Back pain is considered the most common musculoskeletal condition in the adult population, with a prevalence around 84%, that is, 5 out of 6 people will experience back pain in adulthood. Despite their high morbidity, only 15% of patients with back pain are diagnosed with a specific type of lumbar pathology, a fact that has contributed to the increase in more than 100% of chronic lumbar pathology (Chronic low back pain (CLBP)). This pathology represents the main cause of disability worldwide.
The diagnostic evaluation of patients with back pain can be very challenging and requires complex clinical decision making. One of the main challenges for specialists is discovering among all the musculoskeletal structures potentially involved which or which are the generators of pain. This is a key factor for the management of these patients avoiding therapeutic errors.
Nuclear magnetic resonance imaging (MRI) has become the imaging tool of choice to assess the pathology of the spine, almost completely replacing CT in the routine study of disc degeneration pathologies, infections, trauma, neoplasms and spinal diseases of the spine.
This project aims to incorporate artificial intelligence tools through the design of a convolutional neural network for the treatment of magnetic resonance imaging for the cataloging and gradation of lumbar pathologies.