We are thrilled to announce we’re joining efforts with Healthplus.ai, assisting them with explainability in their journey to improve post-operative care through AI-powered complication predictions.

Healthplus.ai’s EHR-integrated clinical decision-support tool PERISCOPE™ re-uses existing data with a powerful AI engine for early warning of post-surgery infections and other complications. Earlier diagnosis and actions reduce the number of complications, their impact, and cost, benefiting patients, doctors, clinics, and hospitals.

Deeploy aims to contribute to this innovation with a new explainability technique tailored to medical professionals. The novelty lies in the application of explainability to a time series model and the explainability of feature combinations.

Moreover, Healthplus.ai and Deeploy will design the UI in close collaboration with medical professionals to come to a good presentation of the prediction and explanation, so that the information shown can support the clinical decision process.

In addition, we aim to provide medical professionals with the option to provide feedback on both predictions and explanations. This gives them control and provides the machine learning model with valuable new input data, improving the quality of predictions and explanations over time.

This project is supported by the MIT grant for Artificial Intelligence (AI) R&D Collaborative Projects. This tender supports the development of AI technology and contributes to a strong knowledge and innovation base in the field of AI for SMEs.