SANDMED: Shape ANalysis for the Diagnosis of METabolic Disorders

Data inizio
17 aprile 2010
Durata (mesi) 
Responsabili (o referenti locali)
Giachetti Andrea

The SANDMED project aims at using the information acquired by whole body scanners, coupled, if possible, with other non-invasive acquisitions, to obtain relevant diagnostic parameters related to obesity and metabolic abnormalities (e.g., hyperinsulinaemia, hyperuricaemia, glucose intolerance, hypertension).
The basic idea of the project is to apply and customize recently developed shape analysis methods (e.g. PCA/LDA, spectral analysis, salient point detection and characterisation, skeleton-based segmentation and modelling) on a large set of acquired body scans to derive several compact parametric representations of the global and local variability of human body anatomy, testing the correlation between the parameters obtained and other diagnostic indicators.
In this way it will be possible to derive indices that are more effective than those commonly used now (e.g., BMI, W/H ratio), showing only a limited ability to predict medium- to long-term consequences of excess fat.
The project aims also at improving the acquisition protocols related to 3D whole body scanning and develop an interactive tool where not only high resolution surface models can be displayed and interactively analyzed, but they can also be enriched with computed indices and additional volumetric information coming from other modalities (e.g. US, DXA, MRI).

Enti finanziatori:

Finanziamento: richiesto
Programma: EUROPA - Progetti Europei

Partecipanti al progetto

Umberto Castellani
Professore associato
Andrea Giachetti
Professore associato
Paolo Moghetti
Professore associato
Carlo Zancanaro
Professore ordinario
Aree di ricerca coinvolte dal progetto
Sistemi intelligenti
Computing methodologies - Computer graphics