• Dr. Pandikumar S
• Dr.M.Sengaliappan

Erythroplakia is an infrequent, quarantined, red, velvety lesion which affects patients mainly in the 6th and 7th decades. This study proposes a hybrid approach for the detection of oral cancer by integrating Support Vector Machine (SVM) and Recurrent Neural Network (RNN) classifiers, utilizing features extracted from Gray Level Run Length Matrix (GLRLM) analysis of medical images. GLRLM, a texture analysis method, captures important textural features that are indicative of malignancy in oral tissues. These features are then fed into SVM and RNN models, which are trained to differentiate between cancerous and non-cancerous lesions. The SVM classifier is employed for its robustness in handling high-dimensional data and separating classes with a maximum margin, while the RNN leverages its capability to learn sequential dependencies, enhancing the model's performance in capturing complex patterns within the image data. Experimental results demonstrate that the combined SVM and RNN approach with GLRLM features significantly improves the accuracy, sensitivity, and specificity of oral cancer detection compared to conventional methods. This hybrid model showcases the potential of advanced machine learning techniques in medical diagnostics, paving the way for more reliable and efficient early detection of oral cancer.

Erythroplakia, GLRLM, RNN, SVM, Naive Bayes

"HYBRID APPROACH FOR THE DETECTION OF ERYTHROPLAKIA ORAL CANCER USING SVM AND RNN WITH GLRLM FUSION MODEL", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 9, page no.e178-e182, September-2024, Available :http://www.jetir.org/papers/JETIR2409422.pdf

"HYBRID APPROACH FOR THE DETECTION OF ERYTHROPLAKIA ORAL CANCER USING SVM AND RNN WITH GLRLM FUSION MODEL", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 9, page no. ppe178-e182, September-2024, Available at : http://www.jetir.org/papers/JETIR2409422.pdf