Intensity-based registration of bright-field and second-harmonic generation images of histopathology tissue sections

Evaluation code and demo code for automatic registration of H&E brightfield image and SHG image of tissue sections.

Demo

1. Proposed method is integrated in CurveAlign program, execute CurveAlign.m in curvealign folder.
2. Running the program without a complete installation of MATLAB is possible, a detailed description of the installation can be found at
   • https://eliceirilab.org/software/curvealign/
   • https://eliceirilab.org/sites/default/files/2017-08/Manual%20%20for%20%20CurveAlign%20V4.0%20Beta%20August%202017.pdf
3. Quick start guide for running in MATLAB
   • Open MATLAB
   • Navigate to curvealign folder and run CurveAlign.m
   • Click BD creation in the main panel
   • Click Get HE Files in the pop-up panel and select the H&E images
   • Click Get SHG Folder in the pop-up panel and seelect the folder containing all the SHG images, each SHG image should have the same file name as the corresponding H&E image.
   • Select the registration method.
     [Auto based on RGB intensity] uses a k-means clustering to segment the H&E images (slower)
     [Auto based on HSV intensity] uses Otsu's method and simple hue channel thresholding to segment the H&E images (faster)
   • Check Reg box at the bottom of the pop-out window
   • Click OK and wait, messages are logged in MATLAB command window

Evaluation

We compared two SIFT-based methods [1-4] and an intensity-based method [5-6] to our proposed method[7].

1. Proposed method can be used by running CurveAlign.m in curvealign folder.
   Detailed instruction of the graphical user interface is in the paper.
2. SIFT can be used by running main_registration.mlx in SIFT-matlab-V1.0 folder.
   File path need to point to the corresponding path storing the dataset. (Input/HE_512 and Input/SHG_512_not_adjusted) Comment out the segmentation part if testing the raw HE input.
3. PSO-SIFT can be used by running main_registration.mlx in PSO-SIFT-matlab-V1.0 folder.
   File path need to point to the corresponding path storing the dataset. (Input/HE_512 and Input/SHG_512_not_adjusted) Comment out the segmentation part if testing the raw HE input.
4. Elastix can be used by running elastix_affine.py.
   SimpleElastix and all dependencies need to be installed. https://simpleelastix.github.io/. File path need to point to the corresponding path storing the dataset. (Input/HE_512 and Input/SHG_512_adjusted for raw HE input; Input/ECM and Input/SHG_512_adjusted for ECM input) Need to change to use either ECM as source image or raw HE as source image in the code.
5. Results are shown in Supplementary figure 1107.docx and comparison folder

Please contact us for any questions

References:
[1] D. G. Lowe, “Distinctive Image Features from Scale-Invariant Keypoints,” Int. J. Comput. Vis. 60, 91–110 (2004).
[2] M. A. Fischler and R. C. Bolles, “Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography,” Commun. ACM 24, 381–395 (1981).
[3] G. Shi, X. Xu, and Y. Dai, “SIFT Feature Point Matching Based on Improved RANSAC Algorithm,” in 2013 5th International Conference on Intelligent Human-Machine Systems and Cybernetics, vol. 1 (2013), pp. 474–477.
[4] W. Ma, Z. Wen, Y. Wu, L. Jiao, M. Gong, Y. Zheng, and L. Liu, “Remote Sensing Image Registration With Modified SIFT and Enhanced Feature Matching,” IEEE Geosci. Remote. Sens. Lett. 14, 3–7 (2017).
[5] S. Klein, M. Staring, K. Murphy, M. Viergever, and J. Pluim, “elastix: A Toolbox for Intensity-Based Medical Image Registration,” IEEE Transactions on Med. Imaging 29, 196–205 (2010).
[6] K. Marstal, F. Berendsen, M. Staring, and S. Klein, “SimpleElastix: A User-Friendly, Multi-lingual Library for Medical Image Registration,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), (2016), pp. 574–582. ISSN: 2160-7516
[7] Adib Keikhosravi, Bin Li, Yuming Liu, and Kevin W. Eliceiri. "Intensity-based registration of bright-field and second-harmonic generation images of histopathology tissue sections." Biomedical Optics Express 11, no. 1 (2020): 160-173.

Citations

@article{keikhosravi_intensity-based_2020,
	title = {Intensity-based registration of bright-field and second-harmonic generation images of histopathology tissue sections},
	volume = {11},
	copyright = {\&\#169; 2019 Optical Society of America},
	issn = {2156-7085},
	url = {https://www.osapublishing.org/boe/abstract.cfm?uri=boe-11-1-160},
	doi = {10.1364/BOE.11.000160},
	number = {1},
	journal = {Biomedical Optics Express},
	author = {Keikhosravi, Adib and Li, Bin and Liu, Yuming and Eliceiri, Kevin W.},
	month = jan,
	year = {2020},
	note = {Publisher: Optical Society of America},
	pages = {160--173}
}