Measuring Digital Pathology Throughput and Tissue Dropouts

Abstract

Background: Digital pathology operations that precede viewing by a pathologist have a substantial impact on costs and fidelity of the digital image. Scan time and file size determine throughput and storage costs, whereas tissue omission during digital capture ("dropouts") compromises downstream interpretation. We compared how these variables differ across scanners.

Methods: A 212 slide set randomly selected from a gynecologic-gestational pathology practice was used to benchmark scan time, file size, and image completeness. Workflows included the Hamamatsu S210 scanner (operated under default and optimized profiles) and the Leica GT450. Digital tissue dropouts were detected by the aligned overlay of macroscopic glass slide camera images (reference) with images created by the slide scanners whole slide images.

Results: File size and scan time were highly correlated within each platform. Differences in GT450, default S210, and optimized S210 performance were seen in average file size (1.4 vs. 2.5 vs. 3.4 GB) and scan time (93 vs. 376 vs. 721 s). Dropouts were seen in 29.5% (186/631) of successful scans overall: from a low of 13.7% (29/212) for the optimized S210 profile, followed by 34.6% (73/211) for the GT450 and 40.4% (84/208) for the default profile S210 profile. Small dislodged fragments, "shards," were dropped in 22.2% (140/631) of slides, followed by tissue marginalized at the glass slide edges, 6.2% (39/631). "Unique dropouts," those for which no equivalent appeared elsewhere in the scan, occurred in only three slides. Of these, 67% (2/3) were "floaters" or contaminants from other cases.

Conclusions: Scanning speed and resultant file size vary greatly by scanner type, scanner operation settings, and clinical specimen mix (tissue type, tissue area). Digital image fidelity as measured by tissue dropout frequency and dropout type also varies according to the tissue type and scanner. Dropped tissues very rarely (1/631) represent actual specimen tissues that are not represented elsewhere in the scan, so in most cases cannot alter the diagnosis. Digital pathology platforms vary in their output efficiency and image fidelity to the glass original and should be matched to the intended application.

Keywords: Digital pathology; dropouts; image analysis; operations; scanner; whole-slide imaging.

Fig. 1

Overlay of aligned reference (camera) and scanner images to detect dropouts. High-contrast grayscale images from a lossless camera (A, reference) and “wsi” digital scanner whole-slide image (B, Hamamatsu S210, default brightfield settings) were rescaled, aligned, and color coded (C, overlay) so that superimposed overlapping areas are blue, and dropouts are red. Details of three dropout regions include translucent fat (c1, c3) and tissue present on the margins of the slide (c2, “edge”). Case BD2019-2222, tissue from aortic lymph node dissection.

Fig. 2

Artifacts excluded from tissue “dropout” tally. The dropout detection process [Figure 1] was capable of detecting very small, refractile, and folded contaminating structures that were artifacts not representative of the tissue section present on the slide. These included dust and fibers (a), compression boundaries caught by the microtome knife along the side edge of the paraffin block (b), individual disaggregated cells (c), and floating contaminants (d, clump of squamous cells in a placental section).

Fig. 3

File size and scan time, by scanner (Leica GT450, Hamamatsu S210 default settings, Hamamatsu S210 optimized settings). File size (x-axis, x^0.5) was proportional to scan time (y-axis, y^0.5) on all conditions. The GT450 had shortest scan times and smallest file size, whereas the S210 using settings optimized to minimize dropouts was longest in scan time and produced the largest files. All 212 slides were scanned successfully with the S210 optimum, with one failure with the GT450 (211 scanned), and four failures with the S210 default (208 scanned).

Fig. 4

Hamamatsu S210 (default profile) scan dropout examples (a-d). Each row shows one slide overlay (Left) and dropout detail (Right). In the overlay dropped tissues appear red and captured tissues cyan-blue (See figure 1). Dashed rectangles in the colored overlay indicate framing of the detail photomicrograph captured from original H&E glass slides. Dropouts were classified according to Table 2.

Fig. 5

Hamamatsu S210 (optimized profile) scan dropout examples (a-d). Each row shows one slide overlay (Left) and dropout detail (Right). In the overlay dropped tissues appear red and captured tissues cyan-blue (See figure 1). Dashed rectangles in the colored overlay indicate framing of the detail photomicrograph captured from original H&E glass slides. Dropouts were classified according to Table 2.

Fig. 6

Leica GT450 scan dropout examples (a-d). Each row shows one slide overlay (Left) and dropout detail (Right). In the overlay dropped tissues appear red and captured tissues cyanblue (See figure 1). Dashed rectangles in the colored overlay indicate framing of the detail photomicrograph captured from original H&E glass slides. Dropouts were classified according to Table 2.

Fig. 7

Surrogate representation of non-unique dropout tissues within scanned area. Some tissues present in the original glass slide (reference camera image, top) were dropped during creation of the digital whole slide image (dropout, red frame). In this example the dropout tissue repertoire was non-unique, as histologically equivalent tissues were included within the final digital scan (green frame). Scanned image for endometrial biopsy case DB2019-2243 from Hamamatsu S210 (default profile). See Fig. 4c for color overlay of reference image with digital scan.

Fig. 8

Unique digital dropout tissues not represented in scanned area. All unique tissue dropouts are illustrated here (left) alongside the source location within the whole-slide reference camera image (right). (a) S210default and GT450 unique dropout. Nondescript fragments of probable contaminating placental villi in endometrial curettings of a 71-year-old patient showing inactive endometrium. (b) S210 default unique dropout. Intact fragment of neoplastic endocervical glands in association with attached stroma, in a cervical biopsy diagnosed as scant fragments of neoplastic glandular epithelium. This dropout is considered unique because all neoplastic glandular epithelia retained in the digital image are detached and unassociated with their stromal context. (c) S210 default unique dropout. Simple epithelium lined fibrous tissue in vulvectomy with reactive stratified squamous epidermis. The epithelial lining of the dropped fragment is not represented in the digital image, and may represent either a contaminant from another case, or separate fragment of a dermal epithelial inclusion.