AI in the microbiology lab amplifies human ingenuity.
Labs across the world have adopted automation for upfront specimen processing of microbiology samples. The next frontier in full laboratory automation focuses on the software, and it includes artificial intelligence (AI) algorithms to automatically read and interpret growth on plates, count colonies, and recognize morphology.
The essential algorithms for artificial intelligence for microbiology can be grouped into four categories:
1. Colony counting with growth/no growth discrimination: This quickly screens negative plates by colony count and segregates no growth or no significant growth plates from those with growth.
2. Chromogenic detection: This automatically detects color of colonies on chromogenic media plates
3. Phenotypiccolonyrecognition: This examines colonies on non-chromogenic plates, comparing against a library of thousands of colony images to match the phenotypic characteristics and assign predictive value
4. User-defined expert rules algorithm: This filters reporting, using patient demographics to determine if growth is significant and relevant.
Two multi-center studies (1,2) validated the automatic detection and segregation of positive MRSA and VRE samples using chromogenic agar. With sensitivity at 100 percent and specificity between 89.5 percent and 96 percent, both studies showed that unique artificial intelligence algorithms accurately segregated negative from non-negative plates.
Another study (3) with more 5,000 samples looked at the use of algorithms on blood and MacConkey plates and showed agreement between the software and manual interpretation was 99.96 percent for positive, 92.6 percent for negative.
Notably, sensitivity can increase to 100 percent by using the user-defined expert rules algorithm, applying the patient demographic and location source to filter reporting.
The software could improve laboratory workflow by removing more than 40 percent of urine cultures that fall below the growth threshold set by the laboratory.
Sr. Channel Marketing Manager
(1.) Faron ML, Buchan BW, Coon C, et al. Automatic digital analysis of chromogenic media for vancomycin-resistant Enterococcus screens using Copan WASPLab. J Clin Microbiol. 2016;54(10):2464-2469.
(2.) Faron ML, Buchan BW, Vismara C, et al. Automated scoring of chromogenic media for detection of methicillin-resistant Staphylococcus aureus by use of WASPLab image analysis software. J Clin Microbiol. 2016;54:620-624.
(3.) Faron ML, Buchan BW, Relich RF, et al. Digital image analysis to interpret urine cultures on blood and MacConkey agar. Poster presented at the 2017 ASM Microbe, http://www.copanusa.com/education/scientificstudies/use-digital-image-analysis-interpret-urine-cultures-blood-and-macconkey-agar/.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||What's the buzz in automation?; Artificial Intelligence|
|Publication:||Medical Laboratory Observer|
|Date:||Feb 1, 2018|
|Previous Article:||Automation eliminates need for laborious protocols.|
|Next Article:||Automation for everyone--well almost everyone ...|