Introduction and use of an AI imaging device (APAS® Independence) to autoverify Urine cultures – Presented by Dr Glen Hansen
13th July 2021
Current challenges in the microbiology lab include:
- Increasing pressure for “on demand” testing and rapid turnaround times for results
- Inefficient use of qualified personnel – Valuable microbiologist time is used to screen negative cultures and report results
- Shortage of microbiologists – 5.93% vacancy rate
Laboratories need tools to allow them to prioritise cases that are likely to have a clinical impact to enable them to organise their workflow to work on those cases sooner.
This study evaluated the performance of an AI imaging device, the APAS Independence to identify cases that meet the criteria for urine culture workup. A total of 6200 clinical specimens were investigated to determine the accuracy of the APAS Independence to detect no growth urine cultures and to correctly enumerate >104 or <104 CFU/ml on urine culture plates. The APAS Independence results were compared to a clinical technologist bench read for comparison.
The APAS Independence successfully removed 30% (1860) of cases as true “no growth” from the clinical laboratory workflow with 100% agreement with manual clinical interpretation. The APAS Independence demonstrated an adjusted percent agreement of 98% of cases successfully enumerated compared to clinical technologist bench read (remainder 2% were either defect in the media or label misread).
The study demonstrated that 69% of cases could be auto-cleared by the APAS Independence from the clinical workflow, enabling technologists time to be focussed on the cases that are clinically significant.
Video Presentation: by Glen Hansen
Conference: ECCMID 2021, Online
Date: July 2021
Authors: Salden J, Nussbaum A, Wesenberg E, Hansen G.
Citation: Personal Communication