Sam Tukra Cohort Four participant

Samyakh Tukra

Imperial College London

Samyakh Tukra

AI for predicting acute health deterioration in critical care

Third Eye has developed an AI-driven software that predicts the risk of organ failure two days before onset in critical care patients. This gives doctors the power to foresee patient deterioration, thereby enabling them to intervene early and save lives

Every year, 85 million patients are admitted to intensive care with organ failure globally. Despite constantly monitoring patients and providing the highest echelon of care, intensive care units have the highest error rate in any given hospital unit. Intensive care doctors who have to make life-or-death decisions in high-pressure settings are often overwhelmed by the mounds of incoming patient data. Leading to delayed intervention, facts being missed, errors being made and, in some cases, death. This keeps happening on a daily basis, to patients world-wide.

To give these patients a fighting chance, Third Eye has developed an AI-driven early warning system for organ failure. Our AI achieves this by utilising routinely collected data in intensive care to predict the risk of organ failure before onset. Trained using the largest hospital dataset from MIT, our AI predicts the risk of organ failure up to 36 hours before onset with an accuracy of 96%. This gives doctors the ability to react & intervene early, benefiting both patients and clinical staff.

For patients, Third Eye improves outcomes by identifying the sickest patient, providing an early warning for doctors to intervene, and through this saving lives. Recognizing the severity of a patient’s illness and intervening early can not only help stop patients from getting sicker and dying, but also expedite their recovery. Furthermore, by decreasing the duration of their hospital stay and preventing acute organ failure from becoming permanent damage, our AI can decrease the risk of post ICU chronic pain syndrome.

For clinical staff, our software optimises resource utilization in terms of staff time, patient flow, and equipment usage based on the predicted risk. This is extremely useful in management of healthcare systems which are overwhelmed by patient numbers and are unable to provide for everyone in need. In these difficult situations, doctors can either choose to give excellent care to the few, or suboptimal care to the many. This was further emphasized by the pandemic, where most healthcare systems chose the humane latter approach.

The Third Eye team is passionate about saving lives through the progression of medical technology. Our founder Samyakh Tukra is an AI research engineer with a background in Medical Engineering and, is currently undertaking his PhD in AI & Computer Vision in surgical robotics. He is the inventor of the evolutionary AI powering the Third Eye platform. Our clinical officer Marta Kedrzycki is a clinician, and is currently undertaking her PhD in surgery and cancer. Her insight into critical care medicine and hospital workflow is integral to tailoring the software to end-user needs. Our business developer and strategist, Luca Meise, has completed his BSc in Chemistry and a MSc in Business Innovation. He is key to strategic growth given his combined scientific and business background. Together we aim to change the face of critical care medicine.

Diagnostics Accelerator Programme: market access planning and evidence generation

Delivered in partnership with the National Institute of Health Research London In Vitro Diagnostics unit this six-week evidence generation focused programme will support medtech innovators to de-risk the path to successful deployment through the development of a business case validating their clinical, financial and social claims. This way, when innovators approach the health system, they are comfortable in the knowledge that the product achieves what it is designed to do.

Applications Open (closing 7 March 2021)