BioMedical Engineering OnLine最新文献

{"title":"Deep learning and electrocardiography: systematic review of current techniques in cardiovascular disease diagnosis and management.","authors":"Zhenyan Wu, Caixia Guo","doi":"10.1186/s12938-025-01349-w","DOIUrl":"10.1186/s12938-025-01349-w","url":null,"abstract":"<p><p>This paper reviews the recent advancements in the application of deep learning combined with electrocardiography (ECG) within the domain of cardiovascular diseases, systematically examining 198 high-quality publications. Through meticulous categorization and hierarchical segmentation, it provides an exhaustive depiction of the current landscape across various cardiovascular ailments. Our study aspires to furnish interested readers with a comprehensive guide, thereby igniting enthusiasm for further, in-depth exploration and research in this realm.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"23"},"PeriodicalIF":2.9,"publicationDate":"2025-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11847366/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143482096","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cross-evaluation of wearable data for use in Parkinson's disease research: a free-living observational study on Empatica E4, Fitbit Sense, and Oura.","authors":"Haakon Reithe, Brice Marty, Juan C Torrado, Elise Førsund, Bettina S Husebo, Ane Erdal, Simon U Kverneng, Erika Sheard, Charalampos Tzoulis, Monica Patrascu","doi":"10.1186/s12938-025-01353-0","DOIUrl":"10.1186/s12938-025-01353-0","url":null,"abstract":"<p><strong>Background: </strong>Established assessment scales used for Parkinson's disease (PD) have several limitations in tracking symptom progression and fluctuation. Both research and commercial-grade wearables show potential in improving these assessments. However, it is not known whether pervasive and affordable devices can deliver reliable data, suitable for designing open-source unobtrusive around-the-clock assessments. Our aim is to investigate the usefulness of the research-grade wristband Empatica E4, commercial-grade smartwatch Fitbit Sense, and the Oura ring, for PD research.</p><p><strong>Method: </strong>The study included participants with PD (N = 15) and neurologically healthy controls (N = 16). Data were collected using established assessment scales (Movement Disorders Society Unified Parkinson's Disease Rating Scale, Montreal Cognitive Assessment, REM Sleep Behavior Disorder Screening Questionnaire, Hoehn and Yahr Stage), self-reported diary (activities, symptoms, sleep, medication times), and 2-week digital data from the three devices collected simultaneously. The analyses comprised three steps: preparation (device characteristics assessment, data extraction and preprocessing), processing (data structuring and visualization, cross-correlation analysis, diary comparison, uptime calculation), and evaluation (usability, availability, statistical analyses).</p><p><strong>Results: </strong>We found large variation in data characteristics and unsatisfactory cross-correlation. Due to output incongruences, only heart rate and movement could be assessed across devices. Empatica E4 and Fitbit Sense outperformed Oura in reflecting self-reported activities. Results show a weak output correlation and significant differences. The uptime was good, but Oura did not record heart rate and movement concomitantly. We also found variation in terms of access to raw data, sampling rate and level of device-native processing, ease of use, retrieval of data, and design. We graded the system usability of Fitbit Sense as good, Empatica E4 as poor, with Oura in the middle.</p><p><strong>Conclusions: </strong>In this study we identified a set of characteristics necessary for PD research: ease of handling, cleaning, data retrieval, access to raw data, score calculation transparency, long battery life, sufficient storage, higher sampling frequencies, software and hardware reliability, transparency. The three analyzed devices are not interchangeable and, based on data features, none were deemed optimal for PD research, but they all have the potential to provide suitable specifications in future iterations.