Tumor image detection method and device based on meta-learning feature fusion strategy

Abstract

The invention discloses a tumor image detection method and device based on a meta-learning feature fusion strategy, and the method comprises the steps: carrying out the pre-training of a plurality of network models on an open-source data set, inputting a tumor image into the network models, and extracting a multi-level source domain feature vector; performing scale alignment on the scales of the source domain feature vectors and the target domain feature vectors of different scales by using a meta-learning convolution kernel, and fusing the source domain feature vectors after scale unification into a single-level source domain feature map; migrating the multi-scale single-level source domain feature map into a target domain feature vector of the input tumor image according to the structure of the target network based on a meta-learning convolution kernel; and inputting the target domain feature vector fused with the active domain feature vector into a detection network, and completing regression of candidate frames and classification of detection targets by using a cascaded multi-target detection network. The method can be applied to auxiliary detection of a tumor image diagnosis technology, and the detection efficiency of the method is superior to that of naked eye diagnosis detection of medical experts.

Description

technical field [0001] The invention relates to the technical field of tumor image detection, in particular to a tumor image detection method based on a meta-learning feature fusion strategy. Background technique [0002] With the rapid development of medical imaging technology, the number of digital medical images has increased sharply, gradually exceeding the limit of the number that can be processed manually, and facing the thorny problems of time-consuming, subjective, and low efficiency of manual detection of tumor images. In recent years, computer technology has played an increasingly important role in medical diagnosis, and the research on medical image detection methods has become an important research content in image-based computer-aided diagnosis systems. [0003] Using computer image processing technology to analyze and process two-dimensional medical images, to realize the positioning and classification of tumor targets in the slice area, and to assist medical e...

AI Technical Summary

Problems solved by technology

However, the existing deep learning methods cannot expand the semantics of each layer of tumor images, and the source domain information trained by the dataset cannot be transferred to the target domain with a small amount of data, and the learning efficiency of the multi-target detection network is not high.

Images