| The objective of our study is to evaluate MRI plain scanning in diagnosis of gross types and intermural invasion of advanced gastric cancer. With hypotonia and water filling, we performed low-field MRI with three sequences such as T1WI, T2WI and GRE in transverse and coronal plane in 29 advanced gastric cancer(AGC), pathologically proved by endoscopic biopsy and /or double-contrast barium studies, and study the structures and characteristics of normal and cancerous gastric wall, and location , length, gross types and intermural invasion of AGC in MRI, correlating them with surgery and pathology. No layered normal gastric wall, with moderate and slightly high signal intensity, was shown in T1WI and GRE sequence. While there were 12 patients with two-layered structure normal gastric wall, of which the inner layer was slightly high signal intensity, and the outer one low signal intensity in T2WI sequence image, and no layer with low-signal intensity shown in the other 17 patients. This layered structure was considered relevant to gastric distention and different sequence. In all 29 patients, distinct boundary between normal gastric wall and cancerous one was found in 11(37.93%),while gradual boundary in 18 (62.07%) ,and ulcer found in 20 patients. In all 29 patients, no layered cancerous wall with iso-signal intensity, compared with normal gastric wall, was found in 26 patients, while layered structure was shown in 3 patients, of which two-layered cancerous gastric wall with inner high signal intensity and outer slightly low signal intensity was found in T1WI sequence, while three-layered structure in T1WI sequence, in which the inner layer was shown hyperintense, the middle slightly hyperintense, the outer hypointense. And all of three patients were pathologically proved mucinous adenocarcinoma. Therefore the cancerous wall with layered structure was probably considered mucinous adenocarcinoma.In Our study, the detecting rate of AGC with transverse and coronal scanning was 100%. The superiority of MRI in judging the location of AGC is its ability to achieve multi-planar image such as transverse, coronal, sagittal and oblique image, especially coronal image, which is similar to that of gastrointestinal dual-contrast barium study , can show clearly gastric great curve, small curve and gastric angle, thus helpful to locate the tumor.When judging gross types of AGC, the difference between MRI and gastrointestinal dual-contrast barium study is that we can achieve cross image, which is helpful to judge the base of cancer and shape of gastric wall around it. Meantime, theability of multi-planar imaging is very useful for three-dimentionally judging gross types of AGC. Our study detected 86.21% gross types of AGC accurately with transverse and coronal plane image, considering the characteristics of boundary between normal and cancerous gastric wall, whether or not ulcer, the length of tumor, and differentiated 85% of Borr2 from Borr3 accurately. In all patients of ulcerous gastric cancer, the length of eight Borr2 AGC was 50.24 ±22.65mm, while the length of thirteen Borr3 AGC was 89.32 ±28.64mm, with t-test, there was statistically significant difference (P<0.01). So this result demonstrated that the length of Borr3 was longer than that of Borr2. Therefore, With multiplanar and multisequence image, the plain scanning of low-field MRI is able to accurately judge the location and gross type of AGC.Accuracy of judging T-staging of advanced gastric cancer of each sequence was 65.52% by T1WI, 72.41% by T2WI and 72.41% by GRE. The accuracy of FSE/ T2WI and GRE are higher than that of T1WI, but through the x2 test, there is no obvious difference (P>0.05). Comprehensively analyzing image of all sequences, total accuracy of judging T-staging increased to 75.86%. The accuracy of GRE sequence in T-staging is high, but because of its inherent defects, it may lead to overstaging and wrong locating, therefore it is essential to combine it with other sequence. FSE/T2WI should be used as primary T-staging sequence for it... |
PDF Full Text Request