This invention discloses a method for collaborative registration and alignment of multi-target curve medical images of the female
pelvic floor, comprising: S1: marking the target curve, SCIPP line, and segmentation markers in each sample data; S2: using a
dimensionality reduction method combining Y-Z plane projection and X-coordinate dynamic compensation to reduce the dimensionality of the three-dimensional medical data to two dimensions; S3:
smoothing the curves using cubic spline interpolation before registration to avoid the influence of
noise in the
original data on the registration accuracy; S4: calculating the translation matrix based on the reference point of the SCIPP line at the lower edge of the
pubic symphysis, and aligning the lower edge of the first reference point at the lower edge of the
pubic symphysis. S5: Move to the origin; S6: Use the SCIPP line as a
reference vector and calculate the
rotation matrix using the difference between the
reference vector and the
target angle; S7: Apply the corresponding translation and rotation matrices to the corresponding sample curve data to establish a unified coordinate
system; S8: Calculate the average value of the curve segment based on the marker points, and then combine the average values calculated segment by segment into the average value of the entire curve; S9: Calculate the
centroid, local
centroid, and global
centroid of each curve, and translate the centroid of the corresponding curve to coincide with the local and global centroids of the corresponding curve to achieve alignment. This invention realizes the standardized registration, two-dimensional alignment, and anatomical segmented averaging analysis of female
pelvic organ contour curve data, providing accurate quantitative basis for the diagnosis and
clinical research of
urological diseases.