دورية أكاديمية
Robust Image Population Based Stain Color Normalization: How Many Reference Slides Are Enough?
| العنوان: | Robust Image Population Based Stain Color Normalization: How Many Reference Slides Are Enough? |
|---|---|
| المؤلفون: | Agraz JL; Center for Biomedical Image Computing and Analytics (CBICA) Philaldelphia PA 19139 USA.; Department of Pathology and Laboratory Medicine, Perelman School of Medicine Philaldelphia PA 19139 USA.; Department of Radiology at Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA., Grenko CM; Department of Pathology and Laboratory Medicine, Perelman School of MedicineUniversity of Pennsylvania and the Center for Interdisciplinary Studies Davidson College NC 28035 USA., Chen AA; Penn Statistical Imaging and Visualization Endeavor (PennSIVE)University of Pennsylvania Philaldelphia PA 19139 USA., Viaene AN; Department of Pathology and Laboratory Medicine, Children's Hospital of PhiladelphiaUniversity of Pennsylvania Philaldelphia PA 19139 USA., Nasrallah MD; Department of Pathology and Laboratory Medicine, Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA., Pati S; CBICA and Department of Pathology and Laboratory Medicine, Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA.; Department of Radiology at Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA., Kurc T; Department of Biomedical InformaticsStony Brook University Stony Brook NY 11794-0751 USA., Saltz J; Department of Biomedical InformaticsStony Brook University Stony Brook NY 11794-0751 USA., Feldman MD; Department of Pathology and Laboratory Medicine, Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA., Akbari H; CBICA and the Department of Radiology, Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA., Sharma P; Symphony Health Philaldelphia PA 19139 USA., Shinohara RT; CBICA and the Penn Statistical Imaging and Visualization Endeavor (PennSIVE)University of Pennsylvania Philaldelphia PA 19139 USA., Bakas S; CBICA, and the Department of Pathology and Laboratory Medicine, Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA.; Department of Radiology, Perelman School of MedicineUniversity of Pennsylvania Philaldelphia PA 19139 USA. |
| المصدر: | IEEE open journal of engineering in medicine and biology [IEEE Open J Eng Med Biol] 2023 Jan 05; Vol. 3, pp. 218-226. Date of Electronic Publication: 2023 Jan 05 (Print Publication: 2022). |
| نوع المنشور: | Journal Article |
| اللغة: | English |
| بيانات الدورية: | Publisher: IEEE Country of Publication: United States NLM ID: 101766631 Publication Model: eCollection Cited Medium: Internet ISSN: 2644-1276 (Electronic) Linking ISSN: 26441276 NLM ISO Abbreviation: IEEE Open J Eng Med Biol Subsets: PubMed not MEDLINE |
| أسماء مطبوعة: | Original Publication: New York, NY : IEEE, [2020]- |
| مستخلص: | Histopathologic evaluation of Hematoxylin & Eosin (H&E) stained slides is essential for disease diagnosis, revealing tissue morphology, structure, and cellular composition. Variations in staining protocols and equipment result in images with color nonconformity. Although pathologists compensate for color variations, these disparities introduce inaccuracies in computational whole slide image (WSI) analysis, accentuating data domain shift and degrading generalization. Current state-of-the-art normalization methods employ a single WSI as reference, but selecting a single WSI representative of a complete WSI-cohort is infeasible, inadvertently