A Neural Network Based Attention Mechanism for Abdominal Lymph Node Segmentation

Abstract

The invention relates to the technical field of abdominal lymph node partitioning, in particular to an abdominal lymph node partitioning method based on the neural network of the attention mechanism, which is used to solve the problem of large differences in the reading results of the same abdominal CT medical image by doctors in the prior art. The problem of inaccurate prediction of lymph node partition. The present invention comprises the following steps: step 1: data preparation; step 2: mask generation, preprocessing the data; step 3: constructing an attention mechanism residual network model; step 4: repeating step 3, constructing and training the relative position of lymph nodes Partitioned model; Step 5: Use the model trained in Step 3 and Step 4 to classify the abdominal lymph nodes automatically detected by the detection task. In the present invention, the superposition of the original CT image and the mask is used as input, and the attention mechanism is introduced into the deep residual neural network, so that the abdominal lymph nodes in the CT image can be accurately partitioned.

Description

Technical field[0001]The present invention relates to the technical field of abdominal lymph node region, and more particularly to a method of abdominal lymph node region based on the attention mechanism neural network.Background technique[0002]The prior art intermunan node metastasis is one of the common metastasis recurrence forms of colorectal cancer. For low-level rectal cancer, lower rectal cancer, lower rectal cancer, lower rectal cancer, and the whole abdomen reinforced CT. Scanning is an important imaging method for clinical discretion to determine the lymph node positioning of colorectal cancer. It is one of the main ways of the clinician for identifying whether the side lymph nodes has metastasis. Due to the use of developers, lymph node detection is not A further difficult task, but because the distribution of abdominal organs is complex, the abdominal lymph node partition is difficult to define, and the lymph nodes cannot be accurately cleared, which is a pretty cleaning...

AI Technical Summary

Problems solved by technology

[0005] Based on the above problems, the present invention provides an abdominal lymph node partition method based on the neural network of the attention mechanism, which is used to solve the problem of large differences in the reading results of the same abdominal CT medical image by doctors in the prior art and the prediction of abdominal lymph node partition. inaccurate question

Images