Artificial neuroradiology: Between human and artificial networks of neurons? - IMT Atlantique Accéder directement au contenu
Article Dans Une Revue Journal de Neuroradiologie / Journal of Neuroradiology Année : 2019

Artificial neuroradiology: Between human and artificial networks of neurons?

Résumé

Increasingly widespread application of advanced image processing and Artificial Intelligence (AI) in the field of neuroradiology highlights changing trends in the availability of emerging technologies. In the past 10 years, publications on AI in radiology have increased from 100-150 to 700-800 per year and neuroradiology appears as the most involved subspecialty, accounting for about one-third of articles. But it should be noted that the application of lower levels of AI through medical image processing has been integrated since the birth of our discipline. The convergence of better technological performance and higher volume of data to process has favored the development of more advanced processes, such as machine learning (ML). AI can be attributed to any machine performing a task normally claiming human cognition. ML is a type of AI that allows computers to learn from data without explicit programming (not by programming it for a specific domain, but by designing a system that can learn from several examples to solve a problem, such as classification algorithms: clustering, support vector machine.. .). ML-based algorithms may differ depending on the approach, the type of data, and the task. Supervised and unsupervised learning are part of this. In this latter approach neither criteria nor ground truth are used to train the algorithm. Deep learning is therefore a supervised machine learning method that uses a specific architecture, mainly a form of neural network to automatically extract relevant features. These neural networks are inspired by the structure of the brain.

Dates et versions

hal-02285990 , version 1 (13-09-2019)

Identifiants

Citer

Arnaud Attye, Julien Ognard, François Rousseau, Douraied Ben Salem. Artificial neuroradiology: Between human and artificial networks of neurons?. Journal de Neuroradiologie / Journal of Neuroradiology, 2019, 46 (5), pp.279-280. ⟨10.1016/j.neurad.2019.07.001⟩. ⟨hal-02285990⟩
47 Consultations
1 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More