Development and validation of a deep learning algorithm for improving Gleason scoring of prostate cancer

Abstract

For prostate cancer patients, the Gleason score is one of the most important prognostic factors, potentially determining treatment independent of the stage. However, Gleason scoring is based on subjective microscopic examination of tumor morphology and suffers from poor reproducibility. Here we present a deep learning system (DLS) for Gleason scoring whole-slide images of prostatectomies. Our system was developed using 112 million pathologist-annotated image patches from 1226 slides, and evaluated on an independent validation dataset of 331 slides. Compared to a reference standard provided by genitourinary pathology experts, the mean accuracy among 29 general pathologists was 0.61 on the validation set. The DLS achieved a significantly higher diagnostic accuracy of 0.70 ( p = 0.002) and trended towards better patient risk stratification in correlations to clinical follow-up data. Our approach could improve the accuracy of Gleason scoring and subsequent therapy decisions, particularly where specialist expertise is unavailable. The DLS also goes beyond the current Gleason system to more finely characterize and quantitate tumor morphology, providing opportunities for refinement of the Gleason system itself.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 58

1. 1.

   Prostate Cancer—Cancer Stat Facts. https://seer.cancer.gov/statfacts/html/prost.html. Accessed 22 August 2018.

   Not in our library

2. 2.

   Epstein JI, et al. A contemporary prostate cancer grading system: a validated alternative to the Gleason score. Eur. Urol. 2016;69:428-435.

   Not in our library10.1016/j.eururo.2015.06.046

3. 3.

   Epstein JI, Allsbrook WC, Amin MB, Egevad LL. The 2005 International Society of Urological Pathology (ISUP) consensus conference on gleason grading of prostatic carcinoma. Am. J. Surg. Pathol. 2005;29:1228-1242.

   Not in our library10.1097/01.pas.0000173646.99337.b1

4. 4.

   Epstein, J. I. et al. The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma: Definition of Grading Patterns and Proposal for a New Grading System. Am. J. Surg. Pathol. 40, 244–252 (2016).

   Not in our library10.1097/pas.0000000000000530

5. 5.

   NCCN Clinical Practice Guidelines in Oncology. https://www.nccn.org/professionals/physician_gls/default.aspx#prostate. Accessed 14 August 2018.

   Not in our library

6. 6.

   Persson J, et al. Interobserver variability in the pathological assessment of radical prostatectomy specimens: findings of the Laparoscopic Prostatectomy Robot Open (LAPPRO) study. Scand. J. Urol. 2014;48:160-167.

   Not in our library10.3109/21681805.2013.820788

7. 7.

   Veloso SG, et al. Interobserver agreement of Gleason score and modified Gleason score in needle biopsy and in surgical specimen of prostate cancer. Int. Braz. J. Urol. 2007;33:639-646.

   Not in our library10.1590/s1677-55382007000500005

8. 8.

   Montironi R, Lopez-Beltran A, Cheng L, Montorsi F, Scarpelli M. Central prostate pathology review: should it be mandatory? Eur. Urol. 2013;64:199-201.

   Not in our library10.1016/j.eururo.2013.04.002

9. 9.

   Bottke D, et al. Phase 3 study of adjuvant radiotherapy versus wait and see in pT3 prostate cancer: impact of pathology review on analysis. Eur. Urol. 2013;64:193-198.

   Not in our library10.1016/j.eururo.2013.03.029

10. 10.

    Egevad L, et al. Standardization of Gleason grading among 337 European pathologists. Histopathology. 2013;62:247-256.

    Not in our library10.1111/his.12008

11. 11.

    Netto, G. J., Eisenberger, M., Epstein, J. I. & TAX 3501 Trial Investigators. Interobserver variability in histologic evaluation of radical prostatectomy between central and local pathologists: findings of TAX 3501 multinational clinical trial. Urology77, 1155–1160 (2011).

    Not in our library10.1016/j.urology.2010.08.031

12. 12.

    Allsbrook WC, et al. Interobserver reproducibility of Gleason grading of prostatic carcinoma: urologic pathologists. Hum. Pathol. 2001;32:74-80.

    Not in our library10.1053/hupa.2001.21134

13. 13.

