APAS Pharma

Proof of Concept
October 2021

APAS® Pharma uses LBT’s cutting-edge artificial intelligent image analysis algorithms to automatically detect microbial growth for environmental monitoring applications.

Microbial quality control (QC) is an essential activity for monitoring critical production environments where sterility is required. Each year over 350 million Microbial QC tests are performed to ensure the safety of drugs.

There is a growing need from pharmaceutical regulators for automated microbial detection systems in pharmaceutical microbiology laboratories to reduce sources of human error, accelerate return of results and improve data integrity for this critical activity.

Challenges for Microbial Quality Control

Inconsistent Results

Variability and inconsistency of microbial counts due to human variation and manual counting process.

Increasing Regulatory Burden

Increased data integrity requirements, including second analyst verification.

Operational Inefficiency

High volume of samples with no microbial growth (up to 90%).

Time Consuming

Critical time spent performing repetitive labour-intensive tasks.

These errors may be considered, by a health authority investigator, to be an indication of data integrity issues.1

Over 200 FDA warning letters per year with data integrity findings since 2015.

LBT Solution

The APAS Pharma project delivers an advanced solution using image analysis and artificial intelligence technology to automate the reading of culture plates used for environmental monitoring.

The APAS Pharma product has been developed using standard 90mm Tryptone Soy Agar culture media. The artificial intelligence algorithms showed excellent performance for growth detection with no microbial growth missed across the plate. Additionally, initial linearity studies were performed with APAS Pharma demonstrating a high accuracy for bacterial enumeration (R2 = >0.9).

Accurate colony enumeration of bacterial growth
L: plate image, R: APAS reconstructed image.

Automatic recognition of different colony morphologies
L: plate image, R: APAS reconstructed image.

APAS Independence

Key Features and Benefits

  • Automatic reporting of plates showing no microbial growth
  • Reduces need for contemporaneous, second analyst verification
  • Improved data integrity through automation and direct image capture
  • Greater standardisation of results – removes reader variability, subjectivity and transcription errors
  • Audit and data reports available in compliance with CFR 21 Part 11
  • Rapid high-throughput laboratory instrument able to read 200 plates/hour
  • Automated integration with laboratory information system to remove transcription errors

1 A Systematic Approach for the Evaluation, Validation, and Implementation of Automated Colony Counting Systems, Sven Deutschmann, Bill Carpenter, Caroline Duignan, et al., PDA Journal of Pharmaceutical Science and Technology 2022