Low-Cost Ultrasound Imaging for In-Home, Rural, and Clinic Care Without The Need For Trained Staff

Abstract As currently practiced, ultrasound imaging, recognized universally as being of great health care importance, requires the operator to have considerable training and anatomic and medical image understanding. Access to low-cost imaging services for millions of patients in-home, rural areas and clinics is restricted due to the lack of experienced sonographers and clinicians who can effectively acquire high quality scans. Sonovance, a Maryland small business, is addressing this need by developing a low-cost system for operators with no prior training to acquire diagnostic ultrasound images with standard and portable systems and probes, including mobile devices. The approach uses a simple geometric model to represent the imaging target region, a position sensor to report the location and orientation of the probe, and a calibration procedure by which to co- register the target region to the patient. The user interface guides the operator’s positioning of the probe to acquire 2D images that intersect the target region, providing continuous visual feedback as to which regions require additional scanning without requiring the user to view the images themselves. The system reconstructs these images into a rectangular 3D image volume covering the target region. Sonovance has completed a prototype system using a standard Interson probe and an attached electromagnetic sensor to capture position information which has been shown to successfully capture and reconstruct 3D images at prescribed locations on an abdominal phantom. Two key aims are to be addressed to accomplish the overall goal of arriving at a low-cost and clinically useful prototype from the current one validated on a phantom. Aim 1 is to optimize and validate the company’s patented system for guiding unskilled operators to obtain acceptable diagnostic ultrasound images of the kidneys in human patients. Aim 2 is to further reduce the system cost by employing low-cost inertial position sensors and implementing an image reconstruction method based on a statistical image model with position correction to address the reduced accuracy of the positional data. The two goals are independent of each other and can be developed in parallel. Following successful proof of concept in this study, Phase II research will focus on clinical trials to evaluate and optimize the accuracy and usability of the system by novice users for kidney and urinary tract ultrasound imaging. The overall goal of the program is to develop the software into a product deployable worldwide and build out a “full body” of clinical scanning protocols that can be delivered with the technology. Health systems will be able to deploy novice user scanning with existing AI solutions to assist in diagnostic interpretation based on the patient population. The clinical imaging protocols can be developed by Sonovance, or others via APIs, and will include abdominal, cardiac, Doppler, and OB-GYN. This research aligns with NIBIB’s mission to support ultrasound image acquisition and display as well as innovative signal processing and supports the mission of the Point of Care Technologies Network (POCTRN).