Computerized image processing applications for diagnosing and treating bacterial infections – Erik Hallström

Abstract: Bacterial infections are a major cause of death worldwide, and treatment is increasingly difficult due to antibiotic resistance. In cases of sepsis, timely treatment with the correct antibiotics is paramount, every hour of delay increases the risk of death or servere injury. In this seminar I will present two image processing applications aimed at improving patient outcomes:
1. An image processing pipeline for grading and measuring distances in an assay-based antibiotic combination test to robustly quantify antibiotic interactions.
2. A method for rapidly identifying bacterial species at the single-cell level in a microfluidic “mother machine” using phase-contrast microscopy video classification, thus helping with a fast and appropriate treatment selection.

The presentation is based on our recent publication available at https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1011181