Improved method for nose classification

Our research work was focused in the medical field, specifically the Otorhinolaryngology (ENT) field in order to improve the diagnostic accuracy of doctors while examining radio images. Our goal was to automatically recognize whether a nose is healthy or fractured, so we developed an enhanced approach that considers many parameters (input images in different sections, complex data ...). We will detail throughout this article the different stages of our work which led to fracture detection results that are accurate enough according to experts.