagnosis are needed. Previous studies used behavioral and neurophysiological data to assess patients with ADHD, yielding an accuracy range from 56.6% to 92%. Several factors were shown to affect the detection rate, including methods and tasks used and the number of electroencephalogram (EEG) channels. Given that children with ADHD have difficulty sustaining attention, in this study, we tested whether data from multiple tasks with different difficulties and prolonged experiment times can probe the levels of brain resources engaged during task performance and increase ADHD detection. Specifically, we proposed a Deep Neural Network-based (DNN) fusion model of multiple tasks to enhance the detection of ADHD. Methods & Results: Forty-nine children with ADHD and thirty-two typically developing children were recruited. Analytic results show that the fusion of multi-task neurophysiological data can increase the separation rate to 89%, whereas a single data type can only achieve a best accuracy of 81%. Moreover, the use of multiple tasks helps distinguish between children with ADHD and typically developing children. Our results suggest that different neurophysiological models from multiple tasks can provide essential information to assist in ADHD screening. In conclusion, the proposed model offers a more efficient, and accurate alternative for early clinical diagnosis and management of ADHD. The application of artificial intelligence and multimodal neurophysiological data in clinical settings sets a precedent for digital health, paving the way for future advancements in the field.
{"title":"Fusion of Multi-Task Neurophysiological Data to Enhance the Detection of Attention- Deficit/Hyperactivity Disorder","authors":"Kai-Feng Zhang;Shih-Ching Yeh;Eric Hsiao-Kuang Wu;Xiu Xu;Ho-Jung Tsai;Chun-Chuan Chen","doi":"10.1109/JTEHM.2024.3435553","DOIUrl":"10.1109/JTEHM.2024.3435553","url":null,"abstract":"Objective: Attention-deficit/hyperactivity disorder (ADHD) is a childhood-onset neurodevelopmental disorder with a prevalence ranging from 6.1 to 9.4%. The main symptoms of ADHD are inattention, hyperactivity, impulsivity, and even destructive behaviors that may have a long-term negative influence on learning performance or social relationships. Early diagnosis and treatment provide the best chance of reducing and managing symptoms. Currently, ADHD diagnosis relies on behavioral observations and ratings by clinicians and parents. Medical diagnosis of ADHD was reported to be delayed because of a global shortage of well-trained clinicians, the heterogeneous nature of ADHD, and combined comorbidities. Therefore, alternative ways to increase the efficiency of early diagnosis are needed. Previous studies used behavioral and neurophysiological data to assess patients with ADHD, yielding an accuracy range from 56.6% to 92%. Several factors were shown to affect the detection rate, including methods and tasks used and the number of electroencephalogram (EEG) channels. Given that children with ADHD have difficulty sustaining attention, in this study, we tested whether data from multiple tasks with different difficulties and prolonged experiment times can probe the levels of brain resources engaged during task performance and increase ADHD detection. Specifically, we proposed a Deep Neural Network-based (DNN) fusion model of multiple tasks to enhance the detection of ADHD. Methods & Results: Forty-nine children with ADHD and thirty-two typically developing children were recruited. Analytic results show that the fusion of multi-task neurophysiological data can increase the separation rate to 89%, whereas a single data type can only achieve a best accuracy of 81%. Moreover, the use of multiple tasks helps distinguish between children with ADHD and typically developing children. Our results suggest that different neurophysiological models from multiple tasks can provide essential information to assist in ADHD screening. In conclusion, the proposed model offers a more efficient, and accurate alternative for early clinical diagnosis and management of ADHD. The application of artificial intelligence and multimodal neurophysiological data in clinical settings sets a precedent for digital health, paving the way for future advancements in the field.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"668-674"},"PeriodicalIF":3.7,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10614196","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141872990","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 : 2024-07-29DOI: 10.1109/JTEHM.2024.3433448
Dong Hyun Choi;Yoon Ha Joo;Ki Hong Kim;Jeong Ho Park;Hyunjin Joo;Hyoun-Joong Kong;Hyunju Lee;Kyoung Jun Song;Sungwan Kim
The objective of this study was to develop a sound recognition-based cardiopulmonary resuscitation (CPR) training system that is accessible, cost-effective, easy-to-maintain and provides accurate CPR feedback. Beep-CPR, a novel device with accordion squeakers that emit high-pitched sounds during compression, was developed. The sounds emitted by Beep-CPR were recorded using a smartphone, segmented into 2-second audio fragments, and then transformed into spectrograms. A total of 6,065 spectrograms were generated from approximately 40 minutes of audio data, which were then randomly split into training, validation, and test datasets. Each spectrogram was matched with the depth, rate, and release velocity of the compression measured at the same time interval by the ZOLL X Series monitor/defibrillator. Deep learning models utilizing spectrograms as input were trained using transfer learning based on EfficientNet to predict the depth (Depth model), rate (Rate model), and release velocity (Recoil model) of compressions. Results: The mean absolute error (MAE) for the Depth model was 0.30 cm (95% confidence interval [CI]: 0.27–0.33). The MAE of the Rate model was 3.6/min (95% CI: 3.2–3.9). For the Recoil model, the MAE was 2.3 cm/s (95% CI: 2.1–2.5). External validation of the models demonstrated acceptable performance across multiple conditions, including the utilization of a newly-manufactured device, a fatigued device, and evaluation in an environment with altered spatial dimensions. We have developed a novel sound recognition-based CPR training system, that accurately measures compression quality during training. Significance: Beep-CPR is a cost-effective and easy-to-maintain solution that can improve the efficacy of CPR training by facilitating decentralized at-home training with performance feedback.
