A Hybrid RNN-CNN Approach for Enhanced Prediction of Tuberculosis and Lung Cancer

Tuberculosis (TB) and lung cancer are major global health concerns, with early detection being crucial for effective treatment. However, healthcare systems often face shortages of medical professionals and diagnostic facilities, hindering timely diagnosis. Current diagnostic systems mostly rely on algorithms like ResNet for TB prediction, but they fall short in diagnosing lung cancer and COVID-19 accurately. This study proposes a novel hybrid approach combining Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN) to improve diagnostic accuracy for both TB and lung cancer. RNNs are effective at analyzing sequential data, which is useful for temporal analysis, while CNNs specialize in extracting spatial features from medical imaging like X-rays and CT scans. By integrating both architectures, the proposed model enhances prediction capabilities, addressing the limitations of existing systems. With the use of both architectures' strengths, our hybrid model realizes a high accuracy of classification as 98.5%, This hybrid approach has the potential to enable earlier diagnoses, improving treatment outcomes and contributing to better healthcare management during overlapping disease outbreaks.