论文部分内容阅读
收集114例胃癌病例、60例慢性胃病病例和25例正常对照者,选择胃癌流行病学与病因学上有关指标及临床体征、大便隐血等61项因素,根据模式识别原理建立程序在计算机上经自学后建立识别系数,并对146例胃癌、170例胃病、92例正常人进行判别,其符合率分别为89%、55.3%和90.2%,均较单纯应用线性回归筛选为高。结合线性回归方法对本市某厂1108名职工普查筛选出“阳性”者79人,全部经胃镜检查并作活检,发现1例早期胃癌,29例萎缩性胃炎(5例伴有不典型增生),3例胃息肉,其余为不同程度的浅表性胃炎,胃癌普查检出率为0.9‰。
A total of 114 cases of gastric cancer, 60 cases of chronic gastritis and 25 cases of normal controls were collected. Sixty-one factors such as epidemiology and etiology of gastric cancer, clinical signs and fecal occult blood were selected. The procedure was established on the computer according to the principle of pattern recognition. After the self-study, a recognition coefficient was established, and 146 cases of gastric cancer, 170 cases of stomach disease, and 92 cases of normal persons were identified. The compliance rates were 89%, 55.3%, and 90.2%, respectively, which were higher than the linear regression screening. In combination with linear regression methods, a total of 79 people were selected from 1108 general surveys of a factory in the city and 79 were positive. All of them were examined by gastroscopy and biopsy. One case of early gastric cancer and 29 cases of atrophic gastritis (5 cases with atypical hyperplasia) were found. Gastric polyps were found in 3 cases and the rest were superficial gastritis in different degrees. The detection rate of gastric cancer was 0.9%.