Data for Gastrointestinal Disorder Prediction

Gastrointestinal (GI) disorders, affecting parts of the digestive system like the stomach and intestines, can be challenging to diagnose, even for skilled gastroenterologists, due to varying presentations. Early diagnosis is crucial for effective treatment, but the review process is time-consuming and labor-intensive. Computer-aided diagnostic (CAD) methods offer a solution, automating the diagnostic process to save time, reduce workload, and minimize the risk of overlooking critical signs. In recent years, many CAD systems have been developed using Machine Learning and Deep Learning approaches to address this issue. Despite preliminary promise, these existing CAD systems still need to be improved to achieve better performance on larger datasets for safety and reliability purposes before being applied in practical medical diagnostics. In our study, we developed a more efficient CAD system to classify eight types of GI images using Transfer Learning (TL) incorporated with the attention mechanism. The combination of TL strategy and attention mechanism showed promising outcomes. Our findings demonstrated that ConvNeXt is an effective pre-trained network for feature extraction and ConvNeXt+Attention (our proposed method) is a robust CAD system that outperformed other state-of-the-arts.