Reduced Traumatic Lumbar Puncture Rates in Pediatric Oncology Patients Through Development of an AI-Assisted, Ultrasound-Based Interventional Guidance Platform

Rivanna Medical, Inc., Charlottesville VA

Investigators

Abstract

PROJECT SUMMARY/ABSTRACT Lumbar punctures (LPs) are critical for the diagnosis and subsequent treatment of pediatric patients with acute lymphoblastic leukemia (ALL), resulting in up to 20 LPs per patient over the course of treatment. Traumatic LPs (TLPs), resulting in blood contamination of the CSF, are highly prevalent in pediatrics, primarily due to challenging patient anatomy. Apart from creating diagnostic confusion, TLPs can negatively impact event-free survival rates of patients with ALL by introducing leukemic cells into the CSF. Although ultrasound guidance has been shown to improve pediatric LP success rates, blind palpation remains the standard of care. The primary barriers to broad utilization of ultrasound guidance for pediatric LPs are challenging workflow requirements and a lack of access to the training required to achieve competency in neuraxial imaging. There is a clear need to decrease TLPs in pediatrics, especially in oncology, and it is likely that a technological advancement that reduces the difficulty of administering bedside LPs is required to meet this need. During this project, an ultrasound-based pediatric LP guidance system will be developed under a quality management system (QMS) certified to ISO 13485:2016 and 21 CFR Part 820. The key technological innovations underpinning this product include the following: imaging algorithms that highlight bone and soft tissue landmarks, anatomical recognition capabilities that automate injection site selection, a novel dual-array ultrasound probe that enables ‘through-the-probe’ needle trajectories, a single-use sterile consumable designed to simplify workflow, and a needle-visualization mode that highlights the advancing needle to the user. The primary technical tasks during the early project stages include the execution of end-user clinical usability studies to guide electromechanical specification development and refactoring of imaging and artificial intelligence algorithms for pediatric patients. Successful completion of these aims will result in the fabrication of pre-production systems for pre-clinical validation studies to be conducted during the project period. Pre-clinical product validation activities will include summative and integrative usability studies, accuracy studies, and learning curve studies performed in collaboration with clinical experts who will verify that the system meets the requirements for the clinical application. The primary endpoint for the pre-clinical accuracy study is direct measurement of the accuracy of Accuro 3S-guided needle placement for simulated LP procedures, as confirmed by CT imaging. Additionally, a second pre-clinical study will characterize the learning curve required to reach competency with the system by conducting a CUSUM analysis with 25 pediatric oncologists over multiple simulated procedures with healthy human volunteers as simulated patients. Completion of this project will result in the development of a clinical-trial-ready ultrasound imaging- guidance system designed to reduce TLP incidence during pediatric LPs through hardware and imaging innovations that minimize the training and workflow barriers currently limiting adoption of US guidance.