Development of Audiogram Database with Pattern Recognition Function for Audiogram Classification

作者研制了一个计算机辅助听力图自动分类系统，该系统采用句法分析方法，具有听力图模式识别功能。该系统是在微机上用MUMPS语言开发的。该系统将听力图模式信息压缩为六个字符串。这些字符串可以作为听力图分类的知识。该系统可以评估气骨导差、平均听力损失，并可以每隔0.58秒检测和检索听力图的特征。作者将该系统的判断与医生的判断进行了比较。该系统可有效地应用于噪声环境下工人听力损失的筛查和自动健康检测。

The authors have developed an automated computer-aided audiogram classification system which has the function of pattern recognition of audiograms by the syntactic analysis method. The system was developed on micro-computer and MUMPS language. This system compressed the audiogram pattern information by six characters strings. This strings could treat as the knowledge of the audiogram classification. Therefore, it is easy to standardize the audiogram classification.This system evaluate the air-bone gap, the average hearing loss, and also detect and retrieve the characteristics of the audiogram in every 0.58 sec. The authors compared the judgment of this system and those of the physicians. There were not so much difference between the two judgments.This new system can be effectively applied in screening audiometry of workmen's hearing loss in noisy environment and automated health testing.