    Allsbrook WC, et al. Interobserver reproducibility of Gleason grading of prostatic carcinoma: general pathologist. Hum. Pathol. 2001;32:81-88.

    Not in our library10.1053/hupa.2001.21135

14. 14.

    Mikami Y, et al. Accuracy of gleason grading by practicing pathologists and the impact of education on improving agreement. Hum. Pathol. 2003;34:658-665.

    Not in our library10.1016/s0046-8177(03)00191-6

15. 15.

    Esteva A, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017;542:115-118.

    Not in our library10.1038/nature21056

16. 16.

    Haenssle HA, et al. Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann. Oncol. 2018;29:1836-1842.

    Not in our library10.1093/annonc/mdy166

17. 17.

    Gulshan V, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA. 2016;316:2402-2410.

    Not in our library10.1001/jama.2016.17216

18. 18.

    Ting DSW, et al. Development and validation of a deep learning system for diabetic retinopathy and related eye diseases using retinal images from multiethnic populations with diabetes. JAMA. 2017;318:2211-2223.

    Not in our library10.1001/jama.2017.18152

19. 19.

    Burlina PM, et al. Automated grading of age-related macular degeneration from color fundus images using deep convolutional neural networks. JAMA Ophthalmol. 2017;135:1170-1176.

    Not in our library10.1001/jamaophthalmol.2017.3782

20. 20.

    De Fauw J, et al. Clinically applicable deep learning for diagnosis and referral in retinal disease. Nat. Med. 2018;24:1342-1350.

    Not in our library10.1038/s41591-018-0107-6

21. 21.

    Kermany DS, et al. Identifying medical diagnoses and treatable diseases by image-based deep learning. Cell. 2018;172:1122-1131.e9.

    Not in our library10.1016/j.cell.2018.02.010

22. 22.

    Rajpurkar, P. et al. CheXNet: radiologist-level pneumonia detection on chest X-rays with deep learning. Preprint at arXiv [cs.CV]. https://arxiv.org/abs/1711.05225 (2017).

    Not in our library

23. 23.

    Chilamkurthy Sasank, Ghosh Rohit, Tanamala Swetha, Biviji Mustafa, Campeau Norbert G, Venugopal Vasantha Kumar, Mahajan Vidur, Rao Pooja, Warier Prashant. Deep learning algorithms for detection of critical findings in head CT scans: a retrospective study. The Lancet. 2018;392(10162):2388-2396.

    Not in our library10.1016/s0140-6736(18)31645-3

24. 24.

    Ehteshami Bejnordi B, et al. Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. JAMA. 2017;318:2199-2210.

    Not in our library10.1001/jama.2017.14585

25. 25.

    Liu, Y. et al. Detecting cancer metastases on gigapixel pathology images. Preprint at arXiv [cs.CV]. https://arxiv.org/abs/1703.02442 (2017).

    Not in our library

26. 26.

    Campanella, G., Silva, V. W. K. & Fuchs, T. J. Terabyte-scale deep multiple instance learning for classification and localization in pathology. Preprint at arXiv [cs.CV]. https://arxiv.org/abs/1805.06983 (2018).

    Not in our library

27. 27.
    Automated Gleason grading of prostate cancer tissue microarrays via deep learning · 2018

    Corrected
28. 28.

    Steiner David F., MacDonald Robert, Liu Yun, Truszkowski Peter, Hipp Jason D., Gammage Christopher, Thng Florence, Peng Lily, Stumpe Martin C. Impact of Deep Learning Assistance on the Histopathologic Review of Lymph Nodes for Metastatic Breast Cancer. The American Journal of Surgical Pathology. 2018;42(12):1636-1646.

    Not in our library10.1097/pas.0000000000001151

29. 29.

    Liu Yun, Kohlberger Timo, Norouzi Mohammad, Dahl George E., Smith Jenny L., Mohtashamian Arash, Olson Niels, Peng Lily H., Hipp Jason D., Stumpe Martin C. Artificial Intelligence–Based Breast Cancer Nodal Metastasis Detection: Insights Into the Black Box for Pathologists. Archives of Pathology & Laboratory Medicine. 2019;143(7):859-868.

    Not in our library10.5858/arpa.2018-0147-oa

30. 30.