本研究的目的是开发一种基于声音识别的心肺复苏（CPR）培训系统，该系统方便使用、经济实惠、易于维护并能提供准确的心肺复苏反馈。Beep-CPR 是一种新型装置，带有风琴式尖叫器，在按压过程中会发出高亢的声音。Beep-CPR 发出的声音由智能手机录制，分割成 2 秒钟的音频片段，然后转换成频谱图。从大约 40 分钟的音频数据中共生成了 6,065 张频谱图，然后随机分成训练数据集、验证数据集和测试数据集。每张频谱图都与 ZOLL X 系列监护仪/除颤器在相同时间间隔内测量到的压缩深度、速率和释放速度相匹配。以频谱图为输入的深度学习模型通过基于 EfficientNet 的迁移学习进行训练，以预测按压的深度（深度模型）、速率（速率模型）和释放速度（反冲模型）。结果：深度模型的平均绝对误差（MAE）为 0.30 厘米（95% 置信区间 [CI]：0.27-0.33）。速率模型的 MAE 为 3.6/分钟（95% 置信区间：3.2-3.9）。后坐力模型的 MAE 为 2.3 厘米/秒（95% CI：2.1-2.5）。模型的外部验证表明，在多种条件下，包括使用新制造的设备、疲劳设备以及在空间尺寸改变的环境中进行评估，其性能都是可以接受的。我们开发了一种新型的基于声音识别的心肺复苏训练系统，可在训练过程中准确测量按压质量。意义重大：Beep-CPR 是一种成本效益高且易于维护的解决方案，可通过提供性能反馈来促进分散的家庭培训，从而提高心肺复苏术培训的效果。
{"title":"A Development of a Sound Recognition-Based Cardiopulmonary Resuscitation Training System","authors":"Dong Hyun Choi;Yoon Ha Joo;Ki Hong Kim;Jeong Ho Park;Hyunjin Joo;Hyoun-Joong Kong;Hyunju Lee;Kyoung Jun Song;Sungwan Kim","doi":"10.1109/JTEHM.2024.3433448","DOIUrl":"10.1109/JTEHM.2024.3433448","url":null,"abstract":"The objective of this study was to develop a sound recognition-based cardiopulmonary resuscitation (CPR) training system that is accessible, cost-effective, easy-to-maintain and provides accurate CPR feedback. Beep-CPR, a novel device with accordion squeakers that emit high-pitched sounds during compression, was developed. The sounds emitted by Beep-CPR were recorded using a smartphone, segmented into 2-second audio fragments, and then transformed into spectrograms. A total of 6,065 spectrograms were generated from approximately 40 minutes of audio data, which were then randomly split into training, validation, and test datasets. Each spectrogram was matched with the depth, rate, and release velocity of the compression measured at the same time interval by the ZOLL X Series monitor/defibrillator. Deep learning models utilizing spectrograms as input were trained using transfer learning based on EfficientNet to predict the depth (Depth model), rate (Rate model), and release velocity (Recoil model) of compressions. Results: The mean absolute error (MAE) for the Depth model was 0.30 cm (95% confidence interval [CI]: 0.27–0.33). The MAE of the Rate model was 3.6/min (95% CI: 3.2–3.9). For the Recoil model, the MAE was 2.3 cm/s (95% CI: 2.1–2.5). External validation of the models demonstrated acceptable performance across multiple conditions, including the utilization of a newly-manufactured device, a fatigued device, and evaluation in an environment with altered spatial dimensions. We have developed a novel sound recognition-based CPR training system, that accurately measures compression quality during training. Significance: Beep-CPR is a cost-effective and easy-to-maintain solution that can improve the efficacy of CPR training by facilitating decentralized at-home training with performance feedback.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"550-557"},"PeriodicalIF":3.7,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10613881","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141868985","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 : 2024-07-29DOI: 10.1109/JTEHM.2024.3434460
Lei Guan;Xiaodong Yang;Nan Zhao;Malik Muhammad Arslan;Muneeb Ullah;Qurat Ul Ain;Abbas Ali Shah;Akram Alomainy;Qammer H. Abbasi
Vital signs are important indicators to evaluate the health status of patients. Channel state information (CSI) can sense the displacement of the chest wall caused by cardiorespiratory activity in a non-contact manner. Due to the influence of clutter, DC components, and respiratory harmonics, it is difficult to detect reliable heartbeat signals. To address this problem, this paper proposes a robust and novel method for simultaneously extracting breath and heartbeat signals using software defined radios (SDR). Specifically, we model and analyze the signal and propose singular value decomposition (SVD)-based clutter suppression method to enhance the vital sign signals. The DC is estimated and compensated by the circle fitting method. Then, the heartbeat signal and respiratory signal are obtained by the modified variational modal decomposition (VMD). The experimental results demonstrate that the proposed method can accurately separate the respiratory signal and the heartbeat signal from the filtered signal. The Bland-Altman analysis shows that the proposed system is in good agreement with the medical sensors. In addition, the proposed system can accurately measure the heart rate variability (HRV) within 0.5m. In summary, our system can be used as a preferred contactless alternative to traditional contact medical sensors, which can provide advanced patient-centered healthcare solutions.