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"22"},"PeriodicalIF":2.9,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11846298/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143472118","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identification of anatomical locations: its relevance for vibrotactile perception of individuals with Parkinson's disease.","authors":"Ankita Raghuvanshi, Priya Pallavi, Rahul Chhatlani, Jayesh Parmar, Manish Rana, Sagar Betai, Uttama Lahiri","doi":"10.1186/s12938-024-01326-9","DOIUrl":"10.1186/s12938-024-01326-9","url":null,"abstract":"<p><strong>Background: </strong>Vibrotactile input is a useful sensory cue for individuals with Parkinson's Disease (PD) to overcome freezing of gait (FoG). For this input to serve as a cue, its accurate perception is required. This needs the input to be delivered at an anatomical location where it can be perceived. This is particularly true for individuals with PD whose tactile perception differs from that of healthy individuals. Literature indicates choice of various anatomical locations e.g., Finger, Wrist, Thigh, Shin, Calf, Ankle, Achilles Tendon, Heel and torso for the application of vibrotactile stimulation. Though studies have focused on the comparison of the vibrotactile perception (based on feedback) at various anatomical locations, yet these have involved only healthy individuals. However, such exploration remains as majorly untouched for individuals with PD.</p><p><strong>Methods: </strong>To bridge this gap, here we have conducted a study using our vibrotactile stimulation system while involving twenty-one individuals with PD to understand the choice of anatomical location with regard to vibrotactile perception. In addition, our study involved twenty-one age-matched healthy individuals to understand possible differences if any in vibrotactile perception between the two groups of participants.</p><p><strong>Results: </strong>Our results showed that for the healthy participants, both 'Wrist' and 'Thigh' were equally strong anatomical locations with regard to vibrotactile perception that were correctly identified 100% of the time closely followed by 'Finger' for which the correct identification was 98% of the time with correct identification for all these three locations being statistically (p < 0.05) higher than the other locations. In contrast, for individuals with PD, the 'Thigh' emerged as a strong candidate anatomical location with regard to vibrotactile perception even for those with severity of symptoms (based on clinical measure) that was correctly identified 96% of the time followed by 'Wrist' for which the correct identification was 92% of the time with the correct identification for only the 'Thigh' being statistically (p < 0.05) higher than all the other locations (except 'Wrist').</p><p><strong>Conclusion: </strong>This finding is clinically significant in deciding the right anatomical location to offer vibrotactile cues for it to be correctly perceived by one with PD, providing assistance to overcome FoG.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"21"},"PeriodicalIF":2.9,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11841177/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143466819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Calculation of recovery coefficients for partial volume effect correction in PET/CT imaging using a customized anthropomorphic body phantom.","authors":"Gunes Yavuz, Bilal Kovan, Turkay Toklu, Tevfik F Çermik, Cengizhan Öztürk","doi":"10.1186/s12938-025-01330-7","DOIUrl":"10.1186/s12938-025-01330-7","url":null,"abstract":"<p><p>Positron Emission Tomography/Computed Tomography (PET/CT) combines metabolic and anatomical information improving the precision and accuracy of oncological diagnostics. The standardized uptake value (SUV) measures tumor metabolism, yet its accuracy is influenced by the partial volume effect (PVE), impacting small lesion detection. This study aims to refine PVE corrections for small lesions using an in-house customized, special anthropomorphic phantom. Scans of this phantom which contained spheres of different sizes were performed across four hospitals at different PET/CT systems from various manufacturers (Siemens and Philips analog PET/CT systems, GE analog and digital PET/CT systems). The phantom contained six custom-designed cylinders with embedded spheres simulating sub-centimeter (0.3, 0.5, 0.9) and centimeter (1.3, 1.9, 2.8) lesions. Scans were performed separately for each sphere in the thorax, abdomen, and pelvis regions at all sites. Recovery Coefficients (RCs) were calculated to correct SUV values, demonstrating that RCs vary by sphere size and anatomical region but not change significantly among scanners. RCs are approaching unity for larger spheres, ensuring accurate SUV measurements. However, small spheres (< 0.5 cm) exhibited significant measurement challenges due to PVE. The anthropomorphic phantom proved effective in obtaining realistic SUV-corrected values, offering a promising tool for enhancing the accuracy and standardization of PET imaging in oncology. This study underscores the necessity for advanced imaging technologies and standardized RC application in clinical settings to improve the quantification of PET imaging, particularly in small lesion detection.