Elena Khivintseva
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Elena Khivintseva has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
- Alexander Vladimirovich Zakharov
Samara Medical University, RUSSIA
14 shared publications - Igor V. Shirolapov
Samara State Medical University, Russia
12 shared publications - Mariya Sergeeva
Samara State Medical University, Russia
11 shared publications - Natalia Romanchuk
Samara State Medical University, Russia
8 shared publications - Yuliya Komarova
Samara State Medical University, Russia
6 shared publications - Vladimir А. Kalinin
Samara State Medical University, Russia
4 shared publications
Latest Publications
- 28 December 2026
- Published by IGI Global Scientific Publishing
- p. 443-474
Scalable Lightweight AI-Enabled EEG Wearables for Ambulatory Epilepsy Diagnosis and Polysomnography
This chapter provides a systematic analysis of hardware platforms for scalable, lightweight artificial intelligence enabled electroencephalographic (EEG) wearables intended for long term ambulatory epilepsy diagnosis and polysomnography. The clinical need for multi day recordings outside hospital settings is first established, highlighting the limitations of routine inpatient EEG and the high prevalence of unreported seizures. The fundamental trade offs between channel count, battery life, and information content are examined, with power consumption identified as the dominant design constraint. Dry and textile electrode technologies are evaluated in terms of electrode skin impedance, motion artifact suppression, and long term stability. A comparative analysis of existing wearable EEG systems is presented, followed by clinical benefits, regulatory pathways, and current limitations. Future trends such as neuromorphic processors, energy harvesting, and hybrid biocompatible interfaces are discussed.
- 28 December 2026
- Published by IGI Global Scientific Publishing
- p. 443-474
Scalable Lightweight AI-Enabled EEG Wearables for Ambulatory Epilepsy Diagnosis and Polysomnography
This chapter provides a systematic analysis of hardware platforms for scalable, lightweight artificial intelligence enabled electroencephalographic (EEG) wearables intended for long term ambulatory epilepsy diagnosis and polysomnography. The clinical need for multi day recordings outside hospital settings is first established, highlighting the limitations of routine inpatient EEG and the high prevalence of unreported seizures. The fundamental trade offs between channel count, battery life, and information content are examined, with power consumption identified as the dominant design constraint. Dry and textile electrode technologies are evaluated in terms of electrode skin impedance, motion artifact suppression, and long term stability. A comparative analysis of existing wearable EEG systems is presented, followed by clinical benefits, regulatory pathways, and current limitations. Future trends such as neuromorphic processors, energy harvesting, and hybrid biocompatible interfaces are discussed.
- 11 November 2026
- Published by Wiley in European Journal of Neuroscience
- Vol. 64 (9)
Adjunctive Multisensory Passive Immersive Virtual Reality With Synchronized Plantar Tactile Stimulation for Gait and Balance Disorders in Parkinson's Disease: An Exploratory Randomized Controlled Trial
Postural‐locomotor disorders, particularly freezing of gait and postural instability, are among the most disabling symptoms of advanced Parkinson's disease (PD) and respond only partially to dopaminergic therapy. Evidence for virtual reality (VR) approaches in moderate to severe PD remains limited. This prospective randomized controlled open‐label exploratory trial evaluated the addition of multisensory passive immersive VR with synchronized plantar tactile stimulation to conventional rehabilitation for gait, balance, and freezing disorders in moderate‐to‐severe PD (Hoehn and Yahr Stages 2–4). In the study, 80 PD patients were randomized into an experimental group (EXP, n = 40) receiving standard pharmacotherapy and conventional rehabilitation plus 10 sessions of passive immersive VR with synchronized plantar tactile stimulation and a comparison group (COM, n = 40) receiving standard pharmacotherapy and conventional rehabilitation only. Because the VR sessions were added to an otherwise identical