Pancreatic tumor image segmentation method based on dense connection network transfer learning

Abstract

The invention discloses a pancreatic tumor image segmentation method based on dense connection network transfer learning, and the method comprises the steps: obtaining PET (positron emission computed tomography) and MRI (magnetic resonance imaging), carrying out the preprocessing of the PET and MRI, and dividing the PET and MRI into a training set and a test set; constructing a segmentation network, and training the segmentation network by using a PET training data set to obtain a network parameter W1 after one-time training; setting an initial parameter of a feature extraction module in the segmentation network as a value of a corresponding module in W1 by using a migration strategy, randomly initializing parameters of other modules, and retraining the segmentation network by using an MRI image training set to obtain a secondarily trained network parameter W2; and inputting the MRI test set into a segmentation network taking W2 as a network parameter to obtain a segmentation result. The method improves the performance of MRI image segmentation, solves the problem that a network is difficult to train for a small data set in the prior art, and can be used for assisting a doctor to complete automatic target region sketching before pancreatic tumor treatment.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a method for segmenting images of pancreatic tumors, which can be used to help doctors complete automatic target area delineation before pancreatic tumor treatment. Background technique [0002] At present, pancreatic tumor is still one of the deadliest malignant tumors in the world, and the incidence rate is increasing year by year. According to GLOBOCAN 2020, the latest global cancer burden report released by the International Cancer Institute in 2020, the estimated number of new cases of pancreatic tumors in the world in 2020 is nearly 495,700, and the number of deaths is about 466,000. Due to the poor prognosis of pancreatic tumors, the number of deaths caused by them is almost as high as the number of new cases, and it is the seventh leading cause of death from malignant tumors in both men and women. Pancreatic tumors are expected to overtake breast can...

AI Technical Summary

Problems solved by technology

[0007] The above-mentioned existing pancreatic tumor segmentation methods do not pay attention to the problem of the small number of labeled images available in medical image segmentation, and only use images of one modality, that is, pancreatic tumor CT images or pancreatic tumor MRI images , without fully combining multimodal image information, resulting in low accuracy of pancreatic tumor segmentation, unable to meet the needs of automatic delineation of pancreatic tumor regions before radiotherapy

Images