register_H

This program was designed to register reconstructed prostate pathology to in-vivo MRI

You are now following this Submission

This function takes in a H&E image (moving image) and processes it for
use in a correlative study with in-vivo prostate imaging (fixed image).
The H&E's annotations are filled, for visibility, and it is scaled to
actual size as determined by ImageScope software's measurement of the
scanned histology and converted to pixels by use of the pixel spacing
data found in the DICOM header. Using control point registration,
manually setting 8 control points along the prostate contour, and
piecewise linear affine transformation, the H&E is stretched to the
in-vivo prostate shape.
The images are tested for registration validation by manually creating
masks of the prostates contour, comparing contour overlap at 5mm
resolution, and by percent area overlap calculations. Qualitatively,
transparent and checkerboard overlays are also created. The Gleason
Grade data is formatted for use in ROC curve analysis - it is converted
into a 1D array.
This program necessitates a directory containing the MRI and histology
images you wish to register. Annotations on the histology should be as
follows: green [0 255 0] for Gleason Score 3; blue [0 0 255] for Gleason
Score 4; red [255 0 0] for Gleason Score 5.
Total output: 1mm contour comparison, 5mm contour comparison, transparent
overlay, checkerboard overlay, opaque overlay, transformed registered
histo image, 1D array for Gleason 3+, 1D array for Gleason 4+, radiologic
image ROI, histo image ROI. Arrays and ROIs can be used to validate
prostate cancer detection; the rest can be used to validate the
registration match.

Cite As

Brandon Caldwell (2026). register_H (https://se.mathworks.com/matlabcentral/fileexchange/68179-register_h), MATLAB Central File Exchange. Retrieved .

General Information

MATLAB Release Compatibility

  • Compatible with any release

Platform Compatibility

  • Windows
  • macOS
  • Linux
Version Published Release Notes Action
1.0.0