program, the trial evaluates the effect of adding this intervention rather than a VR‐ or tactile‐specific effect. Outcomes were assessed at baseline and after 10 days using the Unified Parkinson's Disease Rating Scale (UPDRS), Berg Balance Scale (BBS), and Freezing of Gait Questionnaire (FoG‐Q). The EXP group showed significant improvements in the UPDRS “Gait” cluster (an eight‐item composite of UPDRS items 2.11, 2.12, 2.13, 3.9, 3.10, 3.11, 3.12, and 3.13, range 0–32; median from 12.5 to 8.0 points, p < 0.001), BBS (from 38.0 to 43.5 points, p < 0.001), and FoG‐Q (from 9.0 to 6.5 points, p < 0.001). A formal group × time analysis confirmed a differential treatment effect for every outcome. For the main outcome, the UPDRS “Gait” cluster, the baseline‐adjusted between‐group difference was −2.48 points (95% CI −3.37 to −1.59; p < 0.001; Hedges' g = 1.16). Large effects also favored the EXP group for the FoG‐Q (−2.14 points, 95% CI −2.62 to −1.66; g = 1.65), the BBS (+2.20 points, 95% CI 1.31 to 3.10; g = 1.06), the UPDRS freezing subscore (items 2.13 + 3.11; −0.89 points, 95% CI −1.13 to −0.66; g = 1.32), and the UPDRS postural subscore (items 3.12 + 3.13; −0.74 points, 95% CI −1.08 to −0.41; g = 0.93). Secondary outcome p ‐values were adjusted for multiple comparisons using the Holm and Benjamini–Hochberg procedures. By contrast, in the COM group, no significant change was observed for freezing of gait ( p = 0.083), and improvements in other outcomes were smaller than in the EXP group. No serious adverse events were recorded. The trial was exploratory, so the effect estimates require confirmation rather than supporting robustness in themselves. Overall, multisensory passive immersive VR with synchronized plantar tactile stimulation was well tolerated and shows preliminary promise as an adjunct to conventional rehabilitation for gait and balance disorders in moderate‐to‐severe PD. Wider clinical implementation requires long‐term multicenter studies with standardized assessment criteria and analysis of long‐term sustainability. These findings support integrating passive VR technologies into personalized neurorehabilitation programs.
- 11 November 2026
- Published by Wiley in European Journal of Neuroscience
- Vol. 64 (9)
Adjunctive Multisensory Passive Immersive Virtual Reality With Synchronized Plantar Tactile Stimulation for Gait and Balance Disorders in Parkinson's Disease: An Exploratory Randomized Controlled Trial
Postural‐locomotor disorders, particularly freezing of gait and postural instability, are among the most disabling symptoms of advanced Parkinson's disease (PD) and respond only partially to dopaminergic therapy. Evidence for virtual reality (VR) approaches in moderate to severe PD remains limited. This prospective randomized controlled open‐label exploratory trial evaluated the addition of multisensory passive immersive VR with synchronized plantar tactile stimulation to conventional rehabilitation for gait, balance, and freezing disorders in moderate‐to‐severe PD (Hoehn and Yahr Stages 2–4). In the study, 80 PD patients were randomized into an experimental group (EXP, n = 40) receiving standard pharmacotherapy and conventional rehabilitation plus 10 sessions of passive immersive VR with synchronized plantar tactile stimulation and a comparison group (COM, n = 40) receiving standard pharmacotherapy and conventional rehabilitation only. Because the VR sessions were added to an otherwise identical program, the trial evaluates the effect of adding this intervention rather than a VR‐ or tactile‐specific effect. Outcomes were assessed at baseline and after 10 days using the Unified Parkinson's Disease Rating Scale (UPDRS), Berg Balance Scale (BBS), and Freezing of Gait Questionnaire (FoG‐Q). The EXP group showed significant improvements in the UPDRS “Gait” cluster (an eight‐item composite of UPDRS items 2.11, 2.12, 2.13, 3.9, 3.10, 3.11, 3.12, and 3.13, range 0–32; median from 12.5 to 8.0 points, p < 0.001), BBS (from 38.0 to 43.5 points, p < 0.001), and FoG‐Q (from 9.0 to 6.5 points, p < 0.001). A formal group × time analysis confirmed a differential treatment effect for every outcome. For the main outcome, the UPDRS “Gait” cluster, the baseline‐adjusted between‐group difference was −2.48 points (95% CI −3.37 to −1.59; p < 0.001; Hedges' g = 1.16). Large effects also favored