Artificial intelligence compared to radiologists for the initial diagnosis of prostate cancer on magnetic resonance imaging: A systematic review and recommendations for future studies

Tom Syer; Pritesh Mehta; Michela Antonelli; Sue Mallett; David Atkinson; Sébastien Ourselin; Shonit Punwani

doi:10.3390/cancers13133318

Artificial intelligence compared to radiologists for the initial diagnosis of prostate cancer on magnetic resonance imaging: A systematic review and recommendations for future studies

Tom Syer, Pritesh Mehta, Michela Antonelli, Sue Mallett, David Atkinson, Sébastien Ourselin, Shonit Punwani^*

^*Corresponding author for this work

Research output: Contribution to journal › Review article › peer-review

41 Citations (Scopus)

Abstract

Computer-aided diagnosis (CAD) of prostate cancer on multiparametric magnetic resonance imaging (mpMRI), using artificial intelligence (AI), may reduce missed cancers and unnecessary biopsies, increase inter-observer agreement between radiologists, and alleviate pressures caused by rising case incidence and a shortage of specialist radiologists to read prostate mpMRI. However, well-designed evaluation studies are required to prove efficacy above current clinical practice. A systematic search of the MEDLINE, EMBASE, and arXiv electronic databases was conducted for studies that compared CAD for prostate cancer detection or classification on MRI against radiologist interpretation and a histopathological reference standard, in treatment-naïve men with a clinical suspicion of prostate cancer. Twenty-seven studies were included in the final analysis. Due to substantial heterogeneities in the included studies, a narrative synthesis is presented. Several studies reported superior diagnostic accuracy for CAD over radiologist interpretation on small, internal patient datasets, though this was not observed in the few studies that performed evaluation using external patient data. Our review found insufficient evidence to suggest the clinical deployment of artificial intelligence algorithms at present. Further work is needed to develop and enforce method-ological standards, promote access to large diverse datasets, and conduct prospective evaluations before clinical adoption can be considered.

Original language	English
Article number	3318
Journal	Cancers
Volume	13
Issue number	13
DOIs	https://doi.org/10.3390/cancers13133318
Publication status	Published - 1 Jul 2021

Keywords

Artificial intelligence
Computer-aided diagnosis
Deep learning
Machine learning
Magnetic resonance imaging
PRISMA-DTA
Prostate cancer
QUADAS-2
Systematic review

UN SDGs

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

Access to Document

10.3390/cancers13133318

Cite this

@article{79060baade584e658a07de0f3c1e9ba8,

title = "Artificial intelligence compared to radiologists for the initial diagnosis of prostate cancer on magnetic resonance imaging: A systematic review and recommendations for future studies",

abstract = "Computer-aided diagnosis (CAD) of prostate cancer on multiparametric magnetic resonance imaging (mpMRI), using artificial intelligence (AI), may reduce missed cancers and unnecessary biopsies, increase inter-observer agreement between radiologists, and alleviate pressures caused by rising case incidence and a shortage of specialist radiologists to read prostate mpMRI. However, well-designed evaluation studies are required to prove efficacy above current clinical practice. A systematic search of the MEDLINE, EMBASE, and arXiv electronic databases was conducted for studies that compared CAD for prostate cancer detection or classification on MRI against radiologist interpretation and a histopathological reference standard, in treatment-na{\"i}ve men with a clinical suspicion of prostate cancer. Twenty-seven studies were included in the final analysis. Due to substantial heterogeneities in the included studies, a narrative synthesis is presented. Several studies reported superior diagnostic accuracy for CAD over radiologist interpretation on small, internal patient datasets, though this was not observed in the few studies that performed evaluation using external patient data. Our review found insufficient evidence to suggest the clinical deployment of artificial intelligence algorithms at present. Further work is needed to develop and enforce method-ological standards, promote access to large diverse datasets, and conduct prospective evaluations before clinical adoption can be considered.",

keywords = "Artificial intelligence, Computer-aided diagnosis, Deep learning, Machine learning, Magnetic resonance imaging, PRISMA-DTA, Prostate cancer, QUADAS-2, Systematic review",

author = "Tom Syer and Pritesh Mehta and Michela Antonelli and Sue Mallett and David Atkinson and S{\'e}bastien Ourselin and Shonit Punwani",

note = "Funding Information: Conflicts of Interest: S.M. receives funding from National Cancer Imaging Translation Accelerator (NCITA), National Institute for Health Research (NIHR) and the University College London/ University College London Hospital (UCL/UCLH) Biomedical Research Centre. D.A. receives research support from Philips and Siemens. S.P. received funding from the University College London/ University College London Hospital (UCL/UCLH) Biomedical Research Centre. Publisher Copyright: {\textcopyright} 2021 by the authors. Licensee MDPI, Basel, Switzerland. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.",

year = "2021",

month = jul,

day = "1",

doi = "10.3390/cancers13133318",

language = "English",

volume = "13",

journal = "Cancers",

issn = "2072-6694",

publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",

number = "13",

}

TY - JOUR

T1 - Artificial intelligence compared to radiologists for the initial diagnosis of prostate cancer on magnetic resonance imaging

