Scholarlink

<p>Parkinson’s disease (PD) is a neurodegenerative disease that affects the neural, physiological, and behavioral systems of the brain. Diagnosis and monitoring are typically based on a thorough assessment of physical signs by medical specialists, which is limited and unable to identify the disease’s prodromal stages. Recent research has shown voice as a digital biomarker for PD detection. In this paper, an intelligent system for PD diagnosis is presented using speech signals. In developing this intelligent system, machine learning (ML) and deep learning (DL) approaches are used. The experiments are carried out on a Parkinson’s speech dataset that includes 195 voice measurements from 31 different people, each with 23 different speech features. The data is first preprocessed by standard scalar and SMOTE methods and then classification was performed by Support Vector Machine (SVM), Random Forest (RF), AdaBoost, XgBoost Classifiers, and Deep Neural Network (DNN). The results were compared using accuracy, sensitivity, specificity, precision, F-Measures, and AUROC. The results show that the DNN outperformed the other ML classifiers in all performance parameters. The accuracy achieved by DNN is 98.87% and AUC is 0.999. In this research work, a DNN is developed for PD diagnosis using speech samples. Developed DNN has one input layer, three hidden layers, and an output layer. There are 22 nodes in the input layer, 44 nodes in the first hidden layer, 22 nodes in the second hidden layer, 11 nodes in the third hidden layer, and one node in the output layer. The current system is employed as a detection tool for PD diagnosis.</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

Speech-driven diagnostic model for early detection of Parkinson’s disease using machine learning and deep learning techniques

  • 期刊论文

  • 发表日期：

    2026年1月22日

查看全文
  • Rohit Lamba,
  • Pooja Rani,
  • Ravi Kumar Sachdeva,
  • Priyanka Bathla,
  • Ayodeji Olalekan Salau

摘要

Parkinson’s disease (PD) is a neurodegenerative disease that affects the neural, physiological, and behavioral systems of the brain. Diagnosis and monitoring are typically based on a thorough assessment of physical signs by medical specialists, which is limited and unable to identify the disease’s prodromal stages. Recent research has shown voice as a digital biomarker for PD detection. In this paper, an intelligent system for PD diagnosis is presented using speech signals. In developing this intelligent system, machine learning (ML) and deep learning (DL) approaches are used. The experiments are carried out on a Parkinson’s speech dataset that includes 195 voice measurements from 31 different people, each with 23 different speech features. The data is first preprocessed by standard scalar and SMOTE methods and then classification was performed by Support Vector Machine (SVM), Random Forest (RF), AdaBoost, XgBoost Classifiers, and Deep Neural Network (DNN). The results were compared using accuracy, sensitivity, specificity, precision, F-Measures, and AUROC. The results show that the DNN outperformed the other ML classifiers in all performance parameters. The accuracy achieved by DNN is 98.87% and AUC is 0.999. In this research work, a DNN is developed for PD diagnosis using speech samples. Developed DNN has one input layer, three hidden layers, and an output layer. There are 22 nodes in the input layer, 44 nodes in the first hidden layer, 22 nodes in the second hidden layer, 11 nodes in the third hidden layer, and one node in the output layer. The current system is employed as a detection tool for PD diagnosis.