Enhancing radiomics and Deep Learning systems through the standardization of medical imaging workflows.

Journal: Scientific data
PMID:

Abstract

Recent advances in computer-aided diagnosis, treatment response and prognosis in radiomics and deep learning challenge radiology with requirements for world-wide methodological standards for labeling, preprocessing and image acquisition protocols. The adoption of these standards in the clinical workflows is a necessary step towards generalization and interoperability of radiomics and artificial intelligence algorithms in medical imaging.

Authors

  • Miriam Cobo
    Advanced Computing and e-Science Research Group, Institute of Physics of Cantabria (IFCA), CSIC - UC, 39005, Santander, Cantabria, Spain. cobocano@ifca.unican.es.
  • Pablo Menéndez Fernández-Miranda
    Servicio de Radiodiagnóstico, Hospital Universitario "Marqués de Valdecilla", Santander, Spain.
  • Gorka Bastarrika
    Clínica Universidad de Navarra, Pamplona, Spain.
  • Lara Lloret Iglesias
    Advanced Computation and e-Science, Instituto de Física de Cantabria (IFCA), Consejo Superior de Investigaciones Científicas (CSIC), Santander, Spain.