A texture image grading method for pituitary tumors based on fine-grained medical image segmentation and truth discovery data augmentation

Abstract

The invention discloses a method for grading pituitary tumor texture images based on fine-grained medical image segmentation and true value discovery data amplification, comprising the following steps: Step 1: medical image segmentation based on fine-grained optimization of Drosophila-density peak clustering; Step 2: Amplification of pituitary tumor data based on true value discovery; Step 3: Grading of pituitary tumor texture images based on Step 1 and Step 2. The present invention accurately segments medical images through the fusion of fine-grained partitioning algorithm and FOA-DPC algorithm; the present invention also realizes medical image data amplification based on truth discovery theory, and solves the problem of too few available medical image data sets. The invention combines the KFOA‑DPC segmentation algorithm with deep learning to solve the problem of dicom format images with complex gray levels and difficult feature extraction, realize the soft and tough classification of pituitary tumors, and assist clinical diagnosis.

Description

Technical field [0001] The present invention relates to a pituitary tumor imaging grading method, and more particularly to a pituitary tumor imaging grading method based on fine-grained medical image division and truth discovery data amplification, belonging to the technical field of medical image processing. Background technique [0002] Pituitoma is a group of tumors that occur from pituitary front leaf and post-leaf and craniopharyngeal residual cells, which is very frequent, accounting for about 10% of intracranial tumors. The texture of pituitary tumors affects the approach and surgical plan of surgery. At present, with the development of minimally invasive technologies, minimally invasive surgery of the butterfly sinus into the road has become the preferred treatment, but it is only suitable for soft texture. Pituitary tumor, and for a small part of the tumor tumors, it is difficult to scrape in the adoption of butterfly sinus surgery, and it is necessary to cut even repeat...

AI Technical Summary

Problems solved by technology

[0008] The technical problem to be solved by the present invention is to provide a pituitary tumor texture image classification method based on fine-grained medical image segmentation and truth value discovery data amplification. The problem of too few medical image data sets has also realized the grading of the soft and tough texture of pituitary tumors, thereby assisting clinical diagnosis

Images