Predicting stroke severity of patients using interpretable machine learning algorithms DOI Creative Commons
Amir Sorayaie Azar,

Tahereh Samimi,

Ghanbar Tavassoli

et al.

European journal of medical research, Journal Year: 2024, Volume and Issue: 29(1)

Published: Nov. 14, 2024

Stroke is a significant global health concern, ranking as the second leading cause of death and placing substantial financial burden on healthcare systems, particularly in low- middle-income countries. Timely evaluation stroke severity crucial for predicting clinical outcomes, with standard assessment tools being Rapid Arterial Occlusion Evaluation (RACE) National Institutes Health Scale (NIHSS). This study aims to utilize Machine Learning (ML) algorithms predict using these two distinct scales. We conducted this datasets collected from hospitals Urmia, Iran, corresponding assessments based RACE NIHSS. Seven ML were applied, including K-Nearest Neighbor (KNN), Decision Tree (DT), Random Forest (RF), Adaptive Boosting (AdaBoost), Extreme Gradient (XGBoost), Support Vector (SVM), Artificial Neural Network (ANN). Hyperparameter tuning was performed grid search optimize model performance, SHapley Additive Explanations (SHAP) used interpret contribution individual features. Among models, RF achieved highest accuracies 92.68% dataset 91.19% NIHSS dataset. The Area Under Curve (AUC) 92.02% 97.86% datasets, respectively. SHAP analysis identified triglyceride levels, length hospital stay, age critical predictors severity. first apply models scales use enhances interpretability increasing clinicians' trust algorithms. best-performing can be valuable tool assisting medical professionals settings.

Language: Английский

Electroencephalogram (EEG) Based Fuzzy Logic and Spiking Neural Networks (FLSNN) for Advanced Multiple Neurological Disorder Diagnosis DOI
Shraddha Jain, Ruchi Srivastava

Brain Topography, Journal Year: 2025, Volume and Issue: 38(3)

Published: Feb. 24, 2025

Language: Английский

Citations

2

Smart Textile Technology for the Monitoring of Mental Health DOI Creative Commons

Shonal Fernandes,

Alberto Ramos,

Mario Vega-Barbas

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(4), P. 1148 - 1148

Published: Feb. 13, 2025

In recent years, smart devices have proven their effectiveness in monitoring mental health issues and played a crucial role providing therapy. The ability to embed sensors fabrics opens new horizons for healthcare, addressing the growing demand innovative solutions objective of this review is understand health, its impact on human body, latest advancements field textiles (sensors, electrodes, garments) physiological signals such as respiration rate (RR), electroencephalogram (EEG), electrodermal activity (EDA), electrocardiogram (ECG), cortisol, all which are associated with disorders. Databases Web Science (WoS) Scopus were used identify studies that utilized monitor specific parameters. Research indicates provide promising results compared traditional methods, offering enhanced comfort long-term monitoring.

Language: Английский

Citations

0

Multiclass classification of epileptic seizure phases using a novel HFO-based feature extraction model DOI

Pelin Sari Tekten,

Soner Kotan,

Fırat Kaçar

et al.

Signal Image and Video Processing, Journal Year: 2025, Volume and Issue: 19(4)

Published: Feb. 22, 2025

Language: Английский

Citations

0

Exploiting adaptive neuro-fuzzy inference systems for cognitive patterns in multimodal brain signal analysis DOI Creative Commons

