Partnership on Lung Tuberculosis Detection Project with Butterfly Network
Butterfly Network, a company based in Burlington, MA, has joined an international research project called the CAD LUS4TB consortium. The project aims to evaluate the impact of AI-assisted point-of-care ultrasound (POCUS) on early detection of tuberculosis (TB) in under-resourced regions, specifically in sub-Saharan Africa.
The CAD LUS4TB consortium, consisting of 10 African and European research institutions, will conduct the trial using Butterfly's portable ultrasound devices. The trial will involve 3,000 adult patients across sub-Saharan Africa and will commence in September.
The AI algorithm used in the project is designed to automate image classification for TB from images obtained with Butterfly's portable ultrasound devices. This AI algorithm is being used in the project for the first time. The trial will focus on the use of AI-enabled POCUS for the detection of TB and will not involve children or adolescents.
The long-term goal of the consortium is to incorporate the technology into national healthcare policies and systems. The trial is focused on early detection of tuberculosis and is not being conducted in regions outside of sub-Saharan Africa.
The new AI algorithm used in the initial trial is developed using Butterfly's software development platform. The trial will not use any AI algorithm other than the one developed using Butterfly's software development platform.
The AI-enabled POCUS is intended to aid nonradiologist frontline healthcare workers in identifying tuberculosis in patients more quickly and accurately. The consortium aims to generate evidence-driven policy supporting AI-enabled lung ultrasound in TB management.
The initial trial for the CAD LUS4TB project will mark a significant step forward in the fight against tuberculosis, particularly in under-resourced regions. The project's success could pave the way for wider adoption of AI-assisted POCUS in healthcare systems worldwide.
