综合声学特征和语音识别错误率的高级优化策略,用于对帕金森病严重程度的多阶段分类
S I M M Raton Mondol1, Ryul Kim2, Sangmin Lee1
1Department of Electrical and Computer Engineering, Inha University, Incheon, Korea.
Biomedical engineering letters
|April 24, 2025
概括
这项研究通过分析语音,区分严重程度来改善帕金森病 (PD) 诊断. 将优化的语音识别与元音声学相结合,可提高早期和高级PD的检测准确度.
科学领域:
- 神经学 神经学
- 语言病理学 语音病理学
- 计算语言学 计算语言学
背景情况:
- 语音分析显示了帕金森病 (PD) 检测的前景.
- 现有的方法往往无法区分不同的PD严重程度.
- 跨PD阶段的不同语言障碍需要精细的诊断方法.
研究的目的:
- 通过区分严重程度来提高帕金森病 (PD) 的诊断准确性.
- 将语音识别错误与声学母音特征集成,以便精确的PD分类.
- 优化语音分析模型,以改善跨PD阶段的检测.
主要方法:
- 利用先进的优化策略,结合语音识别 (字符错误率) 和持续的韩国元音 (/a/, /i/, /u/) 的声学特征.
- 使用的机器学习分类器:随机森林,支持矢量机,k-最近邻居和多层感知器.
- 微调了PD语音识别的Whisper模型,优化了不同的PD严重程度.
主要成果:
- 在PD"ON"状态数据集上,在3级严重程度分类中实现了5.87%的更好的检测准确度.
- 报告了PD"OFF"状态数据集上的3级严重程度分类的7.8%的改进检测准确度.
- 通过综合语音和元音分析,通过PD严重程度水平之间的增强辨别能力.
结论:
- 优化语音识别和母音声学的综合方法有效地检测出不同程度的PD严重程度.
- 与以前的语音分析技术相比,这种方法提供了更精确的帕金森病的分类.
- 针对特定PD阶段的微调模型显著提高了诊断能力.
相关概念视频
Parkinson's Disease: Treatment
148
Neurodegenerative disorders, such as Parkinson's Disease (PD), involve the gradual and irreversible destruction of neurons in particular brain areas. These disorders exhibit standard features like proteinopathies, selective vulnerability of some neurons, and an interaction of intrinsic properties, genetics, and environmental influences in neural injury.
Parkinson's Disease is primarily a result of the loss of dopaminergic neurons in the substantia nigra pars compacta. The cornerstone of...
Parkinson's Disease is primarily a result of the loss of dopaminergic neurons in the substantia nigra pars compacta. The cornerstone of...
148
Parkinson's Disease: Overview
310
Neurodegenerative disorders are progressive diseases that cause irreversible damage and loss to neurons in specific brain areas. Examples of these disorders include Parkinson's disease, Alzheimer's disease, Multiple Sclerosis (MS), and Amyotrophic Lateral Sclerosis (ALS). These disorders share characteristics such as proteinopathies, selective neuronal vulnerability, and a complex interplay between genetic and environmental factors. The primary therapeutic goal for these conditions is...
310


