Abstract: Ultrasound (US) modules and methods in US image-guided systems for thermal ablation are provided, which utilize regular two-dimensional (2D) B mode US images to evaluate in real-time the damage achieved by the thermal ablation, including the prediction of tissue damage immediately after completion of the thermal ablation procedure, as well as the expected damage after 24 hours. Three-dimensional (3D) biotrace map (BTM) representation(s) may be constructed from the received US images by applying deep neural networks (DNN) to segment tissue damage in US frames, and present the damage to the user in 3D and/or in virtual sections through the BTM representation(s). Using regular 2D US probes allows much flexibility in operating the US probe and imaging the target successfully, while the construction of the 3D BTM representation(s) enables accumulating, updating and analyzing the volume data to provide a full and updating representation of the ablation procedure.

Abstract: Systems and methods for training deep neural networks (DNNs) for blood vessel and lesion segmentation using synthetically generated training data, are provided. Systems may comprise parametric simulation modules for generating 3D branching blood vessels and 3D lesion structures and an augmentation module configured to add noise, background and organ boundaries to the 3D models, to yield the synthetic training data for training the DNNs. The 3D vessel model may be generated as a hierarchical tree comprising segments that are generated as anti-aliased lines with specified length and start and end thicknesses, which are elongated by segment addition(s) and/or by branching to follow semi-linear or curved lines, while avoiding overlapping of segments. The 3D lesion model and/or combined vessel/lesion models may be generated using multiple input images such as contrast enhancement phases.

Abstract: Systems, displays and methods are provided, for performing ultrasound image registration and for using ultrasound images to guide thermal ablation. Registration is carried out by correlating sequential ultrasound images, identifying key frames from the correlation values, identifying periodic change(s) corresponding to breathing and heart beating, and correlating pixels in sequential key frames that have a same phase with respect to the identifying periodic change(s). Based on the registration, the start of ablation is detected, bubbles formed in the ablation procedure are identified and their movements are followed—all using B-mode ultrasound images only. Using the identified bubbles, the thermally damaged tissue region is demarcated and provided in real-time—at an accuracy similar to prior art post-ablation results. Disclosed systems and methods therefore provide the physician with an ultrasound-based way of monitoring thermal ablation in real-time.