the EXP group for the FoG‐Q (−2.14 points, 95% CI −2.62 to −1.66; g = 1.65), the BBS (+2.20 points, 95% CI 1.31 to 3.10; g = 1.06), the UPDRS freezing subscore (items 2.13 + 3.11; −0.89 points, 95% CI −1.13 to −0.66; g = 1.32), and the UPDRS postural subscore (items 3.12 + 3.13; −0.74 points, 95% CI −1.08 to −0.41; g = 0.93). Secondary outcome p ‐values were adjusted for multiple comparisons using the Holm and Benjamini–Hochberg procedures. By contrast, in the COM group, no significant change was observed for freezing of gait ( p = 0.083), and improvements in other outcomes were smaller than in the EXP group. No serious adverse events were recorded. The trial was exploratory, so the effect estimates require confirmation rather than supporting robustness in themselves. Overall, multisensory passive immersive VR with synchronized plantar tactile stimulation was well tolerated and shows preliminary promise as an adjunct to conventional rehabilitation for gait and balance disorders in moderate‐to‐severe PD. Wider clinical implementation requires long‐term multicenter studies with standardized assessment criteria and analysis of long‐term sustainability. These findings support integrating passive VR technologies into personalized neurorehabilitation programs.
- 25 September 2026
- Published by Clinical Dentistry in Clinical Dentistry (Russia)
- No. 3,p. 210-216
Prospects for respiratory training with immersive technologies and biofeedback in patients of outpatient dental practice
This study evaluates the effectiveness of structured respiratory training using virtual reality (VR) and biofeedback for reducing psychophysiological tension in patients of outpatient dental practice. Thirty-four patients were enrolled in this pilot randomised study and allocated to a main group (training with VR and biofeedback, n=17) and a comparison group (n=17). In the main group, breathing awareness increased from 35.8±7.5 to 41.4±9.0 points (p=0.052 by the sign test and p=0.011 by the Wilcoxon signed-rank test) and relaxation self-efficacy increased from 4.8±3.6 to 7.0±3.5 points (p=0.043 by the sign test and p=0.006 by the Wilcoxon signed-rank test). Between-group comparison of changes showed a statistically significant advantage of VR-assisted training for the gain in breathing awareness (p=0.038 by the Mann–Whitney U-test); no between-group difference was found for relaxation self-efficacy (p=0.092). The findings suggest that a single session of respiratory training with immersive technologies and biofeedback promotes the development of breathing self-regulation skills. Given the sample size, the single-session design and the use of non-validated scales, these data are preliminary and require confirmation in larger studies assessing anxiety, autonomic measures and treatment adherence.
- 25 September 2026
- Published by Clinical Dentistry in Clinical Dentistry (Russia)
- No. 3,p. 210-216
Prospects for respiratory training with immersive technologies and biofeedback in patients of outpatient dental practice
This study evaluates the effectiveness of structured respiratory training using virtual reality (VR) and biofeedback for reducing psychophysiological tension in patients of outpatient dental practice. Thirty-four patients were enrolled in this pilot randomised study and allocated to a main group (training with VR and biofeedback, n=17) and a comparison group (n=17). In the main group, breathing awareness increased from 35.8±7.5 to 41.4±9.0 points (p=0.052 by the sign test and p=0.011 by the Wilcoxon signed-rank test) and relaxation self-efficacy increased from 4.8±3.6 to 7.0±3.5 points (p=0.043 by the sign test and p=0.006 by the Wilcoxon signed-rank test). Between-group comparison of changes showed a statistically significant advantage of VR-assisted training for the gain in breathing awareness (p=0.038 by the Mann–Whitney U-test); no between-group difference was found for relaxation self-efficacy (p=0.092). The findings suggest that a single session of respiratory training with immersive technologies and biofeedback promotes the development of breathing self-regulation skills. Given the sample size, the single-session design and the use of non-validated scales, these data are preliminary and require confirmation in larger studies assessing anxiety, autonomic measures and treatment adherence.
