Aisthesis Medical is participating in ESICM LIVES 2026 in Lisbon, taking place from 10–14 October at the Lisboa Congress Centre. We are represented by our AI Lead, Petros Mitseas, who is presenting three ePosters on AI-powered sepsis prediction, generalisation across hospitals and clinical interpretability.

The studies examine key questions behind the development of VIOSync SPI: how early sepsis risk can be identified, how models perform in different hospitals and how clinicians understand and use their outputs.

Our Research at ESICM LIVES 2026

Explainable Deep Learning for Multi-Horizon Sepsis Prediction: A Multicentre Retrospective Validation Study

This retrospective study evaluates sepsis risk prediction across multiple time horizons using data from three institutions. It examines predictive performance against established scoring systems and the potential to provide earlier insight into evolving risk.

Cross-Institutional Generalisation of AI-Based Sepsis Prediction: Zero-Shot Performance Degradation and Recovery via Parameter-Efficient Fine-Tuning

This study examines the challenges of transferring a model to a new hospital and evaluates adaptation approaches to recover predictive performance. It highlights the importance of local evaluation when preparing clinical AI for deployment.

Interpretability and Clinical Usability of an AI-Based Sepsis Prediction Model: Multicentre SHAP Attribution, Embedding Analysis, and Usability Evaluation

This study explores how model explanations vary across hospitals and care settings, alongside an interface usability evaluation with 22 healthcare professionals. It examines clinicians’ ability to interpret AI-generated risk information.

Connecting Research with Clinical Implementation

Our participation brings this research into direct discussion with the intensive care community. For Aisthesis Medical, these conversations help connect model development with the practical needs of clinicians and hospital teams.

VIOSync SPI turns routinely collected clinical data into explainable sepsis risk scores and trends. Combining predictive AI with Digital Patient Twin technology, it is designed to help clinicians recognise risk earlier, prioritise patient review and support timely assessment and intervention.

About ESICM LIVES 2026

ESICM LIVES 2026 is the 39th Annual Congress of the European Society of Intensive Care Medicine. Held in Lisbon from 10–14 October, it brings together intensive care professionals for scientific sessions, educational courses and workshops.

About Aisthesis Medical

Aisthesis Medical is a deep-tech medtech startup developing cutting-edge AI-driven solutions to revolutionize acute care. Founded in 2022, our mission is to save lives by predicting and preventing sepsis before it becomes fatal. Aisthesis Medical is revolutionizing acute care with VIOSync, the first holistic AI-driven digital patient twin platform that goes beyond early sepsis prediction to guide clinical intervention and optimize patient outcomes. VIOSync, predicts sepsis up to 48 hours earlier, integrates directly into hospital workflows, and optimizes treatment pathways to reduce unnecessary antibiotic use.

Want to Learn More or Partner With Us?

If you’re a hospital, NHS trust, or healthcare investor interested in learning more about VIOSync or our security and compliance posture, we’d love to connect.

📩 Get in touch: https://aisthesismed.com/contact/

🌐 Learn more about VIOSync: https://aisthesismed.com/product/