Intelligent Automation – the first US Use of the APAS Independence Delivering Artificial Intelligence for Clinical Microbiology Automation

25 June 2019

Clinical Microbiology relies on technically skilled laboratorians to process, and interpret cultures from clinical specimens. Laboratory automation provides improved accuracy, decreased turn-around times, improves efficiency and reduces reliance on maintaining, acquiring and training qualified workforce. However, the impact of automation remains unfulfilled, in part, because robotics fails to incorporate interpretive processes needed to fully prioritize microbiology work needed to maximize efficiencies and impact reporting. The APAS independence (APAS) is an in-vitro diagnostic device incorporating machine learning algorithms with digital image capture. We report the first US experience of the APAS using urine cultures plated to sheep blood agar (SBA).

Total of 369 urine cultures (REMEL automated-SBA) were placed in the APAS for review and reporting allocation. Reports were recorded based on the ability of the APAS to correctly identify bacterial growth and clear negative cultures from the workflow. The APAS was able to report and finalize negative SBA urine cultures in 13 seconds. The system correctly identified growth form SBA in 100% of the cases and correctly screened negative cases, including those requiring review with 100% NPV. The APAS incorporates machine learning/artificial intelligence into routine microbiology workflow, applying decision making processes capable of prioritizing postive cases, screening negative growth, and reducing time-to-report.

Poster Presentation: by Dr Glen Hansen

Conference: ASM Microbe 2019, San Francisco

Date: June 2019

Authors: Bujold A, Cox A, Hansen G, Hanson K, Wesenberg E

Citation: Bujold A, Cox A, Hansen G, Hanson K, Wesenberg E., ASM Microbe 2019, Intelligent Automation- the first US Use of the APAS Independence Delivering Artificial Intelligence for Clinical Microbiology Automation.