Procesamiento de radiografía de tórax mediante morfología matemática para facilitar la identificación de problemas óseos

Abstract

Chest x-ray is the most frequently performed radiological analysis on patients; However, the assessment of bone pathologies through these studies has been lagged by doctors, so the present research work focuses on the detection of cervical rib pathologies and enchondroma or bone infarction, with the in order to provide assistance and support to doctors, so that they have a tool that allows them to make a more efficient and timely diagnosis of their patients. The study focuses on the development of an algorithm that allows determining whether or not the anomalies under analysis are suffering, and subsequently highlighting and displaying them for the knowledge of the doctor and the patient. To carry out the work, the information contained in the database provided by BIMCV (Medical Image Bank of the Valencian Community) was taken; the processing of the images was carried out with Mathematical Morphology, mainly based on the use of filters. The results obtained indicate a detection sensitivity greater than 90%, and a specificity greater than 95%, which proves the behavior of the developed algorithm, this is corroborated with the validation carried out by a Traumatologist. The determining factor for bone pathologies to be detected and highlighted is the establishment of detection thresholds, this is because x-rays have certain differences in positioning and image quality, in addition to the location, shape and size of the pathologies