{"title":"Non-Contact Measurement of Cardiopulmonary Activity Using Software Defined Radios","authors":"Lei Guan;Xiaodong Yang;Nan Zhao;Malik Muhammad Arslan;Muneeb Ullah;Qurat Ul Ain;Abbas Ali Shah;Akram Alomainy;Qammer H. Abbasi","doi":"10.1109/JTEHM.2024.3434460","DOIUrl":"10.1109/JTEHM.2024.3434460","url":null,"abstract":"Vital signs are important indicators to evaluate the health status of patients. Channel state information (CSI) can sense the displacement of the chest wall caused by cardiorespiratory activity in a non-contact manner. Due to the influence of clutter, DC components, and respiratory harmonics, it is difficult to detect reliable heartbeat signals. To address this problem, this paper proposes a robust and novel method for simultaneously extracting breath and heartbeat signals using software defined radios (SDR). Specifically, we model and analyze the signal and propose singular value decomposition (SVD)-based clutter suppression method to enhance the vital sign signals. The DC is estimated and compensated by the circle fitting method. Then, the heartbeat signal and respiratory signal are obtained by the modified variational modal decomposition (VMD). The experimental results demonstrate that the proposed method can accurately separate the respiratory signal and the heartbeat signal from the filtered signal. The Bland-Altman analysis shows that the proposed system is in good agreement with the medical sensors. In addition, the proposed system can accurately measure the heart rate variability (HRV) within 0.5m. In summary, our system can be used as a preferred contactless alternative to traditional contact medical sensors, which can provide advanced patient-centered healthcare solutions.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"558-568"},"PeriodicalIF":3.7,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10613608","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141873113","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}
Brain microstructural changes already occur in the earliest phases of Alzheimer’s disease (AD) as evidenced in diffusion magnetic resonance imaging (dMRI) literature. This study investigates the potential of the novel dMRI Apparent Measures Using Reduced Acquisitions (AMURA) as imaging markers for capturing such tissue modifications.Tract-based spatial statistics (TBSS) and support vector machines (SVMs) based on different measures were exploited to distinguish between amyloid-beta/tau negative (A $beta $ -/tau-) and A $beta $ +/tau+ or A $beta $ +/tau- subjects. Moreover, eXplainable Artificial Intelligence (XAI) was used to highlight the most influential features in the SVMs classifications and to validate the results by seeing the explanations’ recurrence across different methods.TBSS analysis revealed significant differences between A $beta $ -/tau- and other groups in line with the literature. The best SVM classification performance reached an accuracy of 0.73 by using advanced measures compared to more standard ones. Moreover, the explainability analysis suggested the results’ stability and the central role of the cingulum to show early sign of AD.By relying on SVM classification and XAI interpretation of the outcomes, AMURA indices can be considered viable markers for amyloid and tau pathology. Clinical impact: This pre-clinical research revealed AMURA indices as viable imaging markers for timely AD diagnosis by acquiring clinically feasible dMR images, with advantages compared to more invasive methods employed nowadays.
