METHOD OF ASSESSING QUALITY OF TELERADIOLOGICAL IMAGES Russian patent published in 2024 - IPC G16H30/00 A61C7/00 G06T7/00 

FIELD: medicine.

SUBSTANCE: invention relates to a method of assessing the quality of teleradiological images. The method is a pre-processing of X-ray images for a diagnostic task in an automated mode, which consists of using algorithms to provide search. Teleradiological images in the lateral projection of the head are analysed to determine the relevance of radiological images for cephalometric analysis. The image is analysed automatically using elements of artificial intelligence based on machine learning and computer vision. Machine learning is carried out by searching for image inconsistencies with the quality criteria of teleradiological images; the following is taken into account: the position of the patient’s head during the X-ray examination, the presence of a craniostat ruler in the area of interest for correct scaling of the resulting image, image deformation that occurred due to the patient’s movement during the X-ray examination, closure of the lips and teeth in the patient’s usual occlusion during the X-ray examination research, correct positioning of ear pads in the patient’s external auditory canals, calibration of x-ray equipment.

EFFECT: taking into account the position of the anatomical formations of the skull, it is possible to analyse telex-ray images automatically using algorithms for determining the relevance of the x-ray image for cephalometric analysis.

1 cl, 1 dwg, 1 ex