Artificial intelligence in musculoskeletal oncology imaging: A critical review of current applications - 04/01/23

Doi : 10.1016/j.diii.2022.10.004

Maxime Lacroix ^a,^b,^c,^⁎, Theodore Aouad ^d, Jean Feydy ^e, David Biau ^b,^f, Frédérique Larousserie ^b,^g, Laure Fournier ^a,^b,^c, Antoine Feydy ^b,^h
^a Department of Radiology, Hôpital Européen Georges Pompidou, Assistance Publique-Hôpitaux de Paris, Paris, 75015, France
^b Université Paris Cité, Faculté de Médecine, Paris, 75006, France
^c PARCC UMRS 970, INSERM, Paris 75015, France
^d Université Paris-Saclay, CentraleSupélec, Inria, Centre for Visual Computing, 91190, Gif-sur-Yvette, France
^e Université Paris Cité, HeKA team, Inria Paris, Inserm, 75006, Paris, France
^f Department of Orthopedic Surgery, Hôpital Cochin, Assistance Publique-Hôpitaux de Paris, Paris, 75014, France
^g Department of Pathology, Hôpital Cochin, Assistance Publique-Hôpitaux de Paris, Paris, 75014, France
^h Department of Radiology, Hôpital Cochin, Assistance Publique-Hôpitaux de Paris, Paris, 75014, France

^⁎Corresponding author: maxime.lacroix@aphp.fr

connectez-vous ou créez un compte

Bienvenue sur EM-consulte, la référence des professionnels de santé.
Article gratuit.

Connectez-vous pour en bénéficier!

Highlights

•	Machine learning, deep learning and radiomic models can be applied to musculoskeletal oncology imaging.
•	One limitation of radiomic studies is the lack of assessement of robustness of extracted features.
•	Further efforts are needed for a routine use of artificial intelligence in musculoskeletal oncology imaging.

Le texte complet de cet article est disponible en PDF.

Abstract

Artificial intelligence (AI) is increasingly being studied in musculoskeletal oncology imaging. AI has been applied to both primary and secondary bone tumors and assessed for various predictive tasks that include detection, segmentation, classification, and prognosis. Still, in the field of clinical research, further efforts are needed to improve AI reproducibility and reach an acceptable level of evidence in musculoskeletal oncology. This review describes the basic principles of the most common AI techniques, including machine learning, deep learning and radiomics. Then, recent developments and current results of AI in the field of musculoskeletal oncology are presented. Finally, limitations and future perspectives of AI in this field are discussed.

Le texte complet de cet article est disponible en PDF.

Keywords : Artificial intelligence, Bone tumor, Deep learning, Machine learning, Metastases

Abbreviations : ADC, AI, AUC, CNN, CT, DL, DSC, FDG, PET/CT, ML, MRI, PET-CT, RF, ROC, SPECT, SVM

Plan

Introduction

Basic principles and process

Why do we do “learning”?

Overfitting and generalization

Conventional statistics

Overcoming the “curse of dimensionality” with machine learning (ML)

Processing markers into hand-crafted features with radiomics

Learning task-specific features with deep learning

Statistical analysis and regularization

Applying machine learning for segmentation

Current applications

Primary bone lesions

Secondary bone lesions

Main limitations of AI

CRediT authorship contribution statement

Export

Vol 104 - N° 1

P. 18-23 - janvier 2023 Retour au numéro

Article précédent

Artificial intelligence: A critical review of applications for lung nodule and lung cancer
Constance de Margerie-Mellon, Guillaume Chassagnon

| Article suivant

Artificial intelligence: A review of current applications in hepatocellular carcinoma imaging
Anna Pellat, Maxime Barat, Romain Coriat, Philippe Soyer, Anthony Dohan

Bienvenue sur EM-consulte, la référence des professionnels de santé.

connectez-vous ou créez un compte

Artificial intelligence in musculoskeletal oncology imaging: A critical review of current applications - 04/01/23

Highlights

Abstract

Plan

Export citations

Fichier

Contenu

Accès rapides

Mon compte

Aide & support

Plateformes Elsevier Masson

Déclaration CNIL