- 7 August 2026
- Published by IGI Global Scientific Publishing
- p. 119-148
Integrating Brain-Computer Interfaces and Artificial Intelligence for Deep Interpretation of Cognitive States in Decision Support
The chapter establishes Hybrid Intelligence (HI) as a synergistic paradigm, integrating human cognition with artificial intelligence via adaptive Brain-Computer Interfaces (BCI). It addresses the persistent “cognitive gap” in decision support by detailing the core components—human cognitive states and enabling technologies like EEG and fNIRS—and their integration into a closed-loop architecture. The discussion focuses on creating a “deep interpretation” engine for real-time cognitive state decoding, explores applications in critical domains, and examines the ethical and practical challenges of deploying such systems to augment human decision-making effectively.
- 7 August 2026
- Published by IGI Global Scientific Publishing
- p. 119-148
Integrating Brain-Computer Interfaces and Artificial Intelligence for Deep Interpretation of Cognitive States in Decision Support
The chapter establishes Hybrid Intelligence (HI) as a synergistic paradigm, integrating human cognition with artificial intelligence via adaptive Brain-Computer Interfaces (BCI). It addresses the persistent “cognitive gap” in decision support by detailing the core components—human cognitive states and enabling technologies like EEG and fNIRS—and their integration into a closed-loop architecture. The discussion focuses on creating a “deep interpretation” engine for real-time cognitive state decoding, explores applications in critical domains, and examines the ethical and practical challenges of deploying such systems to augment human decision-making effectively.
- 20 March 2026
- Published by IGI Global Scientific Publishing
- p. 207-238
Transcranial Magnetic Stimulation as a Novel Therapeutic Strategy for Parkinson's Disease in Under-Resourced Settings
Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized not only by motor deficits but also by cognitive impairment and network-level dysfunction. This chapter explores the therapeutic potential of Transcranial Magnetic Stimulation (TMS), a non-invasive neuromodulation technique. We posit that TMS can directly target maladaptive cortico-subcortical dynamics—such as pathological beta oscillations—and induce beneficial neuroplasticity, thereby addressing both motor and cognitive symptoms. The rationale for network-level intervention is supported by empirical data linking functional connectivity patterns, measured via EEG and source localization, to working memory performance. By reviewing physiological mechanisms, clinical protocols, and methodological integration, we argue that TMS represents a paradigm shift towards personalized, circuit-based remediation in PD, offering a promising adjunct to standard therapies for treatment-resistant symptoms.
- 20 March 2026
- Published by IGI Global Scientific Publishing
- p. 207-238
Transcranial Magnetic Stimulation as a Novel Therapeutic Strategy for Parkinson's Disease in Under-Resourced Settings
Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized not only by motor deficits but also by cognitive impairment and network-level dysfunction. This chapter explores the therapeutic potential of Transcranial Magnetic Stimulation (TMS), a non-invasive neuromodulation technique. We posit that TMS can directly target maladaptive cortico-subcortical dynamics—such as pathological beta oscillations—and induce beneficial neuroplasticity, thereby addressing both motor and cognitive symptoms. The rationale for network-level intervention is supported by empirical data linking functional connectivity patterns, measured via EEG and source localization, to working memory performance. By reviewing physiological mechanisms, clinical protocols, and methodological integration, we argue that TMS represents a paradigm shift towards personalized, circuit-based remediation in PD, offering a promising adjunct to standard therapies for treatment-resistant symptoms.