正如弥散磁共振成像（dMRI）文献所证实的那样，阿尔茨海默病（AD）的早期阶段已经出现大脑微结构变化。本研究探讨了新型 dMRI 表观测量（Apparent Measures Using Reduced Acquisitions，AMURA）作为成像标记捕捉此类组织变化的潜力。研究人员利用基于不同测量方法的肽段空间统计（Tract-based spatial statistics，TBSS）和支持向量机（Support vector machines，SVMs）来区分淀粉样蛋白-β/tau 阴性（A $beta $ -/tau-）和 A $beta $ +/tau+ 或 A $beta $ +/tau- 受试者。此外，eXplainable 人工智能（XAI）被用来突出 SVMs 分类中最有影响力的特征，并通过查看不同方法中解释的重复性来验证结果。与更标准的方法相比，使用高级方法的 SVM 分类准确率达到了 0.73。此外，可解释性分析表明了结果的稳定性以及蝶鞍在显示 AD 早期迹象方面的核心作用。通过依赖 SVM 分类和 XAI 结果解释，AMURA 指数可被视为淀粉样蛋白和 tau 病理学的可行标记。临床影响：这项临床前研究通过获取临床上可行的dMR图像，揭示了AMURA指数是及时诊断AD的可行成像标记物，与目前采用的更具侵入性的方法相比具有优势。
{"title":"XAI-Based Assessment of the AMURA Model for Detecting Amyloid-β and Tau Microstructural Signatures in Alzheimer’s Disease","authors":"Lorenza Brusini;Federica Cruciani;Gabriele Dall’Aglio;Tommaso Zajac;Ilaria Boscolo Galazzo;Mauro Zucchelli;Gloria Menegaz","doi":"10.1109/JTEHM.2024.3430035","DOIUrl":"10.1109/JTEHM.2024.3430035","url":null,"abstract":"Brain microstructural changes already occur in the earliest phases of Alzheimer’s disease (AD) as evidenced in diffusion magnetic resonance imaging (dMRI) literature. This study investigates the potential of the novel dMRI Apparent Measures Using Reduced Acquisitions (AMURA) as imaging markers for capturing such tissue modifications.Tract-based spatial statistics (TBSS) and support vector machines (SVMs) based on different measures were exploited to distinguish between amyloid-beta/tau negative (A\u0000<inline-formula> <tex-math>$beta $ </tex-math></inline-formula>\u0000-/tau-) and A\u0000<inline-formula> <tex-math>$beta $ </tex-math></inline-formula>\u0000+/tau+ or A\u0000<inline-formula> <tex-math>$beta $ </tex-math></inline-formula>\u0000+/tau- subjects. Moreover, eXplainable Artificial Intelligence (XAI) was used to highlight the most influential features in the SVMs classifications and to validate the results by seeing the explanations’ recurrence across different methods.TBSS analysis revealed significant differences between A\u0000<inline-formula> <tex-math>$beta $ </tex-math></inline-formula>\u0000-/tau- and other groups in line with the literature. The best SVM classification performance reached an accuracy of 0.73 by using advanced measures compared to more standard ones. Moreover, the explainability analysis suggested the results’ stability and the central role of the cingulum to show early sign of AD.By relying on SVM classification and XAI interpretation of the outcomes, AMURA indices can be considered viable markers for amyloid and tau pathology. Clinical impact: This pre-clinical research revealed AMURA indices as viable imaging markers for timely AD diagnosis by acquiring clinically feasible dMR images, with advantages compared to more invasive methods employed nowadays.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"569-579"},"PeriodicalIF":3.7,"publicationDate":"2024-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10601188","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141746410","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 : 2024-07-16DOI: 10.1109/JTEHM.2024.3429422
Hyungsoo Lim;Dong Ah Shin;Jaehoon Sim;Jaeheung Park;Taegyun Kim;Kyung Su Kim;Gil Joon Suh;Jung Chan Lee
Objective: This study introduces a novel system that can simulate diverse mechanical properties of the human chest to enhance the experience of CPR training by reflecting realistic chest conditions of patients. Methods: The proposed system consists of Variable stiffness mechanisms (VSMs) and Variable damper (VD) utilizing stretching silicone bands and dashpot dampers with controllable valves to modulate stiffness and damping, respectively. Cyclic loading was applied with a robot manipulator to the system. Compression force and displacement were measured and analyzed to evaluate the system’s mechanical response. Long-term stability of the system was also validated. Results: A non-linear response of the human chest under compression is realized through this design. Test results indicated non-linear force-displacement curves with hysteresis, similar to those observed in the chest of patients. Controlling the VSM and VD allowed for intentional changes in the slope and area of curves that are related to stiffness and damping, respectively. Stiffness and damping of the system were computed using performance test results. The stiffness ranged from 5.34 N/mm to 13.59 N/mm and the damping ranges from 0.127 N $cdot $ s/mm to 0.511 N $cdot $ s/mm. These properties cover a significant portion of the reported mechanical properties of the human chests. The system demonstrated satisfactory stability even when it was subjected to maximum stiffness conditions of the long-term compression test. Conclusion: The system is capable of emulating the mechanical properties and behavior of the human chests, thereby enhancing the CPR training experience.
