Diaphragm prediction system and method based on PCA and TSSM models

Abstract

The invention discloses a diaphragm prediction system and method based on PCA and TSSM models. The implementation thought of the diaphragm prediction system and method is as follows: reconstructing breathing time phase data obtained by 4DCT scanning into a human body three-dimensional image of each phase; segmenting a diaphragm image from the human body three-dimensional image, and calculating diaphragm displacement; segmenting a thoracic-abdominal surface image from the human body three-dimensional image, and calculating thoracic-abdominal surface displacement; carrying out principal component analysis on the diaphragm displacement and the thoracic and abdominal surface displacement respectively to reduce dimensions of the diaphragm displacement and the thoracic and abdominal surface displacement and map the diaphragm displacement and the thoracic and abdominal surface displacement into respective d-dimensional subspaces; and calculating a transformation matrix of diaphragm displacement mapped from the thoracic and abdominal surface displacement of the d-dimensional subspace to the d-dimensional subspace by ridge regression optimization. The method has the advantages that less harm is caused to a patient, the selected diaphragm mass center is more accurate, and the anti-interference performance is better.

Description

technical field [0001] The present invention belongs to the field of image processing technology, and further relates to a diaphragm prediction system based on PCA (Principal Components Analysis) and TSSM (Two-Step Subspace Mapping) model in the field of medical image-assisted intervention technology and method. The invention predicts the movement of the human diaphragm through the displacement of the surface of the chest and abdomen outside the body without using markers. Background technique [0002] The dynamic target tracking method is currently the most effective exhalation management technology. The direct method is to image the markers implanted in the body, and then perform target tracking. But the subjects required expensive, invasive surgery. An alternative to implanting markers is to use breathing information to model the motion of internal anatomy. Correlation of lung tumor motion with respiratory motion by spirometry and real-time position tracking system. T...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the above-mentioned deficiencies in the prior art, and propose a diaphragm prediction system and method based on PCA and TSSM models, which is used to solve the problem that the accuracy of the tumor location prediction model established by the markers outside the PLS model is affected by individual patient differences. , measurement accuracy, tumor-surrogate correlation, training data selection, and sinusoidal quadratic model may not be able to hold your breath when collecting patient data, resulting in incomplete data collection

Images