- Open Access
- 13 August 2025
- Published by National Medical Research Center For Rehabilitation And Balneology in Bulletin of Rehabilitation Medicine
- Vol. 24 (4) , 156-167
Cognitive rehabilitation in multiple sclerosis: effectiveness and potential of virtual reality technologies. A review
Introduction. Innovative technologies in the rehabilitation of multiple sclerosis are aimed at introducing into clinical practice advanced methods and techniques for deep and comprehensive assessment and effective treatment of disorders not only in the motor, but also in the cognitive area. The aim of this review is to summarize and analyze current data on the potential of using virtual reality technologies in cognitive rehabilitation in multiple sclerosis. Aim. To summarise and analyse current evidence on the potential of using virtual reality technologies in cognitive rehabilitation in multiple sclerosis. Materials and methods. The search was conducted in PubMed, Scopus, RSCI databases using search queries and keywords in Russian and English: multiple sclerosis, cognitive, virtual reality/VR, immersive technologies, neurorehabilitation from 2014 to 2024. Main content of the review. Currently, in order to prevent the progression of cognitive deficit, 3D computer simulation technologies are being introduced, which provide realistic user interaction using special electronic equipment, visual and sound effects and multisensory feedback. Virtual reality systems demonstrate effectiveness in the rehabilitation of many diseases, allowing for the correction of impaired functions, stimulation of residual abilities and promotion of global health improvement. The restorative potential of virtual reality (VR) is realized through the complex involvement of the senses and stimulation of neuroplasticity, which develops specific cognitive and behavioral aspects of patient functioning. Increased feedback obtained through exercises performed in a virtual environment allows for the development of awareness of the results of the movements performed and the quality of the movements themselves, which has a positive effect on cognitive activity and motor control. The advantage of using VR technologies, especially with immersive tools, is to create a positive, motivating learning experience for the patient, which requires individual control over several sensorimotor and cognitive domains. Conclusion. This review analyzes the current scientific information in this area of research, provides a detailed assessment of the current state and potential of VR technologies in cognitive rehabilitation in multiple sclerosis, and discusses the fundamental mechanisms underlying neurorehabilitation using immersive technologies.
- 13 August 2025
- Published by National Medical Research Center For Rehabilitation And Balneology in Bulletin of Rehabilitation Medicine
- Vol. 24 (4) , 156-167
Open AccessCognitive rehabilitation in multiple sclerosis: effectiveness and potential of virtual reality technologies. A review
Introduction. Innovative technologies in the rehabilitation of multiple sclerosis are aimed at introducing into clinical practice advanced methods and techniques for deep and comprehensive assessment and effective treatment of disorders not only in the motor, but also in the cognitive area. The aim of this review is to summarize and analyze current data on the potential of using virtual reality technologies in cognitive rehabilitation in multiple sclerosis. Aim. To summarise and analyse current evidence on the potential of using virtual reality technologies in cognitive rehabilitation in multiple sclerosis. Materials and methods. The search was conducted in PubMed, Scopus, RSCI databases using search queries and keywords in Russian and English: multiple sclerosis, cognitive, virtual reality/VR, immersive technologies, neurorehabilitation from 2014 to 2024. Main content of the review. Currently, in order to prevent the progression of cognitive deficit, 3D computer simulation technologies are being introduced, which provide realistic user interaction using special electronic equipment, visual and sound effects and multisensory feedback. Virtual reality systems demonstrate effectiveness in the rehabilitation of many diseases, allowing for the correction of impaired functions, stimulation of residual abilities and promotion of global health improvement. The restorative potential of virtual reality (VR) is realized through the complex involvement of the senses and stimulation of neuroplasticity, which develops specific cognitive and behavioral aspects of patient functioning. Increased feedback obtained through exercises performed in a virtual environment allows for the development of awareness of the results of the movements performed and the quality of the movements themselves, which has a positive effect on cognitive activity and motor control. The advantage of using VR technologies, especially with immersive tools, is to create a positive, motivating learning experience for the patient, which requires individual control over several sensorimotor and cognitive domains. Conclusion. This review analyzes the current scientific information in this area of research, provides a detailed assessment of the current state and potential of VR technologies in cognitive rehabilitation in multiple sclerosis, and discusses the fundamental mechanisms underlying neurorehabilitation using immersive technologies.
