Enhancing Predictive Accuracy in Lung Disease Diagnosis Through Hybrid ResNet and Transfer Learning Models

Accurate and timely diagnosis of lung diseases are crucial for effective medical intervention and improved patient outcomes. This research investigates the efficacy of deep learning algorithms, including CNN, ResNet, Transfer Learning, and a novel hybrid ResNet-Transfer Learning model, to predict lung diseases from medical images. Utilizing diverse datasets encompassing lung conditions such as cancer and chronic diseases, the study meticulously compares the performance of four deep learning models. The hybrid ResNet- Transfer Learning model emerges as a key focus, leveraging the combined strengths of ResNet and Transfer Learning. Results from comprehensive experiments demonstrate the superior performance of the hybrid model, surpassing baseline algorithms in terms of accuracy, sensitivity, and F1-score. Comparative analysis reveals a significant average accuracy improvement of 8.5% and a 6.2 % increase in F1-score, highlighting the promising potential of the proposed model. This research contributes to advancing methodologies for lung disease prediction, emphasizing the impact of hybrid deep learning approaches on diagnostic precision. The findings underscore the pivotal role of algorithmic advancements in enhancing medical image analysis, paving the way for more effective tools in clinical settings.