An Artificial Intelligence-based Support Tool for Automation and Standardisation of Gleason Grading in Prostate Biopsies

Felicia Marginean; Ida Arvidsson; Athanasios Simoulis; Niels Christian Overgaard; Kalle Åström; Anders Heyden; Anders Bjartell; Agnieszka Krzyzanowska

doi:10.1016/j.euf.2020.11.001

An Artificial Intelligence-based Support Tool for Automation and Standardisation of Gleason Grading in Prostate Biopsies

Felicia Marginean, Ida Arvidsson, Athanasios Simoulis, Niels Christian Overgaard, Kalle Åström, Anders Heyden, Anders Bjartell, Agnieszka Krzyzanowska

Skåne University Hospital

Research output: Contribution to journal › Article › peer-review

Abstract

BACKGROUND: Gleason grading is the standard diagnostic method for prostate cancer and is essential for determining prognosis and treatment. The dearth of expert pathologists, the inter- and intraobserver variability, as well as the labour intensity of Gleason grading all necessitate the development of a user-friendly tool for robust standardisation.

OBJECTIVE: To develop an artificial intelligence (AI) algorithm, based on machine learning and convolutional neural networks, as a tool for improved standardisation in Gleason grading in prostate cancer biopsies.

DESIGN, SETTING, AND PARTICIPANTS: A total of 698 prostate biopsy sections from 174 patients were used for training. The training sections were annotated by two senior consultant pathologists. The final algorithm was tested on 37 biopsy sections from 21 patients, with digitised slide images from two different scanners.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Correlation, sensitivity, and specificity parameters were calculated.

RESULTS AND LIMITATIONS: The algorithm shows high accuracy in detecting cancer areas (sensitivity: 100%, specificity: 68%). Compared with the pathologists, the algorithm also performed well in detecting cancer areas (intraclass correlation coefficient [ICC]: 0.99) and assigning the Gleason patterns correctly: Gleason patterns 3 and 4 (ICC: 0.96 and 0.94, respectively), and to a lesser extent, Gleason pattern 5 (ICC: 0.82). Similar results were obtained using two different scanners.

CONCLUSIONS: Our AI-based algorithm can reliably detect prostate cancer and quantify the Gleason patterns in core needle biopsies, with similar accuracy as pathologists. The results are reproducible on images from different scanners with a proven low level of intraobserver variability. We believe that this AI tool could be regarded as an efficient and interactive tool for pathologists.

PATIENT SUMMARY: We developed a sensitive artificial intelligence tool for prostate biopsies, which detects and grades cancer with similar accuracy to pathologists. This tool holds promise to improve the diagnosis of prostate cancer.

Original language	English
Pages (from-to)	995-1001
Journal	European Urology Focus
Volume	7
Issue number	5
Early online date	2020 Dec 7
DOIs	https://doi.org/10.1016/j.euf.2020.11.001
Publication status	Published - 2021

Subject classification (UKÄ)

Urology and Nephrology
Medical Laboratory and Measurements Technologies

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1016/j.euf.2020.11.001Licence: CC BY-NC-ND

1 Doctoral Thesis (compilation)

Prostate Cancer Tissue Biomarkers
Marginean, F.-E., 2022, Lund: Lund University, Faculty of Medicine. 75 p.
Research output: Thesis › Doctoral Thesis (compilation)

Open Access
File

Cite this

@article{1535e8a45f5f4c319f82d42acb570f23,

title = "An Artificial Intelligence-based Support Tool for Automation and Standardisation of Gleason Grading in Prostate Biopsies",

abstract = "BACKGROUND: Gleason grading is the standard diagnostic method for prostate cancer and is essential for determining prognosis and treatment. The dearth of expert pathologists, the inter- and intraobserver variability, as well as the labour intensity of Gleason grading all necessitate the development of a user-friendly tool for robust standardisation.OBJECTIVE: To develop an artificial intelligence (AI) algorithm, based on machine learning and convolutional neural networks, as a tool for improved standardisation in Gleason grading in prostate cancer biopsies.DESIGN, SETTING, AND PARTICIPANTS: A total of 698 prostate biopsy sections from 174 patients were used for training. The training sections were annotated by two senior consultant pathologists. The final algorithm was tested on 37 biopsy sections from 21 patients, with digitised slide images from two different scanners.OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Correlation, sensitivity, and specificity parameters were calculated.RESULTS AND LIMITATIONS: The algorithm shows high accuracy in detecting cancer areas (sensitivity: 100%, specificity: 68%). Compared with the pathologists, the algorithm also performed well in detecting cancer areas (intraclass correlation coefficient [ICC]: 0.99) and assigning the Gleason patterns correctly: Gleason patterns 3 and 4 (ICC: 0.96 and 0.94, respectively), and to a lesser extent, Gleason pattern 5 (ICC: 0.82). Similar results were obtained using two different scanners.CONCLUSIONS: Our AI-based algorithm can reliably detect prostate cancer and quantify the Gleason patterns in core needle biopsies, with similar accuracy as pathologists. The results are reproducible on images from different scanners with a proven low level of intraobserver variability. We believe that this AI tool could be regarded as an efficient and interactive tool for pathologists.PATIENT SUMMARY: We developed a sensitive artificial intelligence tool for prostate biopsies, which detects and grades cancer with similar accuracy to pathologists. This tool holds promise to improve the diagnosis of prostate cancer.",

author = "Felicia Marginean and Ida Arvidsson and Athanasios Simoulis and {Christian Overgaard}, Niels and Kalle {\AA}str{\"o}m and Anders Heyden and Anders Bjartell and Agnieszka Krzyzanowska",

year = "2021",

doi = "10.1016/j.euf.2020.11.001",

language = "English",

volume = "7",

pages = "995--1001",

journal = "European Urology Focus",

issn = "2405-4569",

publisher = "Elsevier",

number = "5",

}

TY - JOUR

T1 - An Artificial Intelligence-based Support Tool for Automation and Standardisation of Gleason Grading in Prostate Biopsies

