<p>Research on automated techniques for analysing medical imaging for lung and colon disease is one of the primary areas of focus. In order to improve patient outcomes and survival rates, lung and colon illnesses must be accurately and promptly identified. This study proposes an innovative deep learning approach that seamlessly integrates CNNs and RNNs to accurately detect lung and colon diseases from medical imaging data. The CNN component excels in extracting discriminative visual features from CT scans, while the RNN component effectively models the inherent temporal dependencies within the data. Proposed integrated CNN-RNN models consistently outperformed existing methods, setting new benchmarks for lung and colon disease detection tasks through rigorous evaluation on multiple publicly available datasets, such as the LUNA16 dataset for lung nodule detection, the ILD-Chest dataset for interstitial lung disease classification, and the CVC-Clinic DB and CVC-Video land datasets for polyp detection in colonoscopy videos. The proposed methodology establishes a framework for diagnosing lung and colon diseases. In the present study, extensive experiments were conducted by integrating various CNN architectures, such as DenseNet201, EfficientNetB3, InceptionV3, InceptionResNetV2, MobileNetV3Large, and Xception, with GRU and LSTM components for sequence modelling. The results demonstrated exceptional performance, with the EfficientNetB3, MobileNetV3Large, and Xception architectures combined with GRU and LSTM achieving an impressive accuracy of 99.96%.</p>