医疗语音分析处理系统的研究与实现/Research and Implementation of a Medical Speech Analyzing and P

2018-08-26 14:41:45

语音 processing speech medical analyzing



医疗语音分析处理系统是一个以辅助诊断嗓音疾病为目的,面向临床应用的系统。该系统融合了多种技术和知识,属于信息学和医学的交叉研究领域。现代医疗迫切要求嗓音诊断的自动化、科学化和客观化,尤其是嗓音诊断的计算机实现,是现今医疗系统应用发展的一个重要研究方向。由于现有的医疗语音分析处理系统采用的技术比较陈旧,而专用的语音诊断设备价格比较昂贵,如何利用医院现有的计算机设备,并结合高新科学技术,使嗓音诊断向着高可靠性、高效性、实用性的方向发展是迫切需要解决的一个问题。
本论文通过研究和分析设计了一个新型的、实用的医疗语音分析处理系统来诊断嗓音疾病。在结合医院实际情况的基础上,提出和设计了医疗语音分析处理系统的总体结构和功能模块。接着对语音分析处理中所涉及的医学知识和参数如频率微扰、振幅微扰、谐噪比、声门噪声能量等进行了研究,并对语音信号计算机化过程中所采用的基本技术如模数变换、WAVE文件格式等进行了讨论和分析。之后对语音分析处理中所使用的关键技术进行了比较深入的研究,采用了先进的小波变换方法来提取嗓音医学参数基音频率,并对传统的临床辅助诊断方法语谱图进行了详细分析。最后在研究的基础上,给出了具体实现过程和结果。
本论文最终实现的医疗语音分析处理系统具有简便性、实用性和准确性等特点。具有一定的应用前景


Medical speech analyzing and processing system is a system that is used in clinic application, and aimed at auxiliary diagnosing voice disease. It combined many technologies and information, and belongs to intersecting research field of informatics and physics. Modern medical treatment imperiously needs automatization, science and impersonality of voice diagnosing, especially the computer realization of voice diagnosing which is an important research direction in the development of modern medical treatment system. Since the technologies used in existing medical speech analyzing and processing system are old, and special speech analyzing devices are expensive, it is a great challenge to make voice diagnosing systems with high reliability, high efficiency and high practicability by using the exiting computer devices in hospital and new scientific technologies.
This thesis designed and implemented a new high efficient medical speech analyzing and processing system to diagnose voice diseases. On the basis of combing the practical conditions of hospital, the whole system structure and function modules of medical speech analyzing and processing system are put forwarded. Then some research has been down in the medical knowledge and parameters in speech analyzing and processing such as Jitter, Shimmer, HNR and NNE, and some basic technologies in the process of transferring acoustic signals into computer signals are discussed and analyzed, such as A/D Conversion, and the WAVE file format. Whereafter deep research is done on some important technologies in speech analyzing and processing, the advanced wavelet transforms technology is used to pick-up pitch which is an important medical parameter of voice, and detailed analysis is done in spectrogram which is a traditional medical auxiliary diagnosing method. Finally on the basis of these researches, detailed implementation steps and results are presented.
The medical speech processing system implemented in this thesis has the advantage of convenience, practicability and accuracy. It has fabulous application trend.