Aim
To address the unmet medical need for equitable access to expert-level care in pediatric trauma centers by leveraging AI-powered ultrasound technology to accurately identify ultrasound views and ultimately isolate areas of bleeding.
“The ultrasound is safe, fast, and effective for children, yet not a lot of physicians are using it in their diagnostic strategy. We set out to figure out why.” – Jeni Janci, CEO, CaptureDX
Partners
CaptureDX stemmed from the UCSF lab of Dr. Aaron Kornblith, underlining a commitment to advancing pediatric care through innovative technology and research. The team is focused on testing the feasibility of an app-based point-of-care ultrasound workflow that is compatible with a wide range of ultrasound machines. The CaptureDX team partnered with SOM Tech to develop an AI-enhanced ultrasound solution tailored for pediatric trauma centers.
Background
The UCSF lab of Dr. Aaron Kornblith sought to enhance the safety and efficacy of treating injured children, recognizing their heightened vulnerability to conventional radiology procedures such as CT scans, which can lead to the development of cancer in up to 1 in 300 cases. To address this concern, the team developed an AI algorithm tailored to match scans to the database of ultrasound views to assess their completeness and suitability for use. This technology evaluates whether a given ultrasound image adequately captures the landmark area, provides recommendations based on scan quality and completeness, and prompts physicians to rescan if necessary. Jeni Janci, CEO of CaptureDX, expressed her personal resonance with Dr. Kornblith's mission, stating, “As a mother myself, I felt passionate about Dr. Kornblith’s mission and solving the issue of why people aren’t using ultrasounds more often and more effectively, especially with the technology being so much safer for kids.”
Recognizing the barriers to adoption of point-of-care ultrasound (POCUS) in emergency medicine, SOM Tech partnered with CaptureDX to address low confidence in performing and interpreting POCUS studies and inefficient workflows for image documentation and archiving.
Approach
CaptureDX and SOM Tech aimed to develop a minimum viable product (MVP) software integrated into the UCSF Health clinical workflow. This software will deploy AI algorithms to provide real-time feedback to clinicians on captured ultrasound images, guide rescans, interpretation assistance, and documentation process automation. The solution will focus on increasing the use of diagnostic POCUS in the emergency department by overcoming confidence and workflow barriers.
Collaboration
With a product vision in mind, SOM Tech and CaptureDX embarked on a collaborative journey. This included creating wireframes and a clickable prototype to test with physicians to gather feedback and iterate, ensuring seamless integration into the UCSF clinical workflow. SOM Tech business solutions architect Lana Milter led the system architecture development, outlining ultrasound machinery integration requirements, ensuring a thorough understanding of CaptureDX’s clinical implementation. Additionally, UX designer Cynthia Milionis worked closely with the team to craft wireframes that incorporated essential product features.
“SOM Tech helped us figure out what the product would look like and how the product would be integrated into the UCSF clinical workflow.” – Jeni Janci, CEO, CaptureDX
Solution
The AI-enhanced software solution developed by CaptureDX and SOM Tech will integrate AI algorithms to provide real-time feedback and guidance to clinicians on captured ultrasound images, improving diagnostic accuracy and efficiency in pediatric trauma cases. This solution aims to increase the adoption of POCUS in the emergency department, reduce reliance on CT scans, and ultimately improve patient outcomes.
Next Steps
In winter 2023, CaptureDX was selected to participate in the iHackHealth Appathon during which UCSF practitioners propose mobile app ideas to improve research, clinical practice, patient care and education utilizing Apple/iOS products. A UC Berkeley team with skills in product design and front-end/back-end development had two weeks, with guidance from Apple developers, to work on CaptureDX as one of the submitted proposals. Additionally, CaptureDX received a 2024 Catalyst Award from Innovation Ventures’ Catalyst Program, which provides industry advisor mentorship and seed funding to promising projects. The team plans to leverage Appathon insights as they continue to move CaptureDX forward.
Written by Madeleine Gesell for SOM Tech (2024)
