Enhancing Diagnostic Accuracy for Lung Disease in Medical Images: A Transfer-Learning Approach

To efficiently control healthcare costs, early and precise diagnosis of lung illnesses is essential. Using deep learning (DL) and transfer learning (TL), this study proposes a fresh method for categorizing five types of lung illnesses. Ten thousand chest X-ray pictures were used as a training and assessment dataset. To improve results, we used a TL model predicated on the MobileNet V2 structure. The testing findings show that the suggested model works, successfully diagnosing lung illnesses with an overall accuracy of 95.05%. Performance parameters showed that the evaluated performance of each class was superior to that of other state-of-the-art models. The model’s utility in diagnosing lung disorders exemplifies its applicability to imaging diagnostics. Furthermore, the TL strategy is shown to be more accurate and computationally efficient when compared to preexisting models. Testing the model on a variety of data sets attested to its sturdiness and generalizability. This study shows promise for improving the diagnosis and detection of lung diseases by utilizing DL and TL. The suggested paradigm makes it easier to implement preventative measures, individualize patient care, and boost health outcomes.