New Micropore Module Distinguished Bacterial Infections with 96% Accuracy Using AI Analysis of Ionic Currents
-
Novel micropore module with 3μm pore detects ionic current changes when bacteria pass through, generates waveforms. Machine learning analyzes waveforms to distinguish bacterial species.
-
Tested on 50 clinical isolates each of S. aureus and S. epidermidis. Machine learning classifier achieved 96.4% sensitivity and 80.8% specificity in distinguishing them.
-
Rapid (few seconds), low-cost (few dollars per test) technique, easy to use (no specialized training needed). Significant for clinical diagnosis and treatment.
-
Surface charge differences likely the key distinguishing feature, though difference subtle. Ongoing improvements to detect more subtle differences.
-
Future applications distinguish more species, detect antibiotic resistance, one-stop bedside diagnostic platform using cloud AI system.