Scholarlink

<p>Parkinson’s disease is a neurodegenerative condition that affects people worldwide. In general, neurologists monitor the changes in handwritten drawings and voice speech signals to detect Parkinson’s disease. The common techniques utilized for the detection of Parkinson’s disease are limited and expensive. A novel Fuzzy Support Vector-Rider optimization algorithm-based Neural Network (FSV-RideNN) approach is employed for the detection of Parkinson’s disease by utilizing voice signals and spiral images. The input hand-drawing spiral image is initially pre-processed by utilizing a Gaussian filter, and the resultant pre-processed images are augmented. Subsequently, features are extracted from the augmented image. Similarly, the pre-processing of input voice signals is executed by utilizing a Non-Local Mean (NLM) filter. Later, the signal-based features are extracted from the pre-processed signal. Thereafter, the proposed method detects Parkinson’s disease using FSV-RideNN from the extracted signal-based and image-based features. Furthermore, the detection performance of FSV-RideNN is assessed by correlating it with the detection performance of baseline approaches, and FSV-RideNN attained superior performance with a Negative Predictive Value (NPV) of 92.05%, accuracy of 88.48%, Positive Predictive Value (PPV) of 91.56%, True Negative Rate (TNR) of 90.57%, and True Positive Rate (TPR) of 89.68% respectively.</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

Fuzzy Based Support Vector RideNN for Parkinson’s Disease Detection Using Voice Signal and Hand Drawing Spiral Image

  • 期刊论文

  • 发表日期：

    2026年1月7日

查看全文
  • Priya Dilip Ghate,
  • Anuradha C. Phadke

摘要

Parkinson’s disease is a neurodegenerative condition that affects people worldwide. In general, neurologists monitor the changes in handwritten drawings and voice speech signals to detect Parkinson’s disease. The common techniques utilized for the detection of Parkinson’s disease are limited and expensive. A novel Fuzzy Support Vector-Rider optimization algorithm-based Neural Network (FSV-RideNN) approach is employed for the detection of Parkinson’s disease by utilizing voice signals and spiral images. The input hand-drawing spiral image is initially pre-processed by utilizing a Gaussian filter, and the resultant pre-processed images are augmented. Subsequently, features are extracted from the augmented image. Similarly, the pre-processing of input voice signals is executed by utilizing a Non-Local Mean (NLM) filter. Later, the signal-based features are extracted from the pre-processed signal. Thereafter, the proposed method detects Parkinson’s disease using FSV-RideNN from the extracted signal-based and image-based features. Furthermore, the detection performance of FSV-RideNN is assessed by correlating it with the detection performance of baseline approaches, and FSV-RideNN attained superior performance with a Negative Predictive Value (NPV) of 92.05%, accuracy of 88.48%, Positive Predictive Value (PPV) of 91.56%, True Negative Rate (TNR) of 90.57%, and True Positive Rate (TPR) of 89.68% respectively.