The present invention provides a
system and method for generating femoral X-
ray films based on
deep learning and digitally reconstructed radiological images, which performs deep multi-task regression through a three-dimensional
convolutional neural network model, and automatically extracts CT slices including the lesser
trochanter and the medial and lateral
femoral condyles Layer, use the conditional generative neural network to segment the medial and lateral
femoral condyles of the lesser
trochanter, perform three-dimensional
surface reconstruction on these two regions, solve the vertices of the lesser
trochanter and the medial and lateral
femoral condyles by calculating the
Gaussian curvature, and calculate the plane of three points The angle between the horizontal plane and the final angle that needs to be rotated is obtained, and the X-
ray film
simulation image of the best position is obtained through
digital reconstruction of the radiological image, replacing the film image used by the traditional CT
simulation positioning
machine. The present invention aims at the problem that the position of the
femur in the current digitally reconstructed radiographic image can only be manually calibrated by doctors, the level of intelligence is not high, the calibration stability is poor, and the actual needs cannot be met, and the computer-aided method is used for
femur CT film correction and X-
ray film
simulation. It can promote the intelligence of
medical equipment.