Background:Endoscopic ultrasonography( EUS)is the first choice for examining submucosal protuberant lesions in the gastrointestinal tract,however,it is very difficult to differentiate gastric stromal tumor from gastric leiomyoma. Photoshop software now is tried to differentiate some easily misdiagnosed diseases. Aims:To investigate the value of Photoshop software in differentiating EUS image of gastric stromal tumor from gastric leiomyoma. Methods:EUS image of 118 cases of gastric stromal tumor and 42 cases of gastric leiomyoma confirmed by pathological and immunohistochemical examination were collected,mean gray value( intensity of echo)and gray value standard deviation( uniformity of echo)of EUS image of lesion were determined by Photoshop software,and the differentiation value was analyzed. Results:The mean gray value of EUS image of gastric stromal tumor was significantly higher than that of gastric leiomyoma(71. 94 vs. 48. 99,P<0. 01). The gray value standard deviation of EUS image of gastric stromal tumor was significantly higher than that of gastric leiomyoma(16. 63 vs. 9. 80,P<0. 01). Conclusions:Under EUS,gastric stromal tumor exhibits higher intensity of echo and less uniformity of echo than gastric leiomyoma. EUS combined with Photoshop software is helpful for differentiating gastric stromal tumor from gastric leiomyoma by analyzing mean gray value and gray value standard deviation of the lesion.