    Leo P, et al. Stable and discriminating features are predictive of cancer presence and Gleason grade in radical prostatectomy specimens: a multi-site study. Sci. Rep. 2018;8:14918.

    Not in our library10.1038/s41598-018-33026-5

31. 31.

    Sparks R, Madabhushi A. Statistical shape model for manifold regularization: Gleason grading of prostate histology. Comput. Vis. Image Under. 2013;117:1138-1146.

    Not in our library10.1016/j.cviu.2012.11.011

32. 32.

    Nguyen, K., Jain, A. K. & Allen, R. L. Automated gland segmentation and classification for gleason grading of prostate tissue images. In: 2010 20th International Conference on Pattern Recognition, 23–26 August 2010, Istanbul (2010).

    Not in our library

33. 33.

    Jiménez del Toro, O. et al. Convolutional neural networks for an automatic classification of prostate tissue slides with high-grade Gleason score. In Medical Imaging 2017: Digital Pathology (Orlando, Florida, USA, 2017).

    Not in our library

34. 34.

    Ma, Z. et al. Semantic segmentation for prostate cancer grading by convolutional neural networks. In Medical Imaging 2018: Digital Pathology (Houston, Texas, USA, 2018).

    Not in our library

35. 35.

    Kallen, H., Molin, J., Heyden, A., Lundstrom, C. & Astrom, K. Towards grading gleason score using generically trained deep convolutional neural networks. 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI) (Prague, Czech Republic, 2016).

    Not in our library

36. 36.

    Zhong Q, et al. A curated collection of tissue microarray images and clinical outcome data of prostate cancer patients. Sci. Data. 2017;4:170014.

    Not in our library10.1038/sdata.2017.14

37. 37.

    van der Kwast TH, et al. Impact of pathology review of stage and margin status of radical prostatectomy specimens (EORTC trial 22911). Virchows Arch. 2006;449:428-434.

    Not in our library10.1007/s00428-006-0254-x

38. 38.

    Srigley JR, et al. Protocol for the examination of specimens from patients with carcinoma of the prostate gland. Arch. Pathol. Lab. Med. 2009;133:1568-1576.

    Not in our library10.5858/133.10.1568

39. 39.

    Humphrey PA, Moch H, Cubilla AL, Ulbright TM, Reuter VE. The 2016 WHO classification of tumours of the urinary system and male genital organs—Part B: prostate and bladder tumours. Eur. Urol. 2016;70:106-119.

    Not in our library10.1016/j.eururo.2016.02.028

40. 40.

    Epstein JI, Amin MB, Reuter VE, Humphrey PA. Contemporary Gleason grading of prostatic carcinoma: an update with discussion on practical issues to implement the 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma. Am. J. Surg. Pathol. 2017;41:e1-e7.

    Not in our library10.1097/pas.0000000000000820

41. 41.

    Sauter G, et al. Clinical utility of quantitative gleason grading in prostate biopsies and prostatectomy specimens. Eur. Urol. 2016;69:592-598.

    Not in our library10.1016/j.eururo.2015.10.029

42. 42.

    Cox, D. R. Regression models and life-tables. in Springer Series in Statistics (eds Kotz, S., & Johnson, N. L.) 527–541 (Springer, New York, NY, 1992).

    Not in our library

43. 43.

    Brimo F, Schultz L, Epstein JI. The value of mandatory second opinion pathology review of prostate needle biopsy interpretation before radical prostatectomy. J. Urol. 2010;184:126-130.

    Not in our library10.1016/j.juro.2010.03.021

44. 44.

    Zhou M, et al. Diagnosis of ‘poorly formed glands’ gleason pattern 4 prostatic adenocarcinoma on needle biopsy: an interobserver reproducibility study among urologic pathologists with recommendations. Am. J. Surg. Pathol. 2015;39:1331-1339.

    Not in our library10.1097/pas.0000000000000457

45. 45.

    Shah RB, et al. Diagnosis of Gleason Pattern 5 prostate adenocarcinoma on core needle biopsy. Am. J. Surg. Pathol. 2015;39:1242-1249.

    Not in our library10.1097/pas.0000000000000442

46. 46.
    Grading of prostatic adenocarcinoma: current state and prognostic implications · 2016

    No notice
47. 47.

