Digital technologies can help improve pathology workflows by automating the review of immunohistochemistry (IHC) slides, which are the cornerstone of diagnostic pathology. However, the inability to simultaneously view multiple digital slides contributes to delays in diagnosis and hinders the widespread clinical adoption of digital pathology.

To accelerate reviewing of IHC slides by pathologists, researchers from the University of Leeds and Leeds Teaching Hospitals NHS Trust developed an algorithm that allows simultaneous reviewing of multiple digital slides.¹ Viewing multiple slides side-by-side is not possible with light microscopy, and manually switching between slides on the microscope stage contributes to delays in diagnosis using conventional microscopes.

The team showed that compared to conventional light microscopy, digital microscopy using their algorithm accelerated the review of IHC slides with relatively little training and without increasing the rate of diagnostic errors.¹

These findings suggest that the integration of new digital technologies into routine clinical practice may accelerate diagnosis and increase the efficiency of pathologists. The study was published in the Journal of Clinical Pathology on January 17, 2022.

Establishing A Novel Digital Microscopy System

The team developed a new system that allows the simultaneous viewing of parallel tissue sections. The system was integrated into a digital pathology workstation. Aperio T3 scanner was used to generate whole slide images of tissue sections, and the whole slide imagining (WSI) software Leeds Virtual Microscope was used to review the digital slides. Leeds Virtual Microscope enables pathologists to simultaneously review immunostained and hematoxylin and eosin (H&E) images in two synchronized panels.

The researchers used a large 6-megapixel medical-grade display to view whole slide images and a smaller 2-megapixel display to navigate between slides. The use of two large monitors enabled pathologists to view more tissue on the screen at one time, thereby reducing the need for interaction by the histopathologist (pan/zoom). A “click and drag” method was used for slide panning, whereas zooming and slide switching were performed using the keyboard.

The pathologists also reviewed tissue slides using a Leica DMR light microscope for comparison. Before the experiment, pathologists were given time to practice manually reviewing slides using the microscope.

Digital Microscopy Reduces the Time to Diagnosis

To compare the time to reach diagnosis using the digital microscope or a conventional light microscope, the team recruited 16 histopathologists (eight trainee pathologists and eight consultant pathologists) to review six IHC-stained needle core biopsies of liver tumors. Three biopsies were examined using the digital workstation, and three were examined using light microscopy.

The use of the digital microscope over a conventional light microscope reduced the mean time to diagnosis by 1 min 21 sec (95% confidence interval [CI], 16 sec to 2 min 26 sec; p = 0.02). The mean time to reach diagnosis was 4 min 3 sec using the digital microscope and 5 min 24 sec using the light microscope. Although the reduction in review time per case may seem small, digital microscopy can save many hours for pathologists who have to review large sets of biopsies.

Because the time to diagnosis varied widely among the cases, the researchers also calculated the normalized time to diagnosis, which was expressed as a percentage of the time to diagnosis on each platform (i.e., digital microscope or light microscope) relative to the time to diagnosis across both platforms.

The normalized mean time to diagnosis was 85% using the digital microscope and 115% using the light microscope. This difference of 26% (95% CI, 15% to 45%) in the normalized mean time to diagnosis between the two platforms was statistically significant (p = 0.0006).

Digital Microscopy Can Accelerate Diagnosis Regardless of Experience

To evaluate whether the experience of the pathologist affected the ability of the digital microscopy system to reduce the time to diagnosis, the team analyzed the time to diagnosis separately for trainee and consultant histopathologists.

Among trainee histopathologists, the use of the digital microscope over a conventional light microscope reduced the mean time to diagnosis by 1 min 54 sec (95% CI, −3 min 11 sec to −0 min 37 sec; p = 0.007). The mean time to diagnosis was 3 min 31 sec using the digital workstation and 5 min 25 sec using the light microscope. Consistently, the use of the digital workstation among trainee histopathologists significantly reduced the normalized mean time to diagnosis by 42% (95% CI, 14% to 70%; 74% for digital microscopy vs. 116% for light microscopy; p = 0.006).

Among consultants, the mean time to diagnosis was 4 min 35 sec using the digital workstation and 5 min 24 sec using the light microscope. Although the use of the digital microscope among consultant histopathologists reduced the time to diagnosis by 0 min 48 sec (95% CI, −2 min 41 sec to 1 min 5 sec), this reduction was not statistically significant (p = 0.37). The normalized mean time to diagnosis for consultants was 96% using the digital workstation and 114% using the light microscope; this 18% (95% CI, −53% to 17%) difference was not statistically significant (p = 0.3).

To adjust for the experience level of histopathologists, the team conducted a multivariable linear regression analysis. The use of the digital microscope over the light microscope significantly reduced the mean time to diagnosis by 1 min 21 sec (95% CI, 0 min 16 sec to 2 min 26 sec; p = 0.017) and the normalized time to diagnosis by 30% (95% CI, –52% to −8%; p = 0.008).

The researchers speculated that the reduction in the time to diagnosis using the digital workstation was likely due to the ability of pathologists to review multiple slides simultaneously, which is impossible using conventional light microscopes.

Future Perspectives

Although the study showed that digital microscopy could accelerate diagnosis by facilitating simultaneous review of multiple IHC-stained slides, the number of cases used in this study was low. The ability of the digital microscope to reduce the time to diagnosis requires further validation in large-scale studies. Additionally, future studies are warranted to evaluate the performance of this new digital microscopy system in real-world settings.

References

1. Clarke E, Doherty D, Randell R, et al. Faster than light (microscopy): superiority of digital pathology over microscopy for assessment of immunohistochemistry. J Clin Pathol. January 2022. doi:10.1136/jclinpath-2021-207961