PAIRUS: Pandemic Artificial Intelligence-based Risk Unified Stratification
BACK TO NEWSROOM
Communications Disseminations 12.04.2021

PAIRUS: Pandemic Artificial Intelligence-based Risk Unified Stratification

iLoF – Intelligent Lab on Fiber

PAIRUS video presentation

COVID-19 pandemics have demonstrated that, globally, we are not prepared with adequate tools to predict and manage infectious outbreaks. In this project, we aim to implement a rapid and low-cost, personalized medicine-based tool capable of distinguishing COVID-19 from other respiratory infections and predicting the evolution of COVID-19 viral infection on a patient-specific basis. This tool would be the basis of a future general pandemic stratification platform, easily adaptable to fight both current and future pandemic outbreaks. The platform is built based on the iLoF (intelligent Lab on Fiber) platform, a fully operational validated solution, based on photonics and Artificial Intelligence (AI) algorithms that allows the detection and identification of bio-nanostructures in liquid dispersions (e.g., plasma) for personalized and precision medicine applications. The iLoF is registering and analyzing the “optical fingerprint” resultant from the interaction of light beams with bio-nanostructures, namely the ones found specifically dysregulated in the plasma due to the inflammatory response to the SARS-CoV-2. By comparing the “fingerprint” with the library of “fingerprints” previously stored in the data set, it was possible to identify “disease fingerprints” and clinical outcomes associated with specific phenotypes and biochemical profiles.

COVID-19 may cause mild/absent symptoms for days and then progress quickly to severe disease requiring ventilation aid. Additionally, there are patients which recover in a fast way, while others in a slow manner without a clear explanation of the biological mechanisms that are involved. While prediction of hospital stay/discharge is uncertain, it is vital to manage the demand of limited resources. Enabling the discharge of low-risk patients earlier could be key for public health or the patient. We propose a low-cost, fast stratification/clinical outcome prediction tool using a validated photonics AI platform. We aim to use our platform for different applications that are currently affecting the public health and, also, for future threats.