Home Research
Home Research

Quantitative annotation of medical images

Research Area: Medical Image Analysis
Status: In progress  
Faculty: Alessandro Verri Participants: Curzio Basso, Gabriele Chiusano, Matteo Santoro

Our effort focuses on the automatic annotation of specific tissues and structures in biomedical images, followed by identification and quantification of their clinically relevant features. Our work so far concentrated on the specific case of automatic estimation of the volume of inflamed synovia from MR images but we are broadening our interests to include other imaging modalities like US and CT.

In order to reach this ambitious objective we pursue research on:

  • learning to segment structures of interest characterized by complex morphologies
  • integrating multiple modalities to increase the accuracy of quantitative estimates (also in the presence of dynamical data)
  • direct quantification via regression (i.e. bypassing explicit annotation steps) of the quantitative parameters relevant for the specific application/p>