LAUSR

Machine learning techniques are widely used for the diagnosis of cancers. In this study, we proposed a classification and prediction system for the diagnosis of gastric cancer based on saliva samples. Gastric cancer (GC) is classified into early gastric cancer (EGC) and advanced gastric cancer (AGC). The diagnosis of GC at an early stage will improve the survival rate. Computer-aided diagnostic (CAD) systems can assist the radiologists in the diagnosis of EGC. 220 saliva samples were collected from the non-cancerous and gastric cancerous persons and analyzed using high-performance liquid chromatography-mass spectrometry (HPLC-MS). Fourteen amino acid biomarkers were sufficient to distinguish the persons from malignant to benign and were observed in the saliva samples with dominant peaks. We used the support vector machine (SVM) for binary classification. The processed Raman dataset was used to train and test the developed model. SVM based neural networks were established using different kernels, which produced different results. Accuracy, specificity, sensitivity, and receiver operating characteristics (ROC) were used to evaluate the proposed classification model, along with mean average error (MAE), mean square Error (MSE), sum average error (SAE), and sum square error (SSE). We achieved an overall accuracy of 97.18%, specificity of 97.44%, and sensitivity of 96.88% for the proposed method. This established method owns the prospect of clinical translation.