- 25 July 2025
- Published by IGI Global Scientific Publishing
- p. 153-184
Robotics and Neuroscience in Rehabilitation
This monograph explores the intersection of robotics and neuroscience in the field of motor neurorehabilitation, focusing on the development and application of brain-computer interface (BCI) systems. Emphasizing principles of neuroplasticity, it examines how technologies such as EEG-based BCI, functional electrical stimulation (FES), virtual reality (VR), and robotic exoskeletons can be integrated into closed-loop, feedback-rich rehabilitation platforms. Special attention is given to hybrid systems combining multiple sensory modalities to enhance user engagement and therapeutic outcomes. The monograph also discusses current challenges in clinical translation, including signal variability, training demands, and hardware limitations, while highlighting promising advances in AI-driven algorithms, wearable EEG solutions, and co-adaptive control strategies. Overall, it argues for a shift toward personalized, home-based rehabilitation technologies that empower patients with motor impairments to achieve greater autonomy and functional recovery.
- 25 July 2025
- Published by IGI Global Scientific Publishing
- p. 153-184
Robotics and Neuroscience in Rehabilitation
This monograph explores the intersection of robotics and neuroscience in the field of motor neurorehabilitation, focusing on the development and application of brain-computer interface (BCI) systems. Emphasizing principles of neuroplasticity, it examines how technologies such as EEG-based BCI, functional electrical stimulation (FES), virtual reality (VR), and robotic exoskeletons can be integrated into closed-loop, feedback-rich rehabilitation platforms. Special attention is given to hybrid systems combining multiple sensory modalities to enhance user engagement and therapeutic outcomes. The monograph also discusses current challenges in clinical translation, including signal variability, training demands, and hardware limitations, while highlighting promising advances in AI-driven algorithms, wearable EEG solutions, and co-adaptive control strategies. Overall, it argues for a shift toward personalized, home-based rehabilitation technologies that empower patients with motor impairments to achieve greater autonomy and functional recovery.
- 23 April 2025
- Published by IGI Global Scientific Publishing
- p. 93-122
AI-Assisted MRI
Artificial intelligence (AI) combined with magnetic resonance imaging (MRI) is transforming multiple sclerosis (MS) management. AI improves lesion detection, segmentation, and tracking, enhancing diagnostic precision and reducing variability. By integrating multimodal MRI data with biomarkers, AI-driven models predict disease progression and personalize treatment strategies, enabling early intervention. These models support therapy response prediction and optimize disease-modifying therapies. AI also automates longitudinal monitoring, assessing lesion changes, brain atrophy, and microstructural damage with high accuracy. Advanced techniques, such as radiomics and functional imaging, reveal subtle patterns of disease progression, providing valuable insights into neurodegeneration. These innovations streamline clinical workflows, improve patient outcomes, and advance precision medicine in MS. By enhancing diagnostic accuracy, treatment planning, and monitoring, AI-powered MRI represents a paradigm shift in neurodegenerative disease management
- 23 April 2025
- Published by IGI Global Scientific Publishing
- p. 93-122
AI-Assisted MRI
Artificial intelligence (AI) combined with magnetic resonance imaging (MRI) is transforming multiple sclerosis (MS) management. AI improves lesion detection, segmentation, and tracking, enhancing diagnostic precision and reducing variability. By integrating multimodal MRI data with biomarkers, AI-driven models predict disease progression and personalize treatment strategies, enabling early intervention. These models support therapy response prediction and optimize disease-modifying therapies. AI also automates longitudinal monitoring, assessing lesion changes, brain atrophy, and microstructural damage with high accuracy. Advanced techniques, such as radiomics and functional imaging, reveal subtle patterns of disease progression, providing valuable insights into neurodegeneration. These innovations streamline clinical workflows, improve patient outcomes, and advance precision medicine in MS. By enhancing diagnostic accuracy, treatment planning, and monitoring, AI-powered MRI represents a paradigm shift in neurodegenerative disease management
- Open Access
- 15 December 2024
- Published by FSBEI of Higher Education SamSMU of Ministry of Health of the Russian Federation in Science and Innovations in Medicine
- Vol. 9 (4) , 284-290
Automatic segmentation of demyelination lesions in multiple sclerosis
Aim – to evaluate the effectiveness of the YOLOv8 algorithm for automatic segmentation of demyelination lesions in various locations in patients with multiple sclerosis. Material and methods. The study included 120 patients with a clinically confirmed diagnosis of multiple sclerosis who underwent contrast-enhanced MRI. The MRI data from patients with different types of disease progression were analyzed. T1-weighted, T2-weighted, and FLAIR sequences were used for the analysis. The YOLOv8 algorithm was adapted for medical imaging and trained on manually annotated MRI scans. Model performance was evaluated using precision, recall, and F1-Score metrics. Results. The YOLOv8 model demonstrated high segmentation performance with a precision of 0.79, recall of 00.73, and F1-Score of 0.65. The model effectively identified demyelination lesions in various locations typical for multiple sclerosis. However, there remains a need to improve recall to minimize the missed lesions. Testing on independent data confirmed the stability of the results of the model. Conclusion. The YOLOv8 algorithm shows significant potential for automatic segmentation of demyelination lesions in multiple sclerosis patients. This method could be successfully implemented in clinical practice, enabling faster diagnosis and improved monitoring of disease progression. Further optimization of the model, through data augmentation techniques and hybrid architectures, may enhance both segmentation accuracy and recall.