AU - Marginean, Felicia

AU - Arvidsson, Ida

AU - Simoulis, Athanasios

AU - Christian Overgaard, Niels

AU - Åström, Kalle

AU - Heyden, Anders

AU - Bjartell, Anders

AU - Krzyzanowska, Agnieszka

PY - 2021

Y1 - 2021

N2 - BACKGROUND: Gleason grading is the standard diagnostic method for prostate cancer and is essential for determining prognosis and treatment. The dearth of expert pathologists, the inter- and intraobserver variability, as well as the labour intensity of Gleason grading all necessitate the development of a user-friendly tool for robust standardisation.OBJECTIVE: To develop an artificial intelligence (AI) algorithm, based on machine learning and convolutional neural networks, as a tool for improved standardisation in Gleason grading in prostate cancer biopsies.DESIGN, SETTING, AND PARTICIPANTS: A total of 698 prostate biopsy sections from 174 patients were used for training. The training sections were annotated by two senior consultant pathologists. The final algorithm was tested on 37 biopsy sections from 21 patients, with digitised slide images from two different scanners.OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Correlation, sensitivity, and specificity parameters were calculated.RESULTS AND LIMITATIONS: The algorithm shows high accuracy in detecting cancer areas (sensitivity: 100%, specificity: 68%). Compared with the pathologists, the algorithm also performed well in detecting cancer areas (intraclass correlation coefficient [ICC]: 0.99) and assigning the Gleason patterns correctly: Gleason patterns 3 and 4 (ICC: 0.96 and 0.94, respectively), and to a lesser extent, Gleason pattern 5 (ICC: 0.82). Similar results were obtained using two different scanners.CONCLUSIONS: Our AI-based algorithm can reliably detect prostate cancer and quantify the Gleason patterns in core needle biopsies, with similar accuracy as pathologists. The results are reproducible on images from different scanners with a proven low level of intraobserver variability. We believe that this AI tool could be regarded as an efficient and interactive tool for pathologists.PATIENT SUMMARY: We developed a sensitive artificial intelligence tool for prostate biopsies, which detects and grades cancer with similar accuracy to pathologists. This tool holds promise to improve the diagnosis of prostate cancer.

AB - BACKGROUND: Gleason grading is the standard diagnostic method for prostate cancer and is essential for determining prognosis and treatment. The dearth of expert pathologists, the inter- and intraobserver variability, as well as the labour intensity of Gleason grading all necessitate the development of a user-friendly tool for robust standardisation.OBJECTIVE: To develop an artificial intelligence (AI) algorithm, based on machine learning and convolutional neural networks, as a tool for improved standardisation in Gleason grading in prostate cancer biopsies.DESIGN, SETTING, AND PARTICIPANTS: A total of 698 prostate biopsy sections from 174 patients were used for training. The training sections were annotated by two senior consultant pathologists. The final algorithm was tested on 37 biopsy sections from 21 patients, with digitised slide images from two different scanners.OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Correlation, sensitivity, and specificity parameters were calculated.RESULTS AND LIMITATIONS: The algorithm shows high accuracy in detecting cancer areas (sensitivity: 100%, specificity: 68%). Compared with the pathologists, the algorithm also performed well in detecting cancer areas (intraclass correlation coefficient [ICC]: 0.99) and assigning the Gleason patterns correctly: Gleason patterns 3 and 4 (ICC: 0.96 and 0.94, respectively), and to a lesser extent, Gleason pattern 5 (ICC: 0.82). Similar results were obtained using two different scanners.CONCLUSIONS: Our AI-based algorithm can reliably detect prostate cancer and quantify the Gleason patterns in core needle biopsies, with similar accuracy as pathologists. The results are reproducible on images from different scanners with a proven low level of intraobserver variability. We believe that this AI tool could be regarded as an efficient and interactive tool for pathologists.PATIENT SUMMARY: We developed a sensitive artificial intelligence tool for prostate biopsies, which detects and grades cancer with similar accuracy to pathologists. This tool holds promise to improve the diagnosis of prostate cancer.

U2 - 10.1016/j.euf.2020.11.001

DO - 10.1016/j.euf.2020.11.001

M3 - Article

C2 - 33303404

SN - 2405-4569

VL - 7

SP - 995

EP - 1001

JO - European Urology Focus

JF - European Urology Focus

IS - 5

ER -

An Artificial Intelligence-based Support Tool for Automation and Standardisation of Gleason Grading in Prostate Biopsies

Abstract

Subject classification (UKÄ)

UN SDGs

Access to Document

Fingerprint

Research output

Prostate Cancer Tissue Biomarkers

Cite this