Scholarlink

<p>Stiff-Person Syndrome (SPS) is a rare neurological disorder characterized by fluctuating muscle stiffness, spasms, and severe impairment of motor functions. Accurate and early detection of SPS is crucial for effective treatment and management. In this study, we propose a Hybrid Deep Learning Optimization (HDLO) Model for the precise detection and diagnosis of SPS, which integrates Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks for feature extraction and temporal sequence handling. The model is further optimized using a Genetic Algorithm (GA) to ensure the most efficient configuration of hyperparameters, enhancing its overall performance. The dataset used in this study consists of clinical data from 2376 patients collected over two years, including demographic information, symptom onset dates, severity levels, treatment types, responses to treatment, follow-up visits, and patient outcomes. Our proposed HDLO model achieves an accuracy of 97.62%, significantly outperforming traditional models such as CNN, LSTM, Random Forest (RF), and Support Vector Machines (SVM), which ranged from 86% to 95%. In addition to accuracy, the HDLO model demonstrated high precision (96.30%), recall (95.75%), and F1-score (96.00%), indicating its robustness in classifying SPS cases. The ensemble learning technique and attention mechanism further improved the model’s ability to focus on critical features, reducing false positives and negatives. This study provides a reliable tool for the early detection of SPS, offering potential improvements in clinical decision-making and patient care. The HDLO model’s ability to analyze both spatial and temporal data contributes to its high classification performance, making it a valuable asset in healthcare.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hybrid Deep Learning Optimization Model for Precise Detection and Diagnosis of Stiff-Person Syndrome

  • 期刊论文

  • 发表日期：

    2026年7月17日

查看全文
  • G. Saravanan,
  • Jeevanantham Vellaichamy

摘要

Stiff-Person Syndrome (SPS) is a rare neurological disorder characterized by fluctuating muscle stiffness, spasms, and severe impairment of motor functions. Accurate and early detection of SPS is crucial for effective treatment and management. In this study, we propose a Hybrid Deep Learning Optimization (HDLO) Model for the precise detection and diagnosis of SPS, which integrates Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks for feature extraction and temporal sequence handling. The model is further optimized using a Genetic Algorithm (GA) to ensure the most efficient configuration of hyperparameters, enhancing its overall performance. The dataset used in this study consists of clinical data from 2376 patients collected over two years, including demographic information, symptom onset dates, severity levels, treatment types, responses to treatment, follow-up visits, and patient outcomes. Our proposed HDLO model achieves an accuracy of 97.62%, significantly outperforming traditional models such as CNN, LSTM, Random Forest (RF), and Support Vector Machines (SVM), which ranged from 86% to 95%. In addition to accuracy, the HDLO model demonstrated high precision (96.30%), recall (95.75%), and F1-score (96.00%), indicating its robustness in classifying SPS cases. The ensemble learning technique and attention mechanism further improved the model’s ability to focus on critical features, reducing false positives and negatives. This study provides a reliable tool for the early detection of SPS, offering potential improvements in clinical decision-making and patient care. The HDLO model’s ability to analyze both spatial and temporal data contributes to its high classification performance, making it a valuable asset in healthcare.