A New Method for Constructing the Correspondence Between Sequence Images and Respiratory Signals

Abstract

The invention discloses a new method for establishing the corresponding relation between sequences image and respiratory signals. The method comprises the steps that in a motion tracking method frame, a binocular stereo vision system is used for collecting displacement of an infrared marker as the human respiratory signals; a new-generation CT scanning machine is used for acquiring complete 4D CT images of a subject in a deep breathing mode; distance movement information of the infrared marker is integrated, a distance measure function is constructed, a distance corresponding correlation method is put forward, and the corresponding relation between the 4D CT images and the respiratory signals is determined by comparing displacement of the marker in the 4D CT images and displacement of the marker in the respiratory signals within the same time period. By means of the method, the operating procedures of acquiring the respiratory signals are simplified, the data measurement accuracy is improved, the experimental time is shortened, and the radiation dosage received by the subject is reduced.

Description

technical field [0001] The invention belongs to the field of motion modeling, and in particular relates to a method for constructing a corresponding relationship between sequence images and respiratory signals in image navigation surgery, which improves the accuracy of the correspondence between sequence images and respiratory signals. Background technique [0002] In image-guided surgery, respiratory motion is one of the important factors causing inaccurate lesion localization. Furthermore, without awareness of respiratory motion when planning radiation therapy for lung cancer patients, the tumor may receive insufficient doses, surrounding healthy tissue may receive unnecessary doses, or both. Therefore, it is necessary to construct a motion model of lung tissue to simulate and predict the influence of respiratory motion on the lesion. [0003] In order to construct a lung breathing motion model, it is necessary to simultaneously acquire 4D lung CT images and respiratory s...

AI Technical Summary

Problems solved by technology

[0009] 1. The steps to obtain external respiration signals are cumbersome, and it is easy to introduce errors in the process of extracting the marker core;

[0010] 2. When acquiring CT images, in the breath-hold mode, it is difficult to control the breathing state when acquiring data, which is not suitable for patients with respiratory dysfunction; and there are differences in lung tissue movement between the breath-hold state and the free breathing state, resulting in corresponding errors;

[0011] 3. In the free breathing mode, the scanning time is long, which may cause discomfort to the subject and increase the radiation dose received by the subject;

[0012] 4. When constructing the corresponding relationship between sequence images and in vitro signals, the method used cannot ensure that the real-time anatomical structure of the subject’s end-inspiration or end-expiration lung tissue is consistent with the 3D lung tissue selected as the end-inspiration or end-expiration state of the lung tissue. The internal CT is consistent, resulting in a corresponding error;

[0013] 5. At the same time, although the lung breathing motion model constructed by this method can estimate the current lung state corresponding to any position in the breathing cycle, it cannot monitor the external signals and make real-time estimates

Images