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"20"},"PeriodicalIF":2.9,"publicationDate":"2025-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11830175/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143424941","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A review on diagnostic assessments of tracheal stenosis.","authors":"Mojdeh Monjezi, Milad Rismanian, Hamidreza Jamaati","doi":"10.1186/s12938-025-01351-2","DOIUrl":"10.1186/s12938-025-01351-2","url":null,"abstract":"<p><p>Tracheal stenosis (TS) is a pathological condition characterized by a reduction in the trachea diameter. It is a common complication after prolonged endotracheal intubation but may also arise from autoimmune or inflammatory processes. Clinicians can select the most appropriate treatment option based on individual patient conditions. Therefore, precise localization and evaluation of the stenosis are essential to ensure safe and effective treatment. This review summarizes current research on TS diagnosis and assessment, encompassing functional, imaging, and bronchoscopy methods. The characteristics, advantages, and disadvantages of each technique are discussed in relation to their application in the diagnosis and assessment of TS. Bronchoscopy is considered the cornerstone of TS diagnosis, and novel adjunct imaging modalities have emerged to enhance its accuracy. We explore advanced endomicroscopic methods, such as endobronchial ultrasound (EBUS), photoacoustic endoscopy (PAE), optical coherence tomography (OCT), and confocal laser endomicroscopy (CLE). Among these, EBUS is clinically approved for diagnosing lesions with high resolution and acceptable penetration depth. OCT and CLE offer real-time imaging for peripheral lesions and potentially malignant nodules, but their use is limited by cost and availability in low-resource settings. Therefore, bronchoscopy, with biopsy techniques as needed, remains the optimal approach for diagnosing tracheal stenosis.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"18"},"PeriodicalIF":2.9,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11827378/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143424940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Preclinical evaluation of a hydraulic actuation system with guide tube for robotic cochlear implant electrode insertion.","authors":"Jakob Cramer, Rolf Salcher, Max Fröhlich, Georg Böttcher-Rebmann, Eralp Artukarslan, Thomas Lenarz, Thomas S Rau","doi":"10.1186/s12938-025-01338-z","DOIUrl":"10.1186/s12938-025-01338-z","url":null,"abstract":"<p><strong>Background: </strong>Automated insertion of the cochlear implant electrode array can reduce the risk of intracochlear trauma. To address this, our group previously developed a hydraulic electrode insertion device, the Cochlea Hydrodrive (CHD), which automates the process using a syringe piston driven by an infusion pump. This study aims to characterize the hydraulic actuation process of the CHD and to preclinically evaluate its design.</p><p><strong>Methods: </strong>A camera-based motion tracking test setup was developed to obtain hydraulic motion profiles. Various syringes were evaluated for their actuation properties and the optimal syringe was selected. The CHD design was adapted based on the selected syringe, incorporating a slotted stainless steel guide tube to surround the electrode during insertion. This enhanced design was tested in ex vivo insertion trials into human head specimens.</p><p><strong>Results: </strong>The final design of the CHD demonstrated smooth and steady motion profiles at all tested velocities (0.4 mm/s, 0.1 mm/s, 0.03 mm/s). Ex vivo insertion trials confirmed these findings, with the guide tube facilitating easy alignment of the CHD in front of the round window and preventing electrode buckling.</p><p><strong>Conclusion: </strong>Our study validates that the CHD provides reliably smooth actuation properties despite its low complexity. The use of a guide tube appears promising and could further enhance the standardization of automated electrode insertion.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"19"},"PeriodicalIF":2.9,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11829445/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143424942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparative analysis of electrical signals in facial expression muscles.","authors":"Luna Adamov, Bojan Petrović, Lazar Milić, Vojin Štrbac, Sanja Kojić, Karunan Joseph, Goran M Stojanović","doi":"10.1186/s12938-025-01350-3","DOIUrl":"10.1186/s12938-025-01350-3","url":null,"abstract":"<p><strong>Background: </strong>Facial expression muscles serve a fundamental role in the orofacial system, significantly influencing the overall health and well-being of an individual. They are essential for performing basic functions such as speech, chewing, and swallowing. The purpose of this study was to determine whether surface electromyography could be used to evaluate the health, function, or dysfunction of three facial muscles by measuring their electrical activity in healthy people. Additionally, to ascertain whether pattern recognition and artificial intelligence may be used for tasks that differ from one another.