Application of Computed Tomography Image Information Technology Based on Optimized Discrete Cosine Transform Filtering Algorithm in Diagnosis of Gastric Neuroendocrine Tumors

Abstract

Objective: To investigate the clinicopathological characteristics and CT manifestations of gastric neuroendocrine tumors using an optimized DCT filtering algorithm. Methods: An optimized DCT filtering algorithm was used to retrospectively analyze the CT image data and clinical pathological data of 31 patients with gastric neuroendocrine tumors. Results: Of the 31 patients, 23 were male and 8 were female, with a median age of 63 years. There were 4 cases of clinical type I, the lesions were located on the major curve side in 3 cases, and the minor curve side in 1 case; 3 cases had CT with polypoid moderately enhanced nodules, and 1 case was missed; pathological grades were all G1 grade; the maximum diameter of the lesion was 1.7∼6.3 cm, average 1.1 cm. There were 2 cases of clinical type II, 1 case was polypoid on the gastric fundus, grade G1; 1 case was on the gastric antrum, CT showed mild gastric wall thickening, grade G2, with liver metastasis. There were 25 cases of clinical type III, 14 cases of cardia gastric fundus, 4 cases of small curvature of the gastric body, 3 cases of large curvature of the gastric body, and 4 cases of gastric antrum; CT signs showed localized thickening of the gastric wall and 3 cases of gastric wall focal 15 cases were irregularly thickened gastric wall with or without soft tissue masses, 6 of which had ulcers on the surface; the largest diameter of the lesions was 1.7 to 6.3 cm with an average of 4.5 cm; the density of the lesions was 11 and 14 were uneven; The enhanced scan showed significant enhancement in 16 cases, moderate enhancement in 9 cases, uniform enhancement in 8 cases, and uneven enhancement in 17 cases, of which the enhancement peak was located in the arterial phase; 25 cases were all grade G3, 10 were local lymph node metastases, and 4 were liver metastases. There were 4 cases of liver metastasis with local lymph node metastasis, and 1 case of intraperitoneal implantation with local lymph node metastasis. Conclusion: The optimized DCT filtering algorithm can enumerate the CT features of gastric neuroendocrine tumors with different clinical and pathological types. The combination of clinical and CT data is valuable for preoperative diagnosis.