目的 探讨智能化抗酸杆菌检测在结核病理诊断中的应用.方法 收集150例肉芽肿性改变的石蜡包埋组织,同时进行手工抗酸染色和自动抗酸染色,分别进行人工阅片病理诊断.使用自主开发的结核杆菌识别人工智能(tuber-culosis bacillus finder,TB-Finder)作为初筛,另请病理医师复核150张自动抗酸染色切片并作出病理诊断,分析两种染色模式的效率及准确率.结果 150例智能化平台染色的阳性率为29.3％(44/150),传统模式染色的阳性率为6.6％(10/150).染色加诊断全流程智能化平台每张切片耗时12 min,传统模式每张切片耗时57 min,智能化平台节约时间成本约78％.结论 智能化抗酸杆菌检测平台的应用可提高病理医师的工作效率和诊断准确率,有助于规范质量控制.

BACKGROUND: Cytologic smear is a routine detection method for pleural effusion, but the results are far from satisfactory. Therefore, pleural effusion is made into cell block for immunochemistry and gene detection, to improve the detection accuracy, which contributes to the treatment and prognosis of pleural effusion. OBJECTIVE: To explore the application value of cell block technology in the pathological diagnosis and molecular detection of pleural effusion. METHODS: Totally 105 cases of pleural effusion samples were collected at Department of Pathology, Peking University First Hospital between January 2016 and August 2017. Effective components extracted after centrifugation were used to make paraffin blocks for hematoxylin-eosin staining and immunocytochemical staining thereafter. Genetic mutations were further detected in the cases of lung cancer diagnosed by cytology.RESULTS AND CONCLUSION: The positive rate of routine cytological smear was 30.5% (32/105), which was significantly lower than that of the cell block section (44.8%, 47/105). The positive rate of immunocytochemical staining combined with cell block was the highest one with 57.1%, 60/105). Besides, this method could be used to identify tumor types. Gene mutation detection was performed in 19 cases of lung cancer, among which 13 cases were positive. These results indicate that cell block technology combined with immunocytochemical staining is an effective method for the diagnosis of malignancy and tumor origin in pleural effusion, and can be further used for mutation detection.