Ultrasound (US) imaging is widely employed for diagnosis and staging of vascular diseases, mainly due to its high availability and the fact it does not emit ionizing radiation. However, high interoperator variability limits the repeatability of US image acquisition. To address this challenge, we propose an end-to-end workflow for automatic robotic US screening of tubular structures using only real-time US imaging feedback. First, a U-Net was trained for real-time segmentation of vascular structure from cross-sectional US images. Then, we represented the detected vascular structure as a 3-D point cloud, which was used to estimate the centerline of the target structure and its local radius by solving a constrained nonlinear optimization problem. Iterating the previous processes, the US probe was automatically aligned to the normal direction of the target structure, while the object was constantly maintained in the center of the US view. The real-time segmentation result was evaluated both on a phantom and in vivo on brachial arteries of volunteers. In addition, the whole process was validated using both simulation and physical phantoms. The mean absolute orientation, centering, and radius error ($\pm$ SD) on a gel phantom were $\text{3.7}\pm \text{1.6}^{\circ }$, $0.2\pm \text{0.2}\,\text{mm}$ and $0.8\pm \text{0.4}\,\text{mm}$, respectively. The results indicate that the method can automatically screen tubular structures with an optimal probe orientation (i.e., normal to the vessel) and accurately estimate the radius of the target structure.

Acquiring good image quality is one of the main challenges for fully-automatic robot-assisted ultrasound systems (RUSS). The presented method aims at overcoming this challenge for orthopaedic applications by optimizing the orientation of the robotic ultrasound (US) probe, i.e. aligning the central axis of the US probe to the tissue's surface normal at the point of contact in order to improve sound propagation within the tissue. We first optimize the in-plane orientation of the probe by analyzing the confidence map [1] of the US image. We then carry out a fan motion and analyze the resulting forces estimated from joint torques to align the central axis of the probe to the normal within the plane orthogonal to the initial image plane. This results in the final 3D alignment of the probe's main axis with the normal to the anatomical surface at the point of contact without using external sensors for surface reconstruction or localizing the point of contact in an anatomical atlas. The algorithm is evaluated both on a phantom and on human tissues (forearm, upper arm and lower back). The mean absolute angular difference ( $\pm$ STD) between true and estimated normal on stationary phantom, forearm, upper arm and lower back was $3.1 \pm 1.0^\circ$ , $3.7 \pm 1.7^\circ$ , $5.3\pm 1.3^\circ$ and $6.9 \pm 3.5^\circ$ , respectively. In comparison, six human operators obtained errors of 3.2 \pm 1.7^\circ on the phantom. Hence the method is able to automatically position the probe normal to the scanned tissue at the point of contact and thus improve the quality of automatically acquired ultrasound images.

The correct orientation of an ultrasound (US) probe is one of the main parameters governing the US image quality. With the rise of robotic ultrasound systems (RUSS), methods that can automatically compute the orientation promise repeatable, automatic acquisition from predefined angles resulting in high-quality US imaging. In this paper, we propose a method to automatically position a US probe orthogonally to the tissue surface, thereby improving sound propagation and enabling RUSS to reach predefined orientations relatively to the surface normal at the contact point. The method relies on the derivation of the underlying mechanical model. Two rotations around orthogonal axes are carried out, while the contact force is being recorded. Then, the force data are fed into the model to estimate the normal direction. Accordingly, the probe orientation can be computed without requiring visual features. The method is applicable to the convex and linear probes. It has been evaluated on a phantom with varying tilt angles and on multiple human tissues (forearm, upper arm, lower back, leg). As a result, it has outperformed existing methods. The mean (±SD) absolute angular difference on in-vivo tissues averaged over all anatomies and probe types is 2.9±1.6 and 2.2±1.5 on the phantom.

Ultrasound (US) is one of the most common medical imaging modalities since it is radiation-free, low-cost, and real-time. In freehand US examinations, sonographers often navigate a US probe to visualize standard examination planes with rich diagnostic information. However, reproducibility and stability of the resulting images often suffer from intra- and inter-operator variation. Reinforcement learning (RL), as an interaction-based learning method, has demonstrated its effectiveness in visual navigating tasks; however, RL is limited in terms of generalization. To address this challenge, we propose a simulation-based RL framework for real-world navigation of US probes towards the standard longitudinal views of vessels. A UNet is used to provide binary masks from US images; thereby, the RL agent trained on simulated binary vessel images can be applied in real scenarios without further training. To accurately characterize actual states, a multi-modality state representation structure is introduced to facilitate the understanding of environments. Moreover, considering the characteristics of vessels, a novel standard view recognition approach based on the minimum bounding rectangle is proposed to terminate the searching process. To evaluate the effectiveness of the proposed method, the trained policy is validated virtually on 3D volumes of a volunteer’s in-vivo carotid artery, and physically on custom-designed gel phantoms using robotic US. The results demonstrate that proposed approach can effectively and accurately navigate the probe towards the longitudinal view of vessels.

Tissue deformation in ultrasound (US) imaging leads to geometrical errors when measuring tissues due to the pressure exerted by probes. Such deformation has an even larger effect on 3D US volumes as the correct compounding is limited by the inconsistent location and geometry. This work proposes a patient-specified stiffness-based method to correct the tissue deformations in robotic 3D US acquisitions. To obtain the patient-specified model, robotic palpation is performed at sampling positions on the tissue. The contact force, US images and the probe poses of the palpation procedure are recorded. The contact force and the probe poses are used to estimate the nonlinear tissue stiffness. The images are fed to an optical flow algorithm to compute the pixel displacement. Then the pixel-wise tissue deformation under different forces is characterized by a coupled quadratic regression. In order to correct the deformation at unseen positions on the trajectory for building 3D volumes, an interpolation is performed based on the stiffness values computed at the sampling positions. With the stiffness and recorded force, the tissue displacement could be corrected. The method was validated on two blood vessel phantoms with different stiffness. The results demonstrate that the method is able to effectively correct the force-induced deformation and finally generate 3D tissue geometries.