Gastric cancer is one of the most common malignant tumors of the digestive system, with a high mortality rate due to late-stage diagnosis. Current clinical diagnosis relies on endoscopic biopsy and histopathological analysis, which are highly dependent on pathologists' expertise and may lead to misdiagnoses. Therefore, there is an urgent need for molecular, digital, and intelligent real-time diagnostic methods. In this study, accurate gastric cancer tissue diagnosis was carried out by integrating time-gated Raman spectroscopy (TG-Raman) with deep learning. The TG-Raman effectively suppresses autofluorescence and enhances Raman signal quality using time-resolved detection. A convolutional neural network (CNN)-based model was developed for spectral denoising and feature extraction. Experimental results demonstrated that the proposed approach achieved a classification accuracy of 98.6% for gastric tumor tissues. This study highlights the potential of TG-Raman combined with deep learning for cancer diagnostics, providing a more accurate, efficient, and noninvasive approach for early gastric cancer detection and clinical applications.