Machine Learning-Based X-ray Images Classification

In hospitals around the world, a vast number of digital X-ray images are taken daily. However, there is a lack of automation in classifying these images, which can lead to time-consuming manual efforts by specialized medical staff. This is especially crucial in situations where immediate intervention is necessary. Additionally, the accuracy of the diagnosis is paramount, but many hospital X-ray datasets are inaccurately populated due to patients having multiple scans scheduled daily. These datasets may contain images of wrong body parts or several empty files, making them unusable for high-level studies without significant time and effort from specialized staff to reorganize them manually. Introducing a classification algorithm using artificial intelligence and machine learning could create a breakthrough in medical imaging and solve this problem. This paper aims to showcase the efficiency specificities of different Deep Learning models and the power of preprocessing images to classify them into different categories.