DETECTION OF KNEE-JOINT DISEASES USING ACOUSTICAL PATTERN-RECOGNITION TECHNIQUE
DETECTION OF KNEE-JOINT DISEASES USING ACOUSTICAL PATTERN-RECOGNITION TECHNIQUE
复制标题
DOI：
10.1016/0021-9290(76)90149-4
复制
发表时间：
1976-01-01
影响因子：
2.4
通讯作者：
BOWLING, GF
中科院分区：
文献类型：
--
作者：
CHU, ML;GRADISAR, IA;BOWLING, GF
开发了一种分析患者主动运动期间正常和病理膝关节发出的声音的方法。该技术简单、无痛，非常适合关节炎筛查项目。在膝盖主动运动期间，膝盖声音被记录在磁带上。利用噪声消除原理将背景噪声降至最低，从而可以分析所谓的声学特征。该签名是应用于随机噪声的模式识别技术的产物。对正常膝关节、类风湿关节炎膝关节和退行性膝关节记录的研究表明，它们各自的波形、频谱模式和自相关统计特性似乎是独特的，并且很可能被证明是一种有前途的非侵入性诊断工具，用于早期检测膝关节损伤的类型和程度。
A method of analyzing the sound emitted by normal and pathological knee joints of patients during active movements was developed. The technique is brief and painless and lends itself well to an arthritis screening program. The knee sound is recorded on magnetic tape during active motion of the knee. Background noise is minimized using the principle of noise cancellation, thereby permitting analysis of what might be called an acoustic signature. This signature is the product of pattern recognition technique applied to random noise. Studies of knee recordings covering normal, rheumatoid arthritic and degenerative knees show that their respective waveforms, spectral patterns and statistical property of auto-correlation appear to be unique, and may well prove to be a promising non-invasive diagnostic tool for early detection of the type and extent of knee joint damage.