Ventricular region segmentation model training method and device and template image determining method and device

Abstract

The embodiment of the invention discloses a training method of a ventricular region segmentation model, and a method, device and equipment for determining a template image of a brain CT image. The method comprises steps of determining candidate slice images from a three-dimensional brain CT image of a patient, determining a ventricle region and a brain parenchyma region of each candidate slice image, and calculating the area ratio of the ventricle region to the brain parenchyma region of each candidate slice image, and determining the candidate slice image corresponding to the maximum value in the area ratio as a typical slice image, and selecting a template image matched with the area ratio of the ventricle region to the brain parenchyma region of the typical slice image from a template image group corresponding to the age of the patient, and taking the template image as a template image corresponding to the three-dimensional brain CT image. Based on the area ratio of the ventricle region to the brain parenchyma region of the typical slice image, the matching degree of the template image selected from the template image group corresponding to the age of the patient and the three-dimensional brain CT image is high, and the registration difficulty of the subsequent brain CT image and the template image can be reduced.

Description

technical field [0001] The present application relates to the field of image processing, in particular to a training method for a segmentation model of a ventricular region, a method, a device and equipment for determining a template image of a brain CT image. Background technique [0002] Ischemic stroke is a major disease that threatens human health. ASPECTS (Alberta Stoke Program Early CT Score, Alberta Stroke Program Early CT Score) can be used to judge the blood supply of multiple blood supply regions in the MCA (Middle Cerebral Artery, Middle Cerebral Artery) of patients with ischemic stroke, which can help doctors understand the blood supply of patients . [0003] MCA blood supply area mapping is the core link in ASPECTS. MCA blood supply area mapping refers to registering the brain CT image and the template image corresponding to the brain CT image to obtain the deformation field between the brain CT image and the template image corresponding to the brain CT image....

AI Technical Summary

Problems solved by technology

When a variety of different brain CT images are registered based on a fixed template image and the MCA brain atlas image corresponding to the template image, the registration difficulty between the brain CT image and the template image will increase, resulting in low registration accuracy. Affects the mapping of the MCA blood supply, making it difficult to accurately determine the blood supply of the MCA

Images