Applications: AMR

Quantitative Phase Microscopy for rapid detection of infection and antimicrobial resistance

Rapid and accurate detection of infection and antimicrobial resistance (AMR) is critical for effective clinical decision-making. Conventional culture-based diagnostic methods typically require 48–96 hours, delaying appropriate treatment and contributing to the spread of resistant pathogens. Our approach integrates highly sensitive label-free quantitative phase microscopy (QPM) with deep learning and whole-genome sequencing (WGS) to enable fast, data-driven identification of infectious agents and their resistance profiles.

Using QPM, individual bacterial cells are imaged without fluorescent labels, capturing intrinsic optical signatures related to cell morphology and biophysical properties. These quantitative phase features are then analyzed using deep convolutional neural networks to classify pathogens by Gram type, species, strain, and antibiotic susceptibility. The optical classification results show strong agreement with genomic characterization obtained by WGS, validating the reliability of this approach.

This combined workflow enables rapid first-stage screening of infections, providing early information on pathogen identity and resistance or susceptibility to antibiotics. WGS can then be used as a follow-up confirmatory tool for detailed genetic profiling. Together, this strategy offers a powerful platform for accelerating AMR diagnostics, supporting timely and targeted antimicrobial therapy, and improving patient outcomes.

By merging label-free optical imaging with artificial intelligence and genomic analysis, this technology represents a transformative pathway toward next-generation, rapid AMR detection and precision microbiology.

References: 

1. Ahmad, Azeem, Ramith Hettiarachchi, Abdolrahman Khezri, Balpreet Singh Ahluwalia, Dushan N. Wadduwage, and Rafi Ahmad. "Highly sensitive quantitative phase microscopy and deep learning aided with whole genome sequencing for rapid detection of infection and antimicrobial resistance." Frontiers in Microbiology 14 (2023): 1154620. Link