Gastrointestinal Disease Classification from Endoscopy Images Using Glrlm-Based Texture Features and Xgboost Algorithm

Resumen

Gastrointestinal (GI) diseases are among the most common health problems worldwide and can lead to serious complications if not diagnosed at an early stage. Endoscopy is widely used by medical professionals to examine the digestive tract and identify abnormalities such as ulcers, polyps, inflammation, and cancerous lesions. However, analyzing a large number of endoscopic images manually can be time-consuming and may result in variations in diagnosis due to human factors. Therefore, the development of an automated and reliable disease classification system is of great importance. This study presents a machine learning-based approach for the classification of gastrointestinal diseases from endoscopy images using Gray Level Run Length Matrix (GLRLM) texture features and the Extreme Gradient Boosting (XGBoost) algorithm. Initially, the endoscopic images are preprocessed to improve image quality and remove unwanted noise. GLRLM is then employed to extract texture features that effectively describe the structural patterns and characteristics of gastrointestinal tissues. These extracted features serve as inputs to the XGBoost classifier, which is known for its efficiency, robustness, and high predictive performance. The proposed model is evaluated using a dataset containing different categories of gastrointestinal diseases. The classification performance is measured using standard evaluation metrics such as accuracy, precision, recall, and F1-score. The experimental results indicate that the XGBoost classifier successfully distinguishes between various gastrointestinal conditions with high accuracy and improved reliability compared to traditional classification techniques. The findings of this study demonstrate that the combination of GLRLM texture analysis and XGBoost classification can serve as an effective tool for computer-aided diagnosis. The proposed system has the potential to assist healthcare professionals in making faster and more accurate diagnostic decisions, thereby improving the overall quality of patient care.