目的：本研究介绍了一种新型系统，该系统可模拟人体胸部的各种机械特性，通过反映患者胸部的真实情况来增强心肺复苏训练的体验。方法：拟议的系统由可变刚度机构（VSM）和可变阻尼器（VD）组成，分别利用拉伸硅胶带和带可控阀门的仪表盘阻尼器来调节刚度和阻尼。使用机器人机械手对系统施加循环加载。对压缩力和位移进行测量和分析，以评估系统的机械响应。同时还验证了系统的长期稳定性。结果该设计实现了人体胸部在压缩下的非线性响应。测试结果表明，非线性力-位移曲线具有滞后性，与在患者胸部观察到的曲线相似。通过控制 VSM 和 VD，可以有意改变曲线的斜率和面积，这分别与刚度和阻尼有关。系统的刚度和阻尼是根据性能测试结果计算得出的。刚度范围为 5.34 N/mm 至 13.59 N/mm，阻尼范围为 0.127 N $cdot $ s/mm 至 0.511 N $cdot $ s/mm。这些特性涵盖了所报道的人体胸部机械特性的很大一部分。即使在长期压缩试验的最大刚度条件下，该系统也表现出令人满意的稳定性。结论该系统能够模拟人体胸腔的机械特性和行为，从而增强心肺复苏训练体验。
{"title":"Variable Stiffness and Damping Mechanism for CPR Manikin to Simulate Mechanical Properties of Human Chest","authors":"Hyungsoo Lim;Dong Ah Shin;Jaehoon Sim;Jaeheung Park;Taegyun Kim;Kyung Su Kim;Gil Joon Suh;Jung Chan Lee","doi":"10.1109/JTEHM.2024.3429422","DOIUrl":"10.1109/JTEHM.2024.3429422","url":null,"abstract":"Objective: This study introduces a novel system that can simulate diverse mechanical properties of the human chest to enhance the experience of CPR training by reflecting realistic chest conditions of patients. Methods: The proposed system consists of Variable stiffness mechanisms (VSMs) and Variable damper (VD) utilizing stretching silicone bands and dashpot dampers with controllable valves to modulate stiffness and damping, respectively. Cyclic loading was applied with a robot manipulator to the system. Compression force and displacement were measured and analyzed to evaluate the system’s mechanical response. Long-term stability of the system was also validated. Results: A non-linear response of the human chest under compression is realized through this design. Test results indicated non-linear force-displacement curves with hysteresis, similar to those observed in the chest of patients. Controlling the VSM and VD allowed for intentional changes in the slope and area of curves that are related to stiffness and damping, respectively. Stiffness and damping of the system were computed using performance test results. The stiffness ranged from 5.34 N/mm to 13.59 N/mm and the damping ranges from 0.127 N\u0000<inline-formula> <tex-math>$cdot $ </tex-math></inline-formula>\u0000 s/mm to 0.511 N\u0000<inline-formula> <tex-math>$cdot $ </tex-math></inline-formula>\u0000 s/mm. These properties cover a significant portion of the reported mechanical properties of the human chests. The system demonstrated satisfactory stability even when it was subjected to maximum stiffness conditions of the long-term compression test. Conclusion: The system is capable of emulating the mechanical properties and behavior of the human chests, thereby enhancing the CPR training experience.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"542-549"},"PeriodicalIF":3.7,"publicationDate":"2024-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10599511","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141718120","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}
The equivalent electrical circuit approach is explored to improve a bioimpedance-based transducer for measuring the bioavailability of synthetic insulin already presented in previous studies. In particular, the electrical parameter most sensitive to the variation of insulin amount injected was identified. Eggplants were used to emulate human electrical behavior under a quasi-static assumption guaranteed by a very low measurement time compared to the estimated insulin absorption time. Measurements were conducted with the EVAL-AD5940BIOZ by applying a sinusoidal voltage signal with an amplitude of 100 mV and acquiring impedance spectra in the range [1–100] kHz. 14 units of insulin were gradually administered using a Lilly’s Insulin Pen having a 0.4 cm long needle. Modified Hayden’s model was adopted as a reference circuit and the electrical component modeling the