Gastrointestinal Stromal Tumor: An Uncommon but Serious Cause of Gastrointestinal Bleeding

Bleeding Gastrointestinal Stromal Tumor (GIST) Results in a Sticky Situation: The Good and Bad of Hemostatic Spray

A 50-year-old female presented with symptomatic anemia and hematemesis due to a 3.3 cm gastric gastrointestinal stromal tumor (GIST), which was located in the fundus. Adequate endoscopic views were only achieved in the retroflexed position and attempts at hemostasis via endoscopic clips were unsuccessful. Subsequently, TC-325 hemostatic powder was sprayed on the bleeding lesion and given retroflexed positioning, the powder also coated the esophagogastroduodenoscopy (EGD) scope where it abutted the gastroesophageal junction (GEJ). Hemostasis was successful, but the scope was unable to be withdrawn due to adherence to the surrounding mucosa. With torque maneuvering and a moderate amount of withdrawal force, the scope was successfully freed. The patient was started on imatinib mesylate and did not experience further bleeding episodes. This case highlights the challenge of achieving hemostasis in a bleeding GIST, the beneficial role of hemostatic powder spray, and the need for caution when utilizing it in a retroflexed manner.

CT Image Examination Based on Virtual Reality Analysis in Clinical Diagnosis of Gastrointestinal Stromal Tumors

With the advancement and development of medical equipment, CT images have become a common lung examination tool. This article mainly studies the application of CT imaging examination based on virtual reality analysis in the clinical diagnosis of gastrointestinal stromal tumors. Before extracting suspected lymph nodes from a CT image of the stomach, the CT image sequence is preprocessed first, which can reduce the cumbersomeness of subsequent extraction of suspected lymph nodes and speed up the subsequent processing. According to medical knowledge, CT images of the stomach show that lymph nodes mainly exist in the adipose tissue around the gastric wall, but there are no lymph nodes in the subcutaneous fat outside the chest. The most basic gray value in the image and the neighborhood average difference feature related to gray level are used as the primary features of visual attention detection. When extracting the neighborhood average difference feature, we use a 3 ∗ 3 sliding window method to traverse each point of the pixel matrix in the image, thereby calculating the feature value of each pixel in the image. After the feature extraction is completed, it is necessary to calibrate the data and make a training data set. The SP immunohistochemical staining method was used. The specimens were fixed with 10% formaldehyde, routinely embedded in paraffin, sectioned, and stained with HE. The tumor tissue was determined by immunohistochemistry, and the reagents were products of Maixin Company. All patients were followed up by regular outpatient review, letters, and visits or phone calls. The data showed that immunohistochemical tumor cells showed positive staining for CD117 (14/15, 93.3%) and CD34 (10/15, 66.7%). The results show that the application of virtual reality technology to CT imaging examination can significantly improve the diagnostic accuracy of gastrointestinal stromal tumors.