Breast Projects

In recent years, we have focused many resources on projects dedicated to the breast. We first carried out a feasibility study thanks to a doctoral thesis financed within the S.H.A.R.M. project [1]. The thesis was intended to investigate the feasibility of a software that supported breast radiologists in reading multimodal 3D breast images.

The positive results of the study encouraged us to continue the work. We have therefore dedicated several resources on two projects with the aim of creating a platform that supports the radiologist in the detection, diagnosis and monitoring of breast cancer. One project is dedicated to the development of algorithms and tools for the analysis and comparison of multimodal breast images, while the second is focused on the synchronization of breast tomosynthesis images.

Both in clinical and in screening, the radiologist must analyze and compare a large number of images acquired at different times, different projections and with different imaging methods. The comparison is subject to the experience of the doctor, and it takes time, as the radiologist has to take into account the deformation of the breast among different acquisitions. This procedure must be repeated every time he wants to retrospectively analyze the patient’s images.

Both projects aim to develop a software platform that supports the radiologist in the interpretation of 3D images. The platform displays images with optimum viewing techniques and provides tools for automatic synchronization of positions between different breast images, in order to facilitate simultaneous navigation. This is made possible by the use of innovative registration algorithms that take into account non-rigid deformations of the structures.

The two projects were co-financed by the Friuli Venezia Giulia region under two funding programs [2] [3]. Both projects were developed in collaboration with Dipartimento ad Attività Integrata di Diagnostica per Immagini dell’Azienda Sanitaria Universitaria Integrata di Trieste, DataMind S.r.l. and G-Squared S.r.l..

Thanks to the knowledge acquired during the projects, we studied methods for the automatic contouring of lesions on breast MRI images. The research centers involved in this project were Azienda Ospedaliera Universitaria Integrata di Verona, Politecnico di Torino and Università degli Studi di Trieste – Dipartimento di Fisica.

[1] PhD Thesis “Automated Deformable Registration of Breast Images: towards a software-assisted multimodal breast image reading” – S.H.A.R.M. (Supporting Human Assets in Research and Mobility) project – European Social Fund 2007/2013
[2] “Unità multitecnologica per la rivelazione, la diagnosi e il monitoraggio del tumore alla mammella” – PAR FSC 2007-2013 FVG.
[3] “Piattaforma per l’analisi e comparazione delle immagini di tomosintesi della mammella” – POR FESR 2014-2020 FVG.

progetti-breast-image

Negli ultimi anni abbiamo investito molto tempo e risorse in progetti dedicati alla mammella. Abbiamo dapprima effettuato uno studio di fattibilità grazie a una tesi di dottorato finanziata all’interno del progetto S.H.A.R.M. [1]. La tesi aveva lo scopo di investigare la fattibilità di un software che supportasse i senologi nella lettura delle immagini 3D multimodali del seno.

I risultati positivi dello studio ci hanno incoraggiato a proseguire il lavoro. Abbiamo quindi dedicato diverse risorse su due progetti con lo scopo di realizzare una piattaforma che supporti il radiologo senologo nella fase di rivelazione, diagnosi e monitoraggio del tumore alla mammella. Un progetto è dedicato allo sviluppo di algoritmi e strumenti per l’analisi e la comparazione di immagini multimodali della mammella, mentre il secondo è focalizzato alla sincronizzazione di immagini di tomosintesi della mammella.

Both in clinical and in screening, the radiologist must analyze and compare a large number of images acquired at different times, different projections and with different imaging methods. The comparison is subject to the experience of the doctor, and it takes time, as the radiologist has to take into account the deformation of the breast among different acquisitions. This procedure must be repeated every time he wants to retrospectively analyze the patient’s images.

Both projects aim to develop a software platform that supports the radiologist in the interpretation of 3D images. The platform displays images with optimum viewing techniques and provides tools for automatic synchronization of positions between different breast images, in order to facilitate simultaneous navigation. This is made possible by the use of innovative registration algorithms that take into account non-rigid deformations of the structures.

The two projects were co-financed by the Friuli Venezia Giulia region under two funding programs [2] [3]. Both projects were developed in collaboration with Dipartimento ad Attività Integrata di Diagnostica per Immagini dell’Azienda Sanitaria Universitaria Integrata di Trieste, DataMind S.r.l. and G-Squared S.r.l..

Thanks to the knowledge acquired during the projects, we studied methods for the automatic contouring of lesions on breast MRI images. The research centers involved in this project were Azienda Ospedaliera Universitaria Integrata di Verona, Politecnico di Torino and Università degli Studi di Trieste – Dipartimento di Fisica.

[1] PhD Thesis “Automated Deformable Registration of Breast Images: towards a software-assisted multimodal breast image reading” – S.H.A.R.M. (Supporting Human Assets in Research and Mobility) project – European Social Fund 2007/2013
[2] “Unità multitecnologica per la rivelazione, la diagnosi e il monitoraggio del tumore alla mammella” – PAR FSC 2007-2013 FVG.
[3] “Piattaforma per l’analisi e comparazione delle immagini di tomosintesi della mammella” – POR FESR 2014-2020 FVG.

MORE

GET MORE DETAILS

DETAILS

Hai bisogno di informazioni?

Start typing and press Enter to search