TY - JOUR
T1 - Explainable artificial intelligence (XAI) in radiology and nuclear medicine
T2 - a literature review
AU - de Vries, Bart M
AU - Zwezerijnen, Gerben J C
AU - Burchell, George L
AU - van Velden, Floris H P
AU - Menke-van der Houven van Oordt, Catharina Willemien
AU - Boellaard, Ronald
N1 - Publisher Copyright: Copyright © 2023 de Vries, Zwezerijnen, Burchell, van Velden, Menke-van der Houven van Oordt and Boellaard.
PY - 2023
Y1 - 2023
N2 - RATIONAL: Deep learning (DL) has demonstrated a remarkable performance in diagnostic imaging for various diseases and modalities and therefore has a high potential to be used as a clinical tool. However, current practice shows low deployment of these algorithms in clinical practice, because DL algorithms lack transparency and trust due to their underlying black-box mechanism. For successful employment, explainable artificial intelligence (XAI) could be introduced to close the gap between the medical professionals and the DL algorithms. In this literature review, XAI methods available for magnetic resonance (MR), computed tomography (CT), and positron emission tomography (PET) imaging are discussed and future suggestions are made.METHODS: PubMed, Embase.com and Clarivate Analytics/Web of Science Core Collection were screened. Articles were considered eligible for inclusion if XAI was used (and well described) to describe the behavior of a DL model used in MR, CT and PET imaging.RESULTS: A total of 75 articles were included of which 54 and 17 articles described post and ad hoc XAI methods, respectively, and 4 articles described both XAI methods. Major variations in performance is seen between the methods. Overall, post hoc XAI lacks the ability to provide class-discriminative and target-specific explanation. Ad hoc XAI seems to tackle this because of its intrinsic ability to explain. However, quality control of the XAI methods is rarely applied and therefore systematic comparison between the methods is difficult. CONCLUSION: There is currently no clear consensus on how XAI should be deployed in order to close the gap between medical professionals and DL algorithms for clinical implementation. We advocate for systematic technical and clinical quality assessment of XAI methods. Also, to ensure end-to-end unbiased and safe integration of XAI in clinical workflow, (anatomical) data minimization and quality control methods should be included.
AB - RATIONAL: Deep learning (DL) has demonstrated a remarkable performance in diagnostic imaging for various diseases and modalities and therefore has a high potential to be used as a clinical tool. However, current practice shows low deployment of these algorithms in clinical practice, because DL algorithms lack transparency and trust due to their underlying black-box mechanism. For successful employment, explainable artificial intelligence (XAI) could be introduced to close the gap between the medical professionals and the DL algorithms. In this literature review, XAI methods available for magnetic resonance (MR), computed tomography (CT), and positron emission tomography (PET) imaging are discussed and future suggestions are made.METHODS: PubMed, Embase.com and Clarivate Analytics/Web of Science Core Collection were screened. Articles were considered eligible for inclusion if XAI was used (and well described) to describe the behavior of a DL model used in MR, CT and PET imaging.RESULTS: A total of 75 articles were included of which 54 and 17 articles described post and ad hoc XAI methods, respectively, and 4 articles described both XAI methods. Major variations in performance is seen between the methods. Overall, post hoc XAI lacks the ability to provide class-discriminative and target-specific explanation. Ad hoc XAI seems to tackle this because of its intrinsic ability to explain. However, quality control of the XAI methods is rarely applied and therefore systematic comparison between the methods is difficult. CONCLUSION: There is currently no clear consensus on how XAI should be deployed in order to close the gap between medical professionals and DL algorithms for clinical implementation. We advocate for systematic technical and clinical quality assessment of XAI methods. Also, to ensure end-to-end unbiased and safe integration of XAI in clinical workflow, (anatomical) data minimization and quality control methods should be included.
KW - computed tomography (CT) imaging
KW - deep learning
KW - explainable artificial intelligence
KW - magnetic resonance (MR) imaging
KW - positron emission tomography (PET) imaging
UR - http://www.scopus.com/inward/record.url?scp=85160047993&partnerID=8YFLogxK
U2 - https://doi.org/10.3389/fmed.2023.1180773
DO - https://doi.org/10.3389/fmed.2023.1180773
M3 - Review article
C2 - 37250654
SN - 2296-858X
VL - 10
SP - 1180773
JO - Frontiers in Medicine
JF - Frontiers in Medicine
M1 - 1180773
ER -