A Deep Learning Model for the Classification of Indeterminate Lung Carcinoma in Biopsy Whole Slide Images

Overview

Journal Sci Rep

Specialty Science

Date 2021 Apr 15

PMID 33854137

Citations 20

Authors

Fahdi Kanavati

Gouji Toyokawa

Seiya Momosaki

Hiroaki Takeoka

Masaki Okamoto

Koji Yamazaki

Sadanori Takeo

Osamu Iizuka

Masayuki Tsuneki

Affiliations

Soon will be listed here.

Abstract

The differentiation between major histological types of lung cancer, such as adenocarcinoma (ADC), squamous cell carcinoma (SCC), and small-cell lung cancer (SCLC) is of crucial importance for determining optimum cancer treatment. Hematoxylin and Eosin (H&E)-stained slides of small transbronchial lung biopsy (TBLB) are one of the primary sources for making a diagnosis; however, a subset of cases present a challenge for pathologists to diagnose from H&E-stained slides alone, and these either require further immunohistochemistry or are deferred to surgical resection for definitive diagnosis. We trained a deep learning model to classify H&E-stained Whole Slide Images of TBLB specimens into ADC, SCC, SCLC, and non-neoplastic using a training set of 579 WSIs. The trained model was capable of classifying an independent test set of 83 challenging indeterminate cases with a receiver operator curve area under the curve (AUC) of 0.99. We further evaluated the model on four independent test sets-one TBLB and three surgical, with combined total of 2407 WSIs-demonstrating highly promising results with AUCs ranging from 0.94 to 0.99.

Citing Articles

Application of the 5th WHO Guidelines for the Diagnosis of Lung Carcinoma in Small Lung Biopsies in a Tertiary Care Center: Is Insecurity of Pathologists for the Accurate Diagnosis Justified?.

Beckert M, Meyer C, Papadopoulos T, Levidou G Diagnostics (Basel). 2024; 14(18).

PMID: 39335769 PMC: 11431320. DOI: 10.3390/diagnostics14182090.

Developing a low-cost, open-source, locally manufactured workstation and computational pipeline for automated histopathology evaluation using deep learning.

Choudhury D, Dolezal J, Dyer E, Kochanny S, Ramesh S, Howard F EBioMedicine. 2024; 107:105276.

PMID: 39197222 PMC: 11399610. DOI: 10.1016/j.ebiom.2024.105276.

Deep learning for dual detection of microsatellite instability and POLE mutations in colorectal cancer histopathology.

Gustav M, Reitsam N, Carrero Z, Loeffler C, van Treeck M, Yuan T NPJ Precis Oncol. 2024; 8(1):115.

PMID: 38783059 PMC: 11116442. DOI: 10.1038/s41698-024-00592-z.

The future of artificial intelligence in thoracic surgery for non-small cell lung cancer treatment a narrative review.

Abbaker N, Minervini F, Guttadauro A, Solli P, Cioffi U, Scarci M Front Oncol. 2024; 14:1347464.

PMID: 38414748 PMC: 10897973. DOI: 10.3389/fonc.2024.1347464.

Identifying Pathological Subtypes of Brain Metastasis from Lung Cancer Using MRI-Based Deep Learning Approach: A Multicenter Study.

Li Y, Yu R, Chang H, Yan W, Wang D, Li F J Imaging Inform Med. 2024; 37(3):976-987.

PMID: 38347392 PMC: 11169103. DOI: 10.1007/s10278-024-00988-0.

References

Kyritsis I, Krebs B, Kampe S, Theegarten D, Aigner C, Welter S . Erroneous diagnosis of small cell lung cancer based on small biopsies with far-reaching consequences: case report of a typical carcinoid tumor. J Thorac Dis. 2017; 9(2):E99-E102. PMC: 5334102. DOI: 10.21037/jtd.2017.02.35. View

Goode A, Gilbert B, Harkes J, Jukic D, Satyanarayanan M . OpenSlide: A vendor-neutral software foundation for digital pathology. J Pathol Inform. 2013; 4:27. PMC: 3815078. DOI: 10.4103/2153-3539.119005. View

Bernardi F, Bernardi M, Takagaki T, Siqueira S, Dolhnikoff M . Lung cancer biopsy: Can diagnosis be changed after immunohistochemistry when the H&E-Based morphology corresponds to a specific tumor subtype?. Clinics (Sao Paulo). 2018; 73:e361. PMC: 6201148. DOI: 10.6061/clinics/2018/e361. View

Wu M, Wang B, Gil J, Sabo E, Miller L, Gan L . p63 and TTF-1 immunostaining. A useful marker panel for distinguishing small cell carcinoma of lung from poorly differentiated squamous cell carcinoma of lung. Am J Clin Pathol. 2003; 119(5):696-702. DOI: 10.1309/P5AB-R5KQ-89RN-JTFH. View

