The invention relates to the technical field of
image analysis, in particular to a dynamic
image registration method for pulmonary atelectasis focus
volume change monitoring. Firstly, a CT
image sequence of a
respiratory cycle of a patient is collected; determining a
breathing gray level change value based on the
gray level difference of the pixel points between the images; dividing a local area according to the position of a pixel point and gray scale information; in a
respiratory cycle,
gas exchange in a pulmonary atelectasis area is limited, expansion capacity during inspiration is limited, a dynamic mode different from a
normal lung tissue is formed, a first abnormal factor is determined by analyzing
gray level change relation of pixel points in a local area and a global area and combining a respiratory gray level change value, and an abnormal area is screened out; due to pulmonary atelectasis,
lung blood vessel displacement and gathering and
lung volume reduction are caused, pixel point number difference characteristics are analyzed, and a second abnormal factor is determined by integrating local region blood vessels, a distribution relation and a
breathing gray level change value; and finally, acquiring an adaptive enhancement factor based on the two, and performing adaptive
sharpening enhancement on the image to improve the subsequent registration precision.