Imperial College London and CHOC Collaborate to Develop AI for Pediatric Intensive Care Decision Support

Imperial College London and CHOC Collaborate on AI Systems for Paediatric Intensive Care

Imperial College London has partnered with the Children’s Hospital of Orange County (CHOC), a leading paediatric healthcare provider in the United States, to develop AI support systems aimed at improving care in paediatric intensive care units (PICUs). This collaboration focuses on training the AI Clinician model, which assists doctors in making complex decisions and delivering faster, more accurate treatment recommendations for critically ill children.

The research is led by Professor Aldo Faisal, Director for AI for Healthcare Centres and Imperial Global USA Academic Theme Lead. His team includes experts from Imperial’s Departments of Bioengineering, Computing, and Surgery and Cancer, working alongside data scientists and clinicians at CHOC. Professor Faisal, who is among the few computer scientists leading clinical trials from algorithm development through to bedside application, highlighted the potential of this work during his keynote at CHOC’s 2nd Annual International Paediatric and Lifespan Data Science Conference in April 2025.

Building the Largest International Paediatric Dataset

Traditional AI models often rely on simple rules. In contrast, Imperial’s AI Clinician uses advanced algorithms that provide personalised recommendations by considering the full complexity of each patient’s condition. A critical factor in the model’s performance is the quality and scope of data used for training.

This partnership has merged 36,000 anonymised records from CHOC with 20,000 anonymised NHS patient records, creating the largest international paediatric dataset of its kind. This extensive dataset ensures the AI Clinician can support a broader range of conditions and diverse medical backgrounds.

Professor Aldo Faisal said, “The gold standard for developing and evaluating patient-ready AI technology is multi-centre international trials. Our UK-US collaboration enables us to make the AI Clinician smarter so it can assist children with a wider variety of medical needs.”

The Role of AI in Paediatric Intensive Care

PICUs generate vast amounts of data—up to 100 different categories every hour for each patient. These data points provide valuable insights but are challenging to interpret quickly and accurately, especially in time-sensitive situations.

Dr Terence Sanger, Chief Scientific Officer at CHOC, explained, “By enabling earlier detection and prediction of critical clinical events, as well as identifying optimal treatment strategies, we aim to improve outcomes and deliver more personalised, data-driven care to every child.” He added that this partnership reflects CHOC’s commitment to advancing AI applications in paediatrics.

Integrating Clinical Expertise with AI Analysis

The AI Clinician model analyses complex, multi-dimensional patient data to support real-world treatment decisions. However, the accuracy of AI-driven recommendations depends heavily on how well the data is prepared and interpreted.

One of the initial tasks involves standardising drug dosages and aligning irregularly recorded data into hourly formats for AI training. Clinicians at CHOC collaborate closely with Imperial’s data scientists to clarify the meaning behind specific data points, ensuring the model provides reliable and personalised guidance.

Dr Padmanabhan Ramnarayan, Deputy Director of Imperial’s Centre for Paediatrics and Child Health, expressed enthusiasm for the collaboration: “By combining data from tens of thousands of critically ill children in the UK and USA, this effort has the potential to significantly improve paediatric critical care.”

Building on Previous AI Clinician Achievements

This partnership with CHOC builds on prior successes where the AI Clinician model supported adult sepsis treatment and AI-guided ventilation management, tested in London hospitals. Previous collaborations have also involved US institutions such as Johns Hopkins University, Emory University, and Sentara Healthcare.

These ongoing efforts demonstrate the practical value of applying AI to complex clinical challenges and the benefits of international data sharing in improving patient outcomes.

For those interested in advancing their skills in AI applications in healthcare, Complete AI Training offers a range of courses tailored to scientific and research professionals.