Ensuring the quality of slides in digital pathology is a meticulous and time-consuming process. The individuals responsible for quality control can vary greatly, ranging from specially trained technicians conducting exhaustive manual reviews of slides, to data scientists or pathologists encountering digital images that fail to meet the required standards for their sensitive work. Labs that have invested heavily in quality control fully understand its value and the associated costs.
Implementing a robust quality control process offers several significant advantages:
- Preventing duplicated work: By catching quality issues early on, extensive re-work by pathologists, technicians and scientists can be avoided. This not only saves time but allows these professionals to feel more fulfilled through focusing their efforts on more rewarding endeavors.
- Swiftly resolving issues: A thorough quality control process ensures that any identified issues are addressed promptly. This expedites the process of delivering results – reducing the likelihood of delayed studies, untimely patient reports, or dissatisfied research customers.
- Guaranteeing high-quality data: Quality control ensures that only reliable and high-quality data enters the realm of AI, image analysis, studies, or the pathologist’s review. This reduces the risk of errors, which can be costly in any setting. For example, a faulty analysis result may erroneously provide confidence in a particular drug candidate which is bound for failure.
- Fostering trust in data quality: A well-executed quality control process instills confidence in the slide and image creation process, thereby reducing the strain that poor data processes can put on employees, and eliminating the need of staff to double-check the work of their colleagues. This, in turn, minimizes the likelihood of burnout and turnover among staff.
- Maximizing data: A comprehensive quality control process is exhaustive, and yields reproducible, accurate, reliable data. It ensures consistent results on every image, every time. By flagging only relevant quality issues and offering complete data coverage, employee time is optimized, available data is optimized, and unnecessary delays from reprocessing are avoided.
Neglecting quality control or diligently adhering to a subpar quality control process can cause labs to miss out on the value it provides and overlook the associated costs. And these issues are not rare. The occurrence of quality issues in digital pathology can vary broadly, frequently estimated between 5 and 10% of scanned images [1,2] with individual labs often reporting rates up to 25% depending on their own practices and standards. For labs that have not implemented a robust and comprehensive quality control process, the costs of these image quality issues are often hidden. Pathologists reviewing poor quality slides do not report a breakdown of their valuable time spent on this task.
A good understanding of the value of robust quality control begs the question: what is the value of automating this process? Proscia’s Automated Quality Control, an add-on to the Concentriq for Research platform, provides all of the benefits listed above, but offers improved efficiency and consistency. Without automation, achieving effective quality control requires hiring technicians to manually scrutinize slides. However, relying on low power examination is insufficient for detecting potential issues. Therefore, in the absence of automated quality control, these technicians will spend up to 5 minutes scanning through each image to find an issue. Unassisted, they can only manage to review a few hundred images per day to find the needles in the haystack that necessitate rework.
In contrast, with automated quality control, this process becomes effortless. Issues are made visible at low power, in search or in aggregate table views, and many images can safely proceed to the next stage without further quality review. Suddenly manual technicians can work up to five times faster, and find their work more rewarding. Alternatively, when manual quality control is not in place, identified issues can be quickly triaged by the appropriate lab personnel or processed through automated workflows. No more will your pathologist be staring down an unsightly out of focus area, and study data will no longer be corrupted with air bubbles interfering with tissue analysis.
With the efficiency gained from Proscia’s Automated Quality Control, laboratories can harness the extra time and resources to drive further advancements in their work. The possibilities are limitless.
Automated Quality Control Resources
Explore Concentriq Webinar: Automated QC
[1] Chen et al. (2021). Assessment of a computerized quantitative quality control tool for whole slide images of kidney biopsies. Journal of Pathology, 253(3), 268-278. doi: 10.1002/path.5590
[2] Rajaganesan S, Kumar R, Rao V, Pai T, Mittal N, Sahay A, Menon S, Desai S. Comparative Assessment of Digital Pathology Systems for Primary Diagnosis. J Pathol Inform. 2021;12:25.
