Ready for the next big leap in making AI truly accessible?
Today, we are excited to release CliMB, our no-code AI-enabled partner for clinical predictive modelling. Despite the significant promise, real-world applications of cutting-edge machine learning remain limited. In fact, clinician scientists face substantial barriers in accessing state-of-the-art tools. While AutoML has been designed as an initial solution to this problem, AutoML tools lack intuitive interfaces and do not address the critical data exploration and engineering phases.
With CliMB, you can build predictive models using natural language. CliMB supports data exploration, engineering, model building, and interpretation – enabling clinician scientists to utilise cutting-edge tools in the fields of data-centric AI, AutoML, and interpretable ML.
This proof of concept is a huge step towards breaking barriers and empowering clinician scientists to build predictive models using cutting-edge tools.
What is CliMB?
CliMB (Clinical Machine learning Builder) is an AI-enabled partner designed to empower clinician scientists to create predictive models from real-world clinical data, all within a single conversation. With its no-code, natural language interface, CliMB guides you through the entire data science pipeline, from data exploration and engineering to model building and interpretation. The intuitive interface combines an interactive chat with a dashboard that displays project progress, data transformations, and visualizations, making it easy to follow along. Leveraging state-of-the-art methods in AutoML, data-centric AI, and interpretability tools, CliMB offers a streamlined solution for developing robust, clinically relevant predictive models.
Our Paper
CliMB: An AI-enabled Partner for Clinical Predictive Modeling
E Saveliev, T Schubert, T Pouplin, V Kosmoliaptsis, M van der Schaar
Abstract
Despite its significant promise and continuous technical advances, real-world applications of artificial intelligence (AI) remain limited. We attribute this to the “domain expert-AI-conundrum”: while domain experts, such as clinician scientists, should be able to build predictive models such as risk scores, they face substantial barriers in accessing state-of-the-art (SOTA) tools. While automated machine learning (AutoML) has been proposed as a partner in clinical predictive modeling, many additional requirements need to be fulfilled to make machine learning accessible for clinician scientists.
To address this gap, we introduce CliMB, a no-code AI-enabled partner designed to empower clinician scientists to create predictive models using natural language. CliMB guides clinician scientists through the entire medical data science pipeline, thus empowering them to create predictive models from real-world data in just one conversation. CliMB also creates structured reports and interpretable visuals.
In evaluations involving clinician scientists and systematic comparisons against a baseline GPT-4, CliMB consistently demonstrated superior performance in key areas such as planning, error prevention, code execution, and model performance. Moreover, in blinded assessments involving 45 clinicians from diverse specialties and career stages, more than 80% preferred CliMB over GPT-4. Overall, by providing a no-code interface with clear guidance and access to SOTA methods in the fields of data-centric AI, AutoML, and interpretable ML, CliMB empowers clinician scientists to build robust predictive models.
Clinical Insights and Demonstration
We recently discussed CliMB extensively with clinical researchers during one of our Revolutionizing Healthcare sessions. In this episode, Prof Mihaela van der Schaar introduces the basics of using AI to empower clinicians and how the new tool (CliMB) or similar virtual AI/ML partners could transform clinical research. This is followed by a panel discussion to hear clinical perspectives from experts.
You can watch the whole video here:
GitHub Repository
You can find and download CliMB on GitHub. Our Repository includes the demonstration and a comprehensive manual that will allow you to understand and run CliMB yourself:
CliMB on Github
Watch the Demo