introducing normalization bias. We seek the optimal number of slides to construct a more representative reference based on composite/aggregate of multiple H&E density histograms and stain-vectors, obtained from a randomly selected WSI population (WSI-Cohort-Subset). We utilized 1,864 IvyGAP WSIs as a WSI-cohort, and built 200 WSI-Cohort-Subsets varying in size (from 1 to 200 WSI-pairs) using randomly selected WSIs. The WSI-pairs' mean Wasserstein Distances and WSI-Cohort-Subsets' standard deviations were calculated. The Pareto Principle defined the optimal WSI-Cohort-Subset size. The WSI-cohort underwent structure-preserving color normalization using the optimal WSI-Cohort-Subset histogram and stain-vector aggregates. Numerous normalization permutations support WSI-Cohort-Subset aggregates as representative of a WSI-cohort through WSI-cohort CIELAB color space swift convergence, as a result of the law of large numbers and shown as a power law distribution. We show normalization at the optimal (Pareto Principle) WSI-Cohort-Subset size and corresponding CIELAB convergence: a) Quantitatively, using 500 WSI-cohorts; b) Quantitatively, using 8,100 WSI-regions; c) Qualitatively, using 30 cellular tumor normalization permutations. Aggregate-based stain normalization may contribute in increasing computational pathology robustness, reproducibility, and integrity. |
| References: | Arch Histol Cytol. 2004 Jun;67(2):115-34. (PMID: 15468952) IEEE Trans Biomed Eng. 2014 May;61(5):1400-11. (PMID: 24759275) IEEE Rev Biomed Eng. 2009;2:147-71. (PMID: 20671804) Comput Med Imaging Graph. 2017 Apr;57:50-61. (PMID: 27373749) Science. 2018 May 11;360(6389):660-663. (PMID: 29748285) IEEE Trans Inf Technol Biomed. 2009 Mar;13(2):236-51. (PMID: 19171527) Med Image Comput Comput Assist Interv. 2014;17(Pt 1):1-8. (PMID: 25333094) IEEE Trans Biomed Eng. 2003 Jun;50(6):697-704. (PMID: 12814236) IEEE Trans Med Imaging. 2016 Aug;35(8):1962-71. (PMID: 27164577) Appl Immunohistochem Mol Morphol. 2014 Nov-Dec;22(10):713-9. (PMID: 24897076) Annu Int Conf IEEE Eng Med Biol Soc. 2014;2014:194-7. (PMID: 25569930) J Am Med Inform Assoc. 2003 Jan-Feb;10(1):39-51. (PMID: 12509356) Annu Int Conf IEEE Eng Med Biol Soc. 2011;2011:3434-8. (PMID: 22255078) J Microsc. 2010 Nov;240(2):164-72. (PMID: 20946383) Anal Quant Cytol Histol. 2001 Aug;23(4):291-9. (PMID: 11531144) Brainlesion. 2020;11992:44-56. (PMID: 32743562) Med Phys. 1998 Jun;25(6):957-64. (PMID: 9650186) IEEE Trans Pattern Anal Mach Intell. 2021 Mar;43(3):766-785. (PMID: 31603771) IEEE Trans Biomed Eng. 2014 Jun;61(6):1729-38. (PMID: 24845283) Proc IEEE Int Symp Biomed Imaging. 2011 Mar-Apr;2011:657-660. (PMID: 27532016) IEEE J Biomed Health Inform. 2021 Feb;25(2):325-336. (PMID: 33085623) Comput Med Imaging Graph. 2015 Jul;43:89-98. (PMID: 25863518) J Epidemiol Community Health. 2007 Oct;61(10):849-52. (PMID: 17873219) J Pathol Inform. 2013 Jun 29;4:15. (PMID: 23858390) |
| معلومات مُعتمدة: | S10 OD023495 United States OD NIH HHS |
| فهرسة مساهمة: | Keywords: Ivy GAP; Pareto principle rule; Power law; Stain normalization bias problem; Whole slide image |
| تواريخ الأحداث: | Date Created: 20230302 Latest Revision: 20230317 |
| رمز التحديث: | 20240628 |
| مُعرف محوري في PubMed: | PMC9970045 |
| DOI: | 10.1109/OJEMB.2023.3234443 |
| PMID: | 36860498 |
| قاعدة البيانات: | MEDLINE |
Full Text Finder | تدمد: | 2644-1276 |
|---|---|
| DOI: | 10.1109/OJEMB.2023.3234443 |