T2 - A systematic review and recommendations for future studies

AU - Syer, Tom

AU - Mehta, Pritesh

AU - Antonelli, Michela

AU - Mallett, Sue

AU - Atkinson, David

AU - Ourselin, Sébastien

AU - Punwani, Shonit

N1 - Funding Information: Conflicts of Interest: S.M. receives funding from National Cancer Imaging Translation Accelerator (NCITA), National Institute for Health Research (NIHR) and the University College London/ University College London Hospital (UCL/UCLH) Biomedical Research Centre. D.A. receives research support from Philips and Siemens. S.P. received funding from the University College London/ University College London Hospital (UCL/UCLH) Biomedical Research Centre. Publisher Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.

PY - 2021/7/1

Y1 - 2021/7/1

N2 - Computer-aided diagnosis (CAD) of prostate cancer on multiparametric magnetic resonance imaging (mpMRI), using artificial intelligence (AI), may reduce missed cancers and unnecessary biopsies, increase inter-observer agreement between radiologists, and alleviate pressures caused by rising case incidence and a shortage of specialist radiologists to read prostate mpMRI. However, well-designed evaluation studies are required to prove efficacy above current clinical practice. A systematic search of the MEDLINE, EMBASE, and arXiv electronic databases was conducted for studies that compared CAD for prostate cancer detection or classification on MRI against radiologist interpretation and a histopathological reference standard, in treatment-naïve men with a clinical suspicion of prostate cancer. Twenty-seven studies were included in the final analysis. Due to substantial heterogeneities in the included studies, a narrative synthesis is presented. Several studies reported superior diagnostic accuracy for CAD over radiologist interpretation on small, internal patient datasets, though this was not observed in the few studies that performed evaluation using external patient data. Our review found insufficient evidence to suggest the clinical deployment of artificial intelligence algorithms at present. Further work is needed to develop and enforce method-ological standards, promote access to large diverse datasets, and conduct prospective evaluations before clinical adoption can be considered.

AB - Computer-aided diagnosis (CAD) of prostate cancer on multiparametric magnetic resonance imaging (mpMRI), using artificial intelligence (AI), may reduce missed cancers and unnecessary biopsies, increase inter-observer agreement between radiologists, and alleviate pressures caused by rising case incidence and a shortage of specialist radiologists to read prostate mpMRI. However, well-designed evaluation studies are required to prove efficacy above current clinical practice. A systematic search of the MEDLINE, EMBASE, and arXiv electronic databases was conducted for studies that compared CAD for prostate cancer detection or classification on MRI against radiologist interpretation and a histopathological reference standard, in treatment-naïve men with a clinical suspicion of prostate cancer. Twenty-seven studies were included in the final analysis. Due to substantial heterogeneities in the included studies, a narrative synthesis is presented. Several studies reported superior diagnostic accuracy for CAD over radiologist interpretation on small, internal patient datasets, though this was not observed in the few studies that performed evaluation using external patient data. Our review found insufficient evidence to suggest the clinical deployment of artificial intelligence algorithms at present. Further work is needed to develop and enforce method-ological standards, promote access to large diverse datasets, and conduct prospective evaluations before clinical adoption can be considered.

KW - Artificial intelligence

KW - Computer-aided diagnosis

KW - Deep learning

KW - Machine learning

KW - Magnetic resonance imaging

KW - PRISMA-DTA

KW - Prostate cancer

KW - QUADAS-2

KW - Systematic review

UR - http://www.scopus.com/inward/record.url?scp=85108889389&partnerID=8YFLogxK

U2 - 10.3390/cancers13133318

DO - 10.3390/cancers13133318

M3 - Review article

AN - SCOPUS:85108889389

SN - 2072-6694

VL - 13

JO - Cancers

JF - Cancers

IS - 13

M1 - 3318

ER -

Artificial intelligence compared to radiologists for the initial diagnosis of prostate cancer on magnetic resonance imaging: A systematic review and recommendations for future studies

Abstract

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this