Generation of 3D ultrasound training volumes from freehand acquired data

We are developing a low cost ultrasound training system running on a laptop in which the user scans a generic 3D curved surface representing the patient using a 5 DoF sensor. A critical component of this system is the generation of ultrasound training image volumes, which need to cover a complete body region in order to provide a realistic scanning experience. This research attempts to develop stitching techniques to generate large global volumes from smaller overlapping volumes acquired using freehand techniques.