MICCAI 2016 Daily - Tuesday

Sergio: These parameters have shown in the past that they are sensitive to disease such as cancer. We believe that our technology can be used to improve cancer screening and differentiation, among other applications. It provides quantitative imaging biomarkers for differentiation of disease. The technology only requires minor hardware additions (a passive reflector) to conventional ultrasound systems. Finally, it does not require compression of the breast, which minimizes patient’s discomfort. MICCAI Daily: Why is this challenging to do? Sergio: Using the hardware in the hand-held setup, the amount of data that we acquire is smaller than with dedicated ultrasound tomography systems. Therefore, the reconstruction is much more challenging than with a conventional tomographic system. MICCAI Daily: What image processing techniques enabled you to solve this challenge? Sergio: We have introduced non- linear optimization methodologies to be able to perform this reconstruction robustly. There are two main aspects of our contribution. One is to find robust ways to track a reflector. The second is to develop novel regularization approaches based on total variation to achieve a robust reconstruction. MICCAI Daily: If you could add one feature to your model that it doesn’t have today, what would it be? Sergio: There are two aspects that I would add. One is to make it a real time solution. At the moment, we are still not there, but we are working very quickly towards this direction. Secondly, we have been successful with phantom and ex- vivo tests, and have just started our clinical tests. Preliminary clinical results show the future potential, however, we observe new challenges that will be addressed in the next steps. Stay tuned! Our Pick for Tuesday 5 MICCAI Daily: Tuesday “ We have introduced non-linear optimization methodologies to be able to perform this reconstruction robustly ”