</p><p><strong>Results: </strong>The study included 24 participants and examined three muscles (m. Orbicularis Oris, m. Zygomaticus Major, and m. Mentalis) during five different facial expressions. Prior to thorough statistical analysis, features were extracted from the acquired electromyographs. Finally, classification was done with the use of logistic regression, random forest classifier and linear discriminant analysis. A statistically significant difference in muscle activity amplitudes was demonstrated between muscles, enabling the tracking of individual muscle activity for diagnostic and therapeutic purposes. Additionally other time domain and frequency domain features were analyzed, showing statistical significance in differentiation between muscles as well. Examples of pattern recognition showed promising avenues for further research and development.</p><p><strong>Conclusion: </strong>Surface electromyography is a useful method for assessing the function of facial expression muscles, significantly contributing to the diagnosis and treatment of oral motor function disorders. Results of this study show potential for further research and development in this field of research.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"17"},"PeriodicalIF":2.9,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11816783/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143405363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diffused Multi-scale Generative Adversarial Network for low-dose PET images reconstruction.","authors":"Xiang Yu, Daoyan Hu, Qiong Yao, Yu Fu, Yan Zhong, Jing Wang, Mei Tian, Hong Zhang","doi":"10.1186/s12938-025-01348-x","DOIUrl":"10.1186/s12938-025-01348-x","url":null,"abstract":"<p><strong>Purpose: </strong>The aim of this study is to convert low-dose PET (L-PET) images to full-dose PET (F-PET) images based on our Diffused Multi-scale Generative Adversarial Network (DMGAN) to offer a potential balance between reducing radiation exposure and maintaining diagnostic performance.</p><p><strong>Methods: </strong>The proposed method includes two modules: the diffusion generator and the u-net discriminator. The goal of the first module is to get different information from different levels, enhancing the generalization ability of the generator to the image and improving the stability of the training. Generated images are inputted into the u-net discriminator, extracting details from both overall and specific perspectives to enhance the quality of the generated F-PET images. We conducted evaluations encompassing both qualitative assessments and quantitative measures. In terms of quantitative comparisons, we employed two metrics, structure similarity index measure (SSIM) and peak signal-to-noise ratio (PSNR) to evaluate the performance of diverse methods.</p><p><strong>Results: </strong>Our proposed method achieved the highest PSNR and SSIM scores among the compared methods, which improved PSNR by at least 6.2% compared to the other methods. Compared to other methods, the synthesized full-dose PET image generated by our method exhibits a more accurate voxel-wise metabolic intensity distribution, resulting in a clearer depiction of the epilepsy focus.</p><p><strong>Conclusions: </strong>The proposed method demonstrates improved restoration of original details from low-dose PET images compared to other models trained on the same datasets. This method offers a potential balance between minimizing radiation exposure and preserving diagnostic performance.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"16"},"PeriodicalIF":2.9,"publicationDate":"2025-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11807330/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143381658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The impact of a 5G-based smart nursing information system and associated mobile hardware on clinical nurses' work stress: a randomized controlled study in a Chinese hospital.","authors":"Xuejiao Ruan, Yuying Lou, Xinhua Zhang, Zhulin Wu, Hongzhi Yuan","doi":"10.1186/s12938-025-01344-1","DOIUrl":"10.1186/s12938-025-01344-1","url":null,"abstract":"<p><strong>Background: </strong>Clinical nurses frequently endure substantial work-related stress, adversely affecting their well-being and potentially compromising patient care quality and safety. The integration of a 5G-based medical private network into smart nursing systems and mobile devices offers a promising solution to reduce this stress. This study evaluates the impact of a Smart Nursing Information System based on a 5G Medical Private Network and its Supporting Mobile Hardware (SNIS-SMH) on mitigating work-related stress among clinical nurses. The goal is to provide a scientific basis for nursing management, reduce error incidents, advance nursing procedures, and enhance productivity.