In this poster, the application of artificial intelligence (AI) in automated culture plate reading specifically addressing concerns related to the detection of mixed microbial populations is presented. Traditional validation approaches for quantitative microbiological methods often emphasize assessing potential interference when multiple organisms are present. However, such methodologies may not be directly applicable to advanced AI-driven systems like the APAS Independence.

Unlike traditional methods that require testing for interference in mixed microbial populations, APAS Independence uses pixel-level analysis to independently detect and classify colonies based on morphology. This approach eliminates bias from neighboring growth and negates the need for mixture testing. The system’s AI models are developed using extensive annotated datasets and validated under established regulatory frameworks, ensuring robust and reliable performance in routine microbiological workflows.

Poster: Clever Culture Systems

Conference: PDA Week 2025

Authors: Steven Giglio, PhD and Chris Ramsey, PhD

‍