The First-of-its-Kind ARM Microscopy Study for Ki-67 Scoring with Augmentiqs was Conducted at UPMC by Dr. Liron Pantanowitz and published in the May 2020 Journal of Cancer Cytopathology
Accepted and published in the Journal of Cancer Cytopathology, the non-sponsored and peer-reviewed study explores the functionality, accuracy and time saving advantages of Augmentiqs as an ARM Microscopy (augmented reality microscope) for Ki-67 neuroendocrine tumor scoring. Titled: “Ki-67 Proliferation Index in Neuroendocrine Tumors: Can Augmented Reality (AR) Microscopy with Image Analysis Improve Scoring?”, the study demonstrated how augmented reality and embedded QuPath image analysis software enables a pathologist to quantify samples in minutes, as opposed to the lengthy manual count.
ARM Microscopy Study
Objective results, cost efficient technology, & time savings.
The Ki-67 index is determined by counting the percentage of positively stained tumor cells in an area of the tumor with the highest nuclear labeling (ie, a hot spot) based on 500 to 2000 tumor cells. Several studies also have attempted to establish the best counting method for determining the Ki-67 index. However, there is no accepted standard as yet and the best scoring method still is debatable.
As manual methods suffer from low reproducibility and high inter-reader variability among pathologists, the authors anticipated that ARM microscopy may overcome issues related to Ki-67 quantification.
The objective of the study was to compare ARM with different scoring methods used for determining the Ki-67 proliferation index, looking at the quality of the results and time required, while keeping in mind the cost factor of the different methods.
ARM Microscopy Study for Ki-67 Scoring – Comparing the Multiple Methods of NET Counting
The ground-breaking study compared Augmentiqs as ARM microscopy, versus all the standard methods of Ki-67 scoring used today including the semi-quantitative eyeball approach, the quantitative print and circle method and Whole Slide Imaging with image analysis.
Understanding the pros and cons of each approach, whether in cost, time or accuracy, is critical for assesing the value of ARM microscopy as a potential tool for digital pathology.
- Eyeball Estimation (EE) – Typically, cytopathologists assess the entire cell block glass slide at a low magnification for the detection of hot-spot areas, and then at an intermediate magnification to estimate the Ki-67 index without performing actual counts of individual cells.
- Printed Image with Manual Counting (PIMC) – Using a digital camera, cytopathologists capture and print out color images of up to 3 hot spot FOVs of each cell block at ×20 magnification. The Ki-67–positive tumor cell nuclei were circled in red and all negative nuclei were crossed out using a gray pencil.
- Augmented Reality Microscopy (ARM Microscopy) – To perform ARM, Augmentiqs was attached to an Olympus light microscope and adjacent computer. For quantitative analytics, this ARM setup was coupled with image analysis software modified from QuPath software for the detection, segmentation, and scoring of nuclear stains. After image analysis results were superimposed on glass slides, cytopathologists manually selected all stained nuclei. These selected stained cells then were divided by the total number of cells that were automatically detected by the software to calculate the Ki-67 index for each FOV. Hence, the actual quantitative assessment component still required human involvement and was not automated.
- Digital image analysis using wholeslide images (DIA) – Ki-67–stained slides were scanned using an Aperio scanner using ×40 magnification and 1 Z-plane. DIA then was performed using the Aperio immunohistochemistry color convoluted, nuclearV9 quantitative image analysis algorithm.
The findings comparing different methods are shown in Table 1. While the overall Ki-67 index varied only slightly with the different counting methods, there were marked differences when scoring individual FOVs. A few grade I tumors were upgraded to grade II when using technology to quantify Ki-67. Relative to the simple eye-ball method, the AR microscope method was quicker than employing camera-captured/printed images for scoring.
An Olympus light microscope with the Augmentiqs ARM Microscopy device fitted between the objective lenses and the eyepiece. (Right) Screenshot showing a field of view using the augmented reality microscopy counting method for Ki-67 Scoring in which the image analysis segments individual nuclei (white circles) and the end user manually selects positive nuclei (red circles).
The Ki-67 index for NETs in cell block material can vary depending on the method using to score positively stained tumor cells. This difference can affect the grade of these NETs, especially for tumors with lower proliferation indices. Manual counting using camera-captured/printed images was the most laborious and time-consuming method. By superimposing image analysis on glass slides in real-time, the AR microscope method simplifies and speeds up this task for pathologists.
Gabe Siegel, Augmentiqs CEO commented on this exciting publication, “Augmentiqs enables pathologists to accomplish what has until now been impossible. The research conducted at UPMC is a clear message that Augmentiqs is not just novel, but more importantly, brings true clinical value to the pathology field. We are looking forward to continuing groundbreaking research like this with other institutions.”
Augmentiqs is a microscope-centric approach to digital pathology, providing pathologists a cost-efficient and low-data method for realizing clinical and workflow enhancements. By connecting the analog microscope to the computer, Augmentiqs maintains the advantages of the microscope for workflow and primary diagnosis, while improving efficiency with the introduction of pathology software applications directly from the microscope.
By functioning as a platform for real-time software deployment within the microscope, Augmentiqs allows pathologists immediate access to imaging, analytical software, telepathology, LIMS integration and other digital pathology applications.
For more information about Augmentiqs, please visit www.augmentiqs.com or contact firstname.lastname@example.org.