Transforming Healthcare with Deep Learning: Chest Condition Diagnosis from CT Scans

This paper examines the Classification of Aortic Aneurysm, Broken Ribs, and Signs of Ischemic Heart Disease utilizing the advanced deep learning framework InceptionV3. A large dataset of computed tomography chest photos was utilised to produce and develop disease classification in this work. The two conditions are connected with a significant death rate. Finally, accessible and accurate diagnostic tools could potentially help treat these ailments before they are life-threatening and extend life. The first steps involved the pre-processing and augmentation of CT chest photos to have enough image diversity and quality. Furthermore, one new deep learning model, InceptionV3, was used throughout the study to raise the models’ feature extraction and classification performance. The model is then fine-tuned for each medical condition to achieve the highest level of predicted accuracy. This study split the dataset that contains numerous annotated images of each condition, to generate a training set and validation set to train and validate the model in a more efficient way. Testing, validation, and training procedures are conducted accorGLQJ WR D VSHFL¿F SODQ WR REVHUYH WKH PRGHO¶V performance. This results from InceptionV3 having an accuracy of 98.1% for these three scenarios implies a good job. According to the results analysis, classification models dependent on deep learning may assist in diagnosing aortic aneurysms, broken ribs, and indicators of ischemic heart disease. Accuracy levels are excellent throughout and suggest that InceptionV3 might truly be an efficient diagnostic tool to reduce healthcare burden and enhance patient outcomes. It could be the base for additional research in the area and even for a healthcare app if deep learning models are realistic for medical image reading.