背景與目的：分析骨轉移素片的錯誤判讀原因。方法：國泰綜合醫院從2004 年8 月至2006 年7 月間共有307 個在素片上被錯誤判讀的骨轉移病變。其中有283 個病變被排除，因為在三位放射線科醫師（包括一位資深骨關節放射線科醫師）仔細看了五分鐘後也找不到病變。其餘24 個病變依它們的位置、臨床徵兆、原發瘤的地點分析，並將其分為感知錯誤或認知錯誤的病變。感知錯誤的病變如果花費少於2 分鐘就察覺到它，定義為「被忽略」的病變；如果花費超過2 分鐘才察覺到它或是無法察覺它，則定義為「不明顯」的病變。如果誤將骨轉移診斷為良性則定義為認知錯誤的病變。結果：錯誤判讀的骨轉移有9 個在脊椎，7 個在骨盆，6 個在肋骨，1 個在肩胛骨，1 個在股骨。錯誤判讀的原因包括17 個被忽略的病變，5 個不明顯的病變及2 個認知錯誤的病變。結論：骨轉移在素片的錯誤判讀原因大多數是感知錯誤，其中多數是因為病變被忽略。每位醫師應該仔細的判讀每一根肋骨、每一節脊椎及被腸氣擋住的骨頭，而即使發現一個病變後也要持續尋找下個病變。唯有遵循以上判讀準則才能降低骨轉移在素片的錯誤判讀。

Background and Purpose: To analyze the causes of misdiagnosis of bone metastases on plain radiography. Methods: We retrospectively reviewed the charts and imaging studies of patients diagnosed with bone metastases at Cathay General Hospital from August 2004 to July 2006. In total, 307 lesions were misdiagnosed by radiologists on reading a plain film, 283 of which were excluded from this analysis because they could not be detected even after scrutiny for ＞5min by 3 radiologists (including 1 senior osteoradiologist). The remaining 24 lesions were analyzed according to their location, clinical symptoms, site of the primary tumor, and whether they were perception-error or recognition-error lesions. A perception-error lesion was classified as having been