運用電腦輔助分類系統分析肩關節超音波影像以診斷肩旋轉肌病變及肩旋轉肌破裂

Abstract

肩膀酸痛是常見的肌肉骨關節疾病，根據文獻結果，肩膀酸痛之終生盛行率達70%，肩旋轉肌腱病變及破裂是最主要的肩部肌腱傷害之一，在老年人族群更為常見，老年人發生肩旋轉肌病變及斷裂的致病機轉為年齡老化及肌腱組織弱化所致，若發生於年輕族群，多為外傷所致。臨床診斷肩旋轉肌疾病有一定的困難度，因為此類病患症狀及臨床表現不具專一性，理學檢查對於鑑別診斷肩旋轉病變之僅有中度準確度(敏感度及特異度) 。僅由病史及理學檢查是無法確診的，X光檢查無法亦直接辨識肌肉、肌腱、韌帶等軟組織的病變，故診斷肩旋轉病變仰賴影像學檢查，其中肩部超音波為診斷工具之一，而超音波診斷雖然由經驗豐富之操作者來執行，可得到高診斷正確率，診斷正確率可核磁共振檢查相當，但操作者間診斷正確率差異大。故本篇論文研究主題，針對肩關節超音波影像進行電腦影像處理來取得並分析取得之資料，並根據超音波影像所運算出來之資料來診斷並分類肩旋轉肌疾病，期望以電腦輔助診斷分類系統，提供肩關節超音波操作者對肩旋轉肌腱病變及破裂之診斷參考，以降低操作者間診斷正確率差異。 論文第一部分之電腦輔助肩旋轉肌病變診斷分類系統[computer-aided diagnosis (CAD) system] ，將分析之肩部超音波影像進行電腦診斷分類為肩旋轉肌腱炎、鈣化性肌腱炎及肩旋轉肌腱破裂(共三種病變)，納入研究之影像包括43張肩旋轉肌腱炎影像、30張鈣化性肌腱炎影像及26張肩旋轉肌腱破裂影像，針對每張超音波影像之病變區域之區域面積及紋理特徵進行分析，並合併多類別邏輯迴歸分類器進行病變分類診斷。此電腦輔助診斷分類系統可達到87.9%之診斷準確率。並可達以下個別準確率: 診斷肩旋轉肌腱炎88.4%、診斷鈣化性肌腱炎83.3%及診斷肩旋轉肌腱破裂92.3%。 論文第二部分之電腦輔助肩旋轉肌腱破裂診斷分類系統[computer-aided tear classification (CTC) system]，針對肩旋轉肌腱破裂進行電腦輔助診斷，納入研究之影像包括89張肩旋轉肌腱炎影像及102張肩旋轉肌腱破裂影像(共136名病患) ，針對每張超音波影像之病變區域之亮度值及紋理特徵進行分析，並合併二元邏輯迴歸分類器進行病變分類診斷。此電腦輔助診斷肩旋轉肌腱破裂診斷分類系統可達到92% (176/191)之診斷準確率。 根據研究結果，電腦輔助肩旋轉肌病變診斷分類系統(CAD)及電腦輔助肩旋轉肌腱破裂診斷分類系統(CTC)可達良好之診斷正確率，臨床上可提供操作者診斷上之參考。

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.