论文部分内容阅读
目的 用形态定量的方法检测纤支镜活检的小细胞肺癌细胞核与淋巴细胞核的差异,为图像分析用于两者的鉴别诊断提供依据。方法 对39例小细胞肺癌及10例慢性炎症(测其淋巴细胞)进行图像分析测量,分别测量其胞核面积、等效直径、周长、形状因子、圆球度及异型指数等六个指标,并与淋巴细胞核进行对比统计、分析。结果 小细胞肺癌细胞核与淋巴细胞核在各项形态学指标之间均有极显著差异,以二者999%可信区间为标准,对39例小细胞肺癌进行验证,诊断符合率分别为949%(37/39)和100%(10/10)。结论 计算机图像分析对于鉴别小细胞肺癌和淋巴细胞很有帮助,尤其对于淋巴细胞样癌的鉴别诊断更有价值
Objective To detect the difference of nuclear and lymphocyte nuclei in small cell lung cancer by fiberoptic biopsy with morphometric method and provide the basis for the differential diagnosis of the two by image analysis. Methods Thirty-nine cases of small cell lung cancer and 10 cases of chronic inflammation (lymphocytes were measured) were measured by image analysis. Six indexes of nucleus area, equivalent diameter, perimeter, shape factor, sphericity and atypia index , And compared with the lymphocyte nucleus statistics, analysis. Results Small cell lung cancer nuclei and lymphocyte nuclei showed significant differences among various morphological indicators. The accuracy of 999% confidence interval was validated in 39 small cell lung cancer patients, and the diagnostic coincidence rates were 949% 37/39) and 100% (10/10). Conclusion Computer image analysis is very useful for the identification of small cell lung cancer and lymphocytes, especially for the differential diagnosis of lymphocytic carcinoma