T. Thamaraimanalan,

Dhanalakshmi Gopal,

S. Vignesh

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 16, 2025

The analysis of cognitive patterns through brain signals offers critical insights into human cognition, including perception, attention, memory, and decision-making. However, accurately classifying these remains a challenge due to their inherent complexity non-linearity. This study introduces novel method, PCA-ANFIS, which integrates Principal Component Analysis (PCA) Adaptive Neuro-Fuzzy Inference Systems (ANFIS), enhance pattern recognition in multimodal signal analysis. PCA reduces the dimensionality EEG data while retaining salient features, enabling computational efficiency. ANFIS combines adaptability neural networks with interpretability fuzzy logic, making it well-suited model non-linear relationships within signals. Performance metrics our proposed such as accuracy, sensitivity, These additions highlight effectiveness method provide concise summary findings. achieves superior classification performance, an unprecedented accuracy 99.5%, significantly outperforming existing approaches. Comprehensive experiments were conducted using diverse dataset, demonstrating method's robustness sensitivity. integration addresses key challenges analysis, artifact contamination non-stationarity, ensuring reliable feature extraction classification. research has significant implications for both neuroscience clinical practice. By advancing understanding processes, PCA-ANFIS facilitates accurate diagnosis treatment disorders neurological conditions. Future work will focus on testing approach larger more datasets exploring its applicability domains neurofeedback, neuromarketing, brain-computer interfaces. establishes capable tool precise efficient processing.

Language: Английский

Citations

0

Multichannel convolutional transformer for detecting mental disorders using electroancephalogrpahy records DOI Creative Commons
Mamadou Dia,

Ghazaleh Khodabandelou,

Syed Muhammad Anwar

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: May 2, 2025

Mental disorders represent a critical global health challenge that affects millions around the world and significantly disrupts daily life. Early accurate detection is paramount for timely intervention, which can lead to improved treatment outcomes. Electroencephalography (EEG) provides non-invasive means observing brain activity, making it useful tool detecting potential mental disorders. Recently, deep learning techniques have gained prominence their ability analyze complex datasets, such as electroencephalography recordings. In this study, we introduce novel deep-learning architecture classification of post-traumatic stress disorder, depression, or anxiety, using data. Our proposed model, multichannel convolutional transformer, integrates strengths both neural networks transformers. Before feeding model low-level features, input pre-processed common spatial pattern filter, signal space projection wavelet denoising filter. Then EEG signals are transformed continuous transform obtain time-frequency representation. The layers tokenize by our pre-processing pipeline, while Transformer encoder effectively captures long-range temporal dependencies across sequences. This specifically tailored process data has been preprocessed transform, technique representation, thereby enhancing extraction relevant features classification. We evaluated performance on three datasets: Psychiatric Dataset, MODMA dataset, Psychological Assessment dataset. achieved accuracies 87.40% 89.84% 92.28% approach outperforms every concurrent approaches datasets used, without showing any sign over-fitting. These results underscore in delivering reliable disorder through analysis, paving way advancements early diagnosis strategies.

Language: Английский

Citations

0

Predicting stroke severity of patients using interpretable machine learning algorithms DOI Creative Commons
Amir Sorayaie Azar,

Tahereh Samimi,

Ghanbar Tavassoli

et al.

European journal of medical research, Journal Year: 2024, Volume and Issue: 29(1)

Published: Nov. 14, 2024

Stroke is a significant global health concern, ranking as the second leading cause of death and placing substantial financial burden on healthcare systems, particularly in low- middle-income countries. Timely evaluation stroke severity crucial for predicting clinical outcomes, with standard assessment tools being Rapid Arterial Occlusion Evaluation (RACE) National Institutes Health Scale (NIHSS). This study aims to utilize Machine Learning (ML) algorithms predict using these two distinct scales. We conducted this datasets collected from hospitals Urmia, Iran, corresponding assessments based RACE NIHSS. Seven ML were applied, including K-Nearest Neighbor (KNN), Decision Tree (DT), Random Forest (RF), Adaptive Boosting (AdaBoost), Extreme Gradient (XGBoost), Support Vector (SVM), Artificial Neural Network (ANN). Hyperparameter tuning was performed grid search optimize model performance, SHapley Additive Explanations (SHAP) used interpret contribution individual features. Among models, RF achieved highest accuracies 92.68% dataset 91.19% NIHSS dataset. The Area Under Curve (AUC) 92.02% 97.86% datasets, respectively. SHAP analysis identified triglyceride levels, length hospital stay, age critical predictors severity. first apply models scales use enhances interpretability increasing clinicians' trust algorithms. best-performing can be valuable tool assisting medical professionals settings.

Language: Английский

Citations

0