During 2024, there were more than 245 thousand cases of pneumonia in Brazil which led to admissions tohospitals, an increase of 30% compared to the previous year. Therefore, our objective was thedevelopment of techniques that provide a quick and reliable identification of this disease. In that regard,we developed an image recognition application, capable of distinguishing unhealthy individuals from apool of X-ray images, and identifying which type of pneumonia (viral or bacterial) present. Thisapplication would be accessible to healthcare professionals such as nurses and physicians via auser-friendly front end application such as a website. In order to identify the images, we trained amachine learning model using a dataset1 containing labelled X-ray images of both healthy andunhealthy lungs, and later on we switched to a deep learning model. When an external test was run, theresults were not impressive so we decided to return to the previous model and improve it using the KFoldtechnique. The result obtained was a model with both accuracy and precision close to 75%, capable ofreliably detecting unhealthy lungs and which type of pneumonia was present if any. This model did notuse CNN and was composed of XGBoost with KFold and image treatment. After that the model wasmade available through a proof of concept website, which let the user upload an X-ray image of theirchest and receive the model’s prediction, so even lay people could use the application.
Lima et al. (Fri,) studied this question.