AI+ML for rapid detection of bacterial infection to reduce hospital-acquired infections and prevent septic shock and heart attack

The antibiotic susceptibility test (AST) is essential in the clinical diagnosis of serious bacterial infections such as sepsis, but it typically takes 2–5 days for sample culture, antibiotic treatment, and result reading. Detecting metabolites secreted from bacteria with surface-enhanced Raman scattering (SERS) and using robust ML+AI algorithms will enable the identification of the “fingerprint” spectra of the molecular components of various bacteria, and classification of various bacteria instantly at the hospital, provide the right medication and prevent heart attack thus reducing the AST time to few hours. Data are generated at QU, the algorithm is built at QCRI, and the biosensor implemented with AI is validated at HMC.