You are here : Home > Scanner: improve medical diagnosis with algorithmic models

News | Focus

Scanner: improve medical diagnosis with algorithmic models

Published on 6 April 2021
  • Imaging has already been revolutionizing medical diagnosis for over four decades. CEA-Leti, which produced and installed the first French scanner in 1976, has been working tirelessly to develop equipment offering increasingly advanced performances.

  • The CT scan is an imaging technique frequently used for diagnoses. It measures the absorption of the X-rays by the tissues through which they pass. Algorithmic models then digitize and reconstruct 3D images of the organs.

  • For an even more accurate, rapid image, researchers at CEA-Leti are improving image reconstruction by introducing an algorithmic model that takes the spectral data into account effectively. 

Here, we have worked on the combination of a physical model and an optimization algorithm making it possible to reduce the image artifacts and noise, states Clarisse Fournier, research engineer at CEA-Leti

​                                  State of the art                                                                       CEA-Leti solution

Caption : Improvement of the image vs. state of the art

  • One of the challenges in developing this method was to offer an algorithmic solution at a reasonable computational cost. The method has been validated both on simulated data and experimental data acquired with a CEA-Leti detector.
  • Work is in progress to enable this model to be industrialized. 

Top page