</p><p><strong>Results: </strong>A total of 226 nurses completed the study. The SNIS-SMH group showed significantly lower total work stress scores (66.16 ± 9.82) compared to the control group (70.65 ± 11.32, P = 0.002). In specific dimensions, the SNIS-SMH group had lower scores for nursing profession and work (14.17 ± 2.37 vs. 15.00 ± 3.06, P = 0.023), workload and time distribution (10.56 ± 2.45 vs. 12.42 ± 2.55, P < 0.001), and patient care (22.55 ± 3.34 vs. 23.70 ± 4.06, P = 0.021). No significant differences were found in the work environment and resource, and management and interpersonal relationships dimensions.</p><p><strong>Conclusions: </strong>The SNIS-SMH system significantly alleviated work-related stress among clinical nurses, particularly in nursing duties, workload and time distribution, and patient care.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"15"},"PeriodicalIF":2.9,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11806821/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143373613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The effect of depth data and upper limb impairment on lightweight monocular RGB human pose estimation models.","authors":"Gloria-Edith Boudreault-Morales, Cesar Marquez-Chin, Xilin Liu, José Zariffa","doi":"10.1186/s12938-025-01347-y","DOIUrl":"10.1186/s12938-025-01347-y","url":null,"abstract":"<p><strong>Background and objectives: </strong>Markerless vision-based human pose estimation (HPE) is a promising avenue towards scalable data collection in rehabilitation. Deploying this technology will require self-contained systems able to process data efficiently and accurately. The aims of this work are to (1) Determine how depth data affects lightweight monocular red-green-blue (RGB) HPE performance (accuracy and speed), to inform sensor selection and (2) Validate HPE models using data from individuals with physical impairments.</p><p><strong>Methods: </strong>Two HPE models were investigated: Dite-HRNet and MobileHumanPose (capable of 2D and 3D HPE, respectively). The models were modified to include depth data as an input using three different fusion techniques: an early fusion method, a simple intermediate fusion method (using concatenation), and a complex intermediate fusion method (using specific fusion blocks, additional convolutional layers, and concatenation). All fusion techniques used RGB-D data, in contrast to the original models which only used RGB data. The models were trained, validated and tested using the CMU Panoptic and Human3.6 M data sets as well as a custom data set. The custom data set includes RGB-D and optical motion capture data of 15 uninjured and 12 post-stroke individuals, while they performed movements involving their upper limbs. HPE model performances were monitored through accuracy and computational efficiency. Evaluation metrics include Mean per Joint Position Error (MPJPE), Floating Point Operations (FLOPs) and frame rates (frames per second).</p><p><strong>Results: </strong>The early fusion architecture consistently delivered the lowest MPJPE in both 2D and 3D HPE cases while achieving similar FLOPs and frame rates to its RGB counterpart. These results were consistent regardless of the data used for training and testing the HPE models. Comparisons between the uninjured and stroke groups did not reveal a significant effect (all p values > 0.36) of motor impairment on the accuracy of any model.</p><p><strong>Conclusions: </strong>Including depth data using an early fusion architecture improves the accuracy-efficiency trade-off of the HPE model. HPE accuracy is not affected by the presence of physical impairments. These results suggest that using depth data with RGB data is beneficial to HPE, and that models trained with data collected from uninjured individuals can generalize to persons with physical impairments.</p>","PeriodicalId":8927,"journal":{"name":"BioMedical Engineering OnLine","volume":"24 1","pages":"12"},"PeriodicalIF":2.9,"publicationDate":"2025-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11804014/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143370373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}