Breast surface reconstruction utilising autonomous robotic assisted ultrasound image acquisition

Abstract

One in eight females will be diagnosed with breast cancer in their lifetime, making it the most diagnosed cancer globally [1]. Three phases are essential for the outcome of breast cancer; early detection, accurate diag- nosis and treatment. Magnetic resonance imaging (MRI) has proven to be highly sensitive in detecting possible tu- mors lesions compared to other image modalities. How- ever, ultrasound (US) guided biopsies are the standard and biopsies on MRI detected lesions is challenging, because these lesions may not be visible on US images. Registration of pre-operative MRI with the intra- operative US combines the benefits of both imaging modalities and subsequently improve tumor localisation. Therefore, this work presents an US image-based sur- face reconstruction based on autonomous acquired US images to increase the accuracy of MRI/US registration. Thereby the proposed work eliminates the need for inter- sensor calibration as would be needed if stereo camera- based, depth camera-based or marker-based breast sur- face reconstruction would be used.