Engineering the future of 3D pathology

Abstract

In recent years, technological advances in tissue preparation, high‐throughput volumetric microscopy, and computational infrastructure have enabled rapid developments in nondestructive 3D pathology, in which high‐resolution histologic datasets are obtained from thick tissue specimens, such as whole biopsies, without the need for physical sectioning onto glass slides. While 3D pathology generates massive datasets that are attractive for automated computational analysis, there is also a desire to use 3D pathology to improve the visual assessment of tissue histology. In this perspective, we discuss and provide examples of potential advantages of 3D pathology for the visual assessment of clinical specimens and the challenges of dealing with large 3D datasets (of individual or multiple specimens) that pathologists have not been trained to interpret. We discuss the need for artificial intelligence triaging algorithms and explainable analysis methods to assist pathologists or other domain experts in the interpretation of these novel, often complex, large datasets.

Three-dimensional assessments are necessary to determine the true, spatially-resolved composition of tissues

“…Recent 3D work has demonstrated the utility of tissue clearing and serial sectioning-based approaches to assess microanatomical maps of large (>1 cm 3 ) volumes of tissue at cellular resolution. [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30] Here, we use the recently developed 3D imaging workflow CODA to assess the spatial composition of key cells types in thick slabs of both grossly normal human pancreas tissue and human pancreas tissue containing pancreatic ductal adenocarcinoma (PDAC), a deadly and common pancreatic neoplasm. 18 The uniquely heterogeneous spatial microenvironment of PDAC makes it an optimal testbed to evaluate the benefits of 3D microanatomic mapping over standard 2D approaches.…”

“…New work has demonstrated the utility of tissue clearing and serial sectioning-based approaches to create quantifiable volumes of human tissues and tumors to assess microanatomical maps of volumes of tissue of >1 cm 3 at cellular resolution. [17][18][19][20][21][22][23] Here, we use the recently developed 3D imaging workflow CODA to assess the spatial composition of key types of cells in large volumes (>1cm 3 ) of human pancreas containing pancreatic ductal adenocarcinoma (PDAC), the deadliest form of pancreatic cancer. 19 CODA aligns 100's-1000's of serial tissue sections with high resolution, automatically annotates tissue components in H&E sections using deep learning semantic segmentation, deconvolves H&E channels to identify individual cells, and reconstructs 3D maps of the tissue sample.…”

PanIN or IPMN? Redefining Lesion Size in 3 Dimensions

“…In the field of digital pathology, researchers have begun to emphasize the importance of 3D information for accurate understanding of tissue morphology, tumor stage, and heterogeneity 18–22 . Our group developed CODA, a novel histology-based 3D mapping technique that allows cellular resolution imaging, and microanatomical labeling, of large volumes of dense tissues such as the pancreas 20,23–25 .…”

“…In the field of digital pathology, researchers have begun to emphasize the importance of 3D information for accurate understanding of tissue morphology, tumor stage, and heterogeneity. [18][19][20][21][22] Our group developed CO-DA, a novel histology-based 3D mapping technique that allows cellular resolution imaging, and microanatomical labeling, of large volumes of dense tissues such as the pancreas. 20,[23][24][25] Here, we utilize CODA to demonstrate that the size classification guidelines of PanIN and IPMN have significant overlap when lesions are evaluated in high-resolution, 3D space.…”