| Summary: | 博士 === 國立臺灣大學 === 生醫電子與資訊學研究所 === 107 === The lifetime prevalence of shoulder pain approaches 70%, which is mostly attributable to rotator cuff lesions such as inflammation, calcific tendinitis, and tears. On clinical examination, shoulder ultrasound is recommended to detect lesions. However, inter-operator variability of diagnostic accuracy exists due to the operator’ experience and expertise. In this study, a computer-aided diagnosis (CAD) system was developed to assist ultrasound operators in diagnosing rotator cuff lesions and to improve practicality of ultrasound examination. The collected cases included 43 inflammations, 30 calcific tendinitis, and 26 tears. For each case, the lesion area and texture features were extracted from the entire lesions and combined in a multinomial logistic regression classifier for lesion classification. The proposed CAD achieved an accuracy of 87.9%. The individual accuracy of this CAD system was 88.4% for inflammation, 83.3% for calcific tendinitis, and 92.3% for tear groups. The k value of Cohen’s Kappa was 0.798.
In another part of this study, a computer-aided tear classification (CTC) system was developed to identify supraspinatus tears in ultrasound examinations and reduce inter-operator variability. The observed cases included 89 ultrasound images of supraspinatus tendinopathy and 102 of supraspinatus tear from 136 patients. For each case, intensity and texture features were extracted from the entire lesion and combined in a binary logistic regression classifier for lesion classification. The proposed CTC system achieved an accuracy rate of 92% (176/191) and an area under receiver operating characteristic curve (Az) of 0.9694. Based on diagnostic performance, the CAD and CTC systems have promise for clinical use.
|
|---|
