论文部分内容阅读
目的：探索胸腹水中腺癌细胞由人工经验诊断走向计算机自动识别的新方法。方法：用胸腹水中腺癌细胞、增生和正常间皮细胞各３００个，每个细胞用图像分析仪测５项指标制定数学模型输入计算机。计算机即可识别上述三种细胞。结果：用该数学模型检查１８例性质不明的胸腹水，计算机自动识别结果均与细胞病理学专家诊断相符。结论：计算机自动识别胸腹水中腺癌细胞是可行的，但有些问题尚需进一步研究。
OBJECTIVE: To explore a new method for the automatic identification of adenocarcinoma cells in the thoracic and abdominal waters by artificial experience diagnosis. METHODS: Three hundred adenocarcinoma cells, three hyperplasias, and three normal mesothelial cells were used in the thoracic and ascitic fluids. Each cell was used to measure 5 indicators using an image analyzer to formulate a mathematical model and enter the computer. The computer can recognize the above three types of cells. Results: Eighteen unexplained hydrothorax and ascites fluids were examined by this mathematical model. The results of computer automatic recognition were in accordance with the expert diagnosis of cytopathology. Conclusion: Computer automatic identification of adenocarcinoma cells in thoracic and abdominal water is feasible, but some problems still need further study.