    Aeffner F, et al. The gold standard paradox in digital image analysis: manual versus automated scoring as ground truth. Arch. Pathol. Lab. Med. 2017;141:1267-1275.

    Not in our library10.5858/arpa.2016-0386-ra

48. 48.

    Wang, D., Khosla, A., Gargeya, R., Irshad, H. & Beck, A. H. Deep learning for identifying metastatic breast cancer. Preprint at arXiv [q-bio.QM]. https://arxiv.org/abs/1606.05718 (2016).

    Not in our library

49. 49.

    Ehteshami Bejnordi, B. et al. Deep learning-based assessment of tumor-associated stroma for diagnosing breast cancer in histopathology images. In 2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017) (Melbourne, Australia, 2017).

    Not in our library10.1109/isbi.2017.7950668

50. 50.

    Weinstein JN, et al. The Cancer Genome Atlas Pan-Cancer analysis project. Nat. Genet. 2013;45:1113-1120.

    Not in our library10.1038/ng.2764

51. 51.

    Liu J, et al. An integrated TCGA Pan-cancer clinical data resource to drive high-quality survival outcome analytics. Cell. 2018;173:400-416.e11.

    Not in our library10.1016/j.cell.2018.02.052

52. 52.

    Stephenson AJ, et al. Defining biochemical recurrence of prostate cancer after radical prostatectomy: a proposal for a standardized definition. J. Clin. Oncol. 2006;24:3973-3978.

    Not in our library10.1200/jco.2005.04.0756

53. 53.

    Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. & Wojna, Z. Rethinking the Inception architecture for computer vision. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (Las Vegas, NV, USA, 2016).

    Not in our library

54. 54.

    Chen, P.-H. C. et al. Microscope 2.0: an augmented reality microscope with real-time artificial intelligence integration. Preprint at arXiv [cs.CV]. https://arxiv.org/abs/1812.00825 (2018).

    Not in our library10.1038/s41591-019-0539-7

55. 55.

    Bejnordi BE, et al. Stain specific standardization of whole-slide histopathological images. IEEE Trans. Med. Imaging. 2016;35:404-415.

    Not in our library10.1109/tmi.2015.2476509

56. 56.

    Zoph, B., Vasudevan, V., Shlens, J. & Le, Q. V. Learning transferable architectures for scalable image recognition. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (Salt Lake City, UT, USA, 2018).

    Not in our library

57. 57.

    Chollet, F. Xception: deep learning with depthwise separable convolutions. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (Honolulu, HI, USA, 2017).

    Not in our library

58. 58.

    Cohen J. A coefficient of agreement for nominal Ssales. Educ. Psychol. Meas. 1960;20:37-46.

    Not in our library10.1177/001316446002000104

Cited by papers we hold 13

• Unveiling key pathomic features for automated diagnosis and Gleason grade estimation in prostate cancer · 2025
  Not checked
• Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models · 2023
  No notice
• Artificial intelligence for diagnosis and Gleason grading of prostate cancer: the PANDA challenge · 2022
  No notice
• Human-interpretable image features derived from densely mapped cancer pathology slides predict diverse molecular phenotypes · 2021
  No notice
• Artificial intelligence assistance significantly improves Gleason grading of prostate biopsies by pathologists · 2021
  No notice
• Review of deep learning: concepts, CNN architectures, challenges, applications, future directions · 2021
  No notice
• Reporting guidelines for clinical trials of artificial intelligence interventions: the SPIRIT-AI and CONSORT-AI guidelines · 2021
  No notice
• PyHIST: A Histological Image Segmentation Tool. · 2020
  Not checked
• Deep learning-based survival prediction for multiple cancer types using histopathology images · 2020
  No notice
• Deep learning interpretation of echocardiograms · 2020
  No notice
• Tumor-Associated Macrophages in Human Breast, Colorectal, Lung, Ovarian and Prostate Cancers · 2020
  No notice
• Next generation diagnostic pathology: use of digital pathology and artificial intelligence tools to augment a pathological diagnosis · 2019
  Not checked
• Translational AI and Deep Learning in Diagnostic Pathology · 2019
  No notice

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.