Generative adversarial networks (GAN)-based data augmentation of rare liver cancers: The SFR 2021 Artificial Intelligence Data Challenge - 04/01/23
, Littisha Lawrance c, Younes Belkouchi c, d, Valérie Vilgrain e, f, Maité Lewin g, h, Hervé Trillaud i, Christine Hoeffel j, k, Valérie Laurent l, Samy Ammari c, m, Eric Morand n, Orphée Faucoz n, Arthur Tenenhaus o, Anne Cotten p, q, Jean-François Meder r, s, Hugues Talbot d, Alain Luciani a, b, Nathalie Lassau c, m
Bienvenue sur EM-consulte, la référence des professionnels de santé.
Article gratuit.
Connectez-vous pour en bénéficier!
Highlights |
• | The SFR 2021 Artificial Intelligence Data Challenge focused on generative adversarial network (GAN)-based data augmentation techniques. |
• | The aim was to create a synthetic dataset of 1000 macrotrabecular-massive hepatocellular carcinomas, which is a rare and aggressive subtype of hepatocellular carcinoma. |
• | This data challenge demonstrates the ability of generative adversarial networks techniques to generate a large number of images from a small sample of rare malignant tumor MRI examinations. |
Abstract |
Purpose |
The 2021 edition of the Artificial Intelligence Data Challenge was organized by the French Society of Radiology together with the Centre National d’Études Spatiales and CentraleSupélec with the aim to implement generative adversarial networks (GANs) techniques to provide 1000 magnetic resonance imaging (MRI) cases of macrotrabecular-massive (MTM) hepatocellular carcinoma (HCC), a rare and aggressive subtype of HCC, generated from a limited number of real cases from multiple French centers.
Materials and methods |
A dedicated platform was used by the seven inclusion centers to securely upload their anonymized MRI examinations including all three cross-sectional images (one late arterial and one portal-venous phase T1-weighted images and one fat-saturated T2-weighted image) in compliance with general data protection regulation. The quality of the database was checked by experts and manual delineation of the lesions was performed by the expert radiologists involved in each center. Multidisciplinary teams competed between October 11th, 2021 and February 13th, 2022.
Results |
A total of 91 MTM-HCC datasets of three images each were collected from seven French academic centers. Six teams with a total of 28 individuals participated in this challenge. Each participating team was asked to generate one thousand 3-image cases. The qualitative evaluation was performed by three radiologists using the Likert scale on ten randomly selected cases generated by each participant. A quantitative evaluation was also performed using two metrics, the Frechet inception distance and a leave-one-out accuracy of a 1-Nearest Neighbor algorithm.
Conclusion |
This data challenge demonstrates the ability of GANs techniques to generate a large number of images from a small sample of imaging examinations of a rare malignant tumor.
Le texte complet de cet article est disponible en PDF.Keywords : Artificial intelligence, Deep learning, Generative adversarial networks, Liver cancer, Magnetic resonance imaging
Abbreviations : AI, CNES, CNN, DL, FID, GANs, GDPR, HCC, MRI, MTM, SFR
Plan
Vol 104 - N° 1
P. 43-48 - janvier 2023 Retour au numéro
Article précédent
- Artificial intelligence in adrenal imaging: A critical review of current applications
- Maxime Barat, Martin Gaillard, Anne-Ségolène Cottereau, Elliot K. Fishman, Guillaume Assié, Anne Jouinot, Christine Hoeffel, Philippe Soyer, Anthony Dohan
Bienvenue sur EM-consulte, la référence des professionnels de santé.