Li D, Xu X, Liu J, Liang D, Shi J, Li S . Small cell lung cancer (SCLC) incidence and trends vary by gender, geography, age, and subcategory based on population and hospital cancer registries in Hebei, China (2008-2017). Thorac Cancer. 2020; 11(8):2087-2093. PMC: 7396395. DOI: 10.1111/1759-7714.13412. View

Ao M, Zhang H, Sakowski L, Sharma R, Illei P, Gabrielson E . The utility of a novel triple marker (combination of TTF1, napsin A, and p40) in the subclassification of non-small cell lung cancer. Hum Pathol. 2014; 45(5):926-34. PMC: 4178947. DOI: 10.1016/j.humpath.2014.01.005. View

Coudray N, Ocampo P, Sakellaropoulos T, Narula N, Snuderl M, Fenyo D . Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nat Med. 2018; 24(10):1559-1567. PMC: 9847512. DOI: 10.1038/s41591-018-0177-5. View

Gertych A, Swiderska-Chadaj Z, Ma Z, Ing N, Markiewicz T, Cierniak S . Convolutional neural networks can accurately distinguish four histologic growth patterns of lung adenocarcinoma in digital slides. Sci Rep. 2019; 9(1):1483. PMC: 6365499. DOI: 10.1038/s41598-018-37638-9. View

Zhang H, Liu J, Cagle P, Allen T, Laga A, Zander D . Distinction of pulmonary small cell carcinoma from poorly differentiated squamous cell carcinoma: an immunohistochemical approach. Mod Pathol. 2004; 18(1):111-8. DOI: 10.1038/modpathol.3800251. View

10.

Grilley-Olson J, Hayes D, Moore D, Leslie K, Wilkerson M, Qaqish B . Validation of interobserver agreement in lung cancer assessment: hematoxylin-eosin diagnostic reproducibility for non-small cell lung cancer: the 2004 World Health Organization classification and therapeutically relevant subsets. Arch Pathol Lab Med. 2012; 137(1):32-40. PMC: 5787023. DOI: 10.5858/arpa.2012-0033-OA. View

11.

Haskell C, Holmes E . Non-small cell lung cancer. Dis Mon. 1988; 34(2):53-108. DOI: 10.1016/0011-5029(88)90024-7. View

12.

Grossman R, Heath A, Ferretti V, Varmus H, Lowy D, Kibbe W . Toward a Shared Vision for Cancer Genomic Data. N Engl J Med. 2016; 375(12):1109-12. PMC: 6309165. DOI: 10.1056/NEJMp1607591. View

13.

Loo P, Thomas S, Nicolson M, Fyfe M, Kerr K . Subtyping of undifferentiated non-small cell carcinomas in bronchial biopsy specimens. J Thorac Oncol. 2010; 5(4):442-7. DOI: 10.1097/JTO.0b013e3181d40fac. View

14.

Yu K, Zhang C, Berry G, Altman R, Re C, Rubin D . Predicting non-small cell lung cancer prognosis by fully automated microscopic pathology image features. Nat Commun. 2016; 7:12474. PMC: 4990706. DOI: 10.1038/ncomms12474. View

15.

Litjens G, Sanchez C, Timofeeva N, Hermsen M, Nagtegaal I, Kovacs I . Deep learning as a tool for increased accuracy and efficiency of histopathological diagnosis. Sci Rep. 2016; 6:26286. PMC: 4876324. DOI: 10.1038/srep26286. View

16.

Warth A, Muley T, Herpel E, Meister M, Herth F, Schirmacher P . Large-scale comparative analyses of immunomarkers for diagnostic subtyping of non-small-cell lung cancer biopsies. Histopathology. 2012; 61(6):1017-25. DOI: 10.1111/j.1365-2559.2012.04308.x. View

17.

Gurda G, Zhang L, Wang Y, Chen L, Geddes S, Cho W . Utility of five commonly used immunohistochemical markers TTF-1, Napsin A, CK7, CK5/6 and P63 in primary and metastatic adenocarcinoma and squamous cell carcinoma of the lung: a retrospective study of 246 fine needle aspiration cases. Clin Transl Med. 2015; 4:16. PMC: 4417108. DOI: 10.1186/s40169-015-0057-2. View

18.

Korbar B, Olofson A, Miraflor A, Nicka C, Suriawinata M, Torresani L . Deep Learning for Classification of Colorectal Polyps on Whole-slide Images. J Pathol Inform. 2017; 8:30. PMC: 5545773. DOI: 10.4103/jpi.jpi_34_17. View

19.

Leslie K, Gruden J, Parish J, Scholand M . Transbronchial biopsy interpretation in the patient with diffuse parenchymal lung disease. Arch Pathol Lab Med. 2007; 131(3):407-23. DOI: 10.5858/2007-131-407-TBIITP. View

20.

Pinheiro F, Teixeira A, de Brito B, da Silva F, Santos M, de Melo F . Immunotherapy - new perspective in lung cancer. World J Clin Oncol. 2020; 11(5):250-259. PMC: 7360520. DOI: 10.5306/wjco.v11.i5.250. View