- 15 December 2024
- Published by FSBEI of Higher Education SamSMU of Ministry of Health of the Russian Federation in Science and Innovations in Medicine
- Vol. 9 (4) , 284-290
Open AccessAutomatic segmentation of demyelination lesions in multiple sclerosis
Aim – to evaluate the effectiveness of the YOLOv8 algorithm for automatic segmentation of demyelination lesions in various locations in patients with multiple sclerosis. Material and methods. The study included 120 patients with a clinically confirmed diagnosis of multiple sclerosis who underwent contrast-enhanced MRI. The MRI data from patients with different types of disease progression were analyzed. T1-weighted, T2-weighted, and FLAIR sequences were used for the analysis. The YOLOv8 algorithm was adapted for medical imaging and trained on manually annotated MRI scans. Model performance was evaluated using precision, recall, and F1-Score metrics. Results. The YOLOv8 model demonstrated high segmentation performance with a precision of 0.79, recall of 00.73, and F1-Score of 0.65. The model effectively identified demyelination lesions in various locations typical for multiple sclerosis. However, there remains a need to improve recall to minimize the missed lesions. Testing on independent data confirmed the stability of the results of the model. Conclusion. The YOLOv8 algorithm shows significant potential for automatic segmentation of demyelination lesions in multiple sclerosis patients. This method could be successfully implemented in clinical practice, enabling faster diagnosis and improved monitoring of disease progression. Further optimization of the model, through data augmentation techniques and hybrid architectures, may enhance both segmentation accuracy and recall.
- 19 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 261-264
Natural Language Processing for Analyzing Neurological Status in Patients With Multiple Sclerosis
Multiple sclerosis is one of the most important diseases affecting the health care system in the modern era. Multiple sclerosis patients' diagnosis and management can be significantly improved with the use of instrumental diagnostic methods, such as magnetic resonance imaging. Based on the diagnostic value of clinical picture data and the unique content of neurological status, the task of learning how to analyze textual sources of a patient's clinical picture arises. To accomplish this task, a model for analyzing textual neurological status was developed, configured, and tested. In developing the model, suitable textual data were collected, exploratory analysis of these data was performed, possible implementations of the neurological status analysis model were analyzed, and the most appropriate option was selected. Currently, the neurological status model is two text data classifiers with prediction of neurological exam score and multiple sclerosis course activity, which have performed well in testing.
- 19 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 261-264
Natural Language Processing for Analyzing Neurological Status in Patients With Multiple Sclerosis
Multiple sclerosis is one of the most important diseases affecting the health care system in the modern era. Multiple sclerosis patients' diagnosis and management can be significantly improved with the use of instrumental diagnostic methods, such as magnetic resonance imaging. Based on the diagnostic value of clinical picture data and the unique content of neurological status, the task of learning how to analyze textual sources of a patient's clinical picture arises. To accomplish this task, a model for analyzing textual neurological status was developed, configured, and tested. In developing the model, suitable textual data were collected, exploratory analysis of these data was performed, possible implementations of the neurological status analysis model were analyzed, and the most appropriate option was selected. Currently, the neurological status model is two text data classifiers with prediction of neurological exam score and multiple sclerosis course activity, which have performed well in testing.