extracellular fluids was found to be the most insulin-sensitive parameter. The trnasducer achieves a state-of-the-art sensitivity of 225.90 ml1. An improvement of 223 % in sensitivity, 44 % in deterministic error, 7 % in nonlinearity, and 42 % in reproducibility was achieved compared to previous experimental studies. The clinical impact of the transducer was evaluated by projecting its impact on a Smart Insulin Pen for real-time measurement of insulin bioavailability. The wide gain in sensitivity of the bioimpedance-based transducer results in a significant reduction of the uncertainty of the Smart Insulin Pen. Considering the same improvement in in-vivo applications, the uncertainty of the Smart Insulin Pen is decreased from $4.2~mu $ l to $1.3~mu $ l.Clinical and Translational Impact Statement: A Smart Insulin Pen based on impedance spectroscopy and equivalent electrical circuit approach could be an effective solution for the non-invasive and real-time measurement of synthetic insulin uptake after subcutaneous administration.
{"title":"Equivalent Electrical Circuit Approach to Enhance a Transducer for Insulin Bioavailability Assessment","authors":"Francesca Mancino;Hanen Nouri;Nicola Moccaldi;Pasquale Arpaia;Olfa Kanoun","doi":"10.1109/JTEHM.2024.3425269","DOIUrl":"10.1109/JTEHM.2024.3425269","url":null,"abstract":"The equivalent electrical circuit approach is explored to improve a bioimpedance-based transducer for measuring the bioavailability of synthetic insulin already presented in previous studies. In particular, the electrical parameter most sensitive to the variation of insulin amount injected was identified. Eggplants were used to emulate human electrical behavior under a quasi-static assumption guaranteed by a very low measurement time compared to the estimated insulin absorption time. Measurements were conducted with the EVAL-AD5940BIOZ by applying a sinusoidal voltage signal with an amplitude of 100 mV and acquiring impedance spectra in the range [1–100] kHz. 14 units of insulin were gradually administered using a Lilly’s Insulin Pen having a 0.4 cm long needle. Modified Hayden’s model was adopted as a reference circuit and the electrical component modeling the extracellular fluids was found to be the most insulin-sensitive parameter. The trnasducer achieves a state-of-the-art sensitivity of 225.90 ml1. An improvement of 223 % in sensitivity, 44 % in deterministic error, 7 % in nonlinearity, and 42 % in reproducibility was achieved compared to previous experimental studies. The clinical impact of the transducer was evaluated by projecting its impact on a Smart Insulin Pen for real-time measurement of insulin bioavailability. The wide gain in sensitivity of the bioimpedance-based transducer results in a significant reduction of the uncertainty of the Smart Insulin Pen. Considering the same improvement in in-vivo applications, the uncertainty of the Smart Insulin Pen is decreased from \u0000<inline-formula> <tex-math>$4.2~mu $ </tex-math></inline-formula>\u0000l to \u0000<inline-formula> <tex-math>$1.3~mu $ </tex-math></inline-formula>\u0000l.Clinical and Translational Impact Statement: A Smart Insulin Pen based on impedance spectroscopy and equivalent electrical circuit approach could be an effective solution for the non-invasive and real-time measurement of synthetic insulin uptake after subcutaneous administration.","PeriodicalId":54255,"journal":{"name":"IEEE Journal of Translational Engineering in Health and Medicine-Jtehm","volume":"12 ","pages":"533-541"},"PeriodicalIF":3.7,"publicationDate":"2024-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10589471","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141570091","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：