Autonomous Robotic Ultrasound System for Liver Follow-up Diagnosis: Pilot Phantom Study

TL;DR

This work tackles the need for reliable, cost-effective liver follow-up ultrasound in local clinics by developing an autonomous robotic ultrasound system guided by hepatic veins as CT–US landmarks. The approach integrates deep HV segmentation, 3D HV-based registration, and a slice-matching pipeline to map CT coordinates to the robot’s base frame and autonomously image targets. Key contributions include two HV segmentation models (full HV and branching point) for 3D HV modeling and HV localization, a segmentation-based CT–US registration workflow, and a target localization/imaging module validated on a liver phantom with multiple trials showing robust HV alignment and target imaging. The framework has potential to reduce workflow time and costs for NAFLD follow-ups and can be extended to other organs and multi-modal communication systems for clinical deployment in local communities.

Abstract

The paper introduces a novel autonomous robot ultrasound (US) system targeting liver follow-up scans for outpatients in local communities. Given a computed tomography (CT) image with specific target regions of interest, the proposed system carries out the autonomous follow-up scan in three steps: (i) initial robot contact to surface, (ii) coordinate mapping between CT image and robot, and (iii) target US scan. Utilizing 3D US-CT registration and deep learning-based segmentation networks, we can achieve precise imaging of 3D hepatic veins, facilitating accurate coordinate mapping between CT and the robot. This enables the automatic localization of follow-up targets within the CT image, allowing the robot to navigate precisely to the target's surface. Evaluation of the ultrasound phantom confirms the quality of the US-CT registration and shows the robot reliably locates the targets in repeated trials. The proposed framework holds the potential to significantly reduce time and costs for healthcare providers, clinicians, and follow-up patients, thereby addressing the increasing healthcare burden associated with chronic disease in local communities.

Figures (17)

• Figure 1: Our proposed robot system conducts ultrasound scans on patients for intermediate follow-up examinations at local clinics. We assume that the CT scan data is provided by the hospital. The autonomous US imaging pipeline involves initial contact, robot coordinate mapping, segmentation/registration of the hepatic vein (HV), image localization integrated with robot control, and so on. Deep segmentation networks assist the coordinate mapping and target localization.
• Figure 2: Our autonomous ultrasound scanning robot.
• Figure 3: The illustration describes the interconnectivity of hardware components. Data flow is as follows: RGB-D camera transmits to PC via USB; US probe sends images wirelessly to smartphone, which relays them to PC via USB. PC communicates with the robot arm via Ethernet for state retrieval and control command issuance.
• Figure 4: The axis alignment of our system. The axes labeled 'x', 'y', 'z' with origin at the robot's base constitute our physical coordinate system. The axes labeled 'inferior', 'left', 'anterior' with origin on the body's upper surface constitute the anatomical coordinate system.
• Figure 5: (a) A target location in the CT image near the branching point. (b) Ultrasound image taken by the robot that matches the target in (a).

Theorems & Definitions (3)

• Remark 1
• Remark 2
• Remark 3