[Aid of artificial intelligence for the anatomo-pathological diagnosis of tumours].

Journal: Revue medicale de Liege
Published Date:

Abstract

The anatomo-pathological diagnosis of tumors is based on many criteria related mainly to image analysis. Currently, in most pathology laboratories, tissues or cells are placed on glass slides and directly analyzed with an optical microscope. Because of technological evolutions, it is currently possible to digitize slides (digital pathology). The digitization of whole slides has allowed the development of computer programs of artificial intelligence (AI) for image analysis. Applied to tumour pathology, this technology allows the detection, diagnosis or evaluation of the prognosis of neoplastic lesions. There are many challenges associated with the use of AI in routine pathology. These are mainly related to the amount of data to be analyzed and to the development of reliable algorithms. Nevertheless, this technology is promising and could become a valuable aid in the field of precision medicine for which the amount of data related to a patient is constantly increasing.

Authors

  • M Reginster
    Service d'Anatomie pathologique, CHU Liège, Belgique.
  • S de Froidmont
    Service d'Anatomie pathologique, CHU Liège, Belgique.
  • J Somja
    Service d'Anatomie pathologique, CHU Liège, Belgique.
  • P Delvenne
    Service d'Anatomie pathologique, CHU Liège, Belgique.