Digital AI in Hematology - Integration of the Scopio Labs x100 Scanner with Newly Implemented AI Capabilities into Routine Clinical Workflow

BACKGROUND Digital pathology and artificial intelligence (AI) are areas of growing interest in pathology. A number of institutes have already integrated digital imaging into routine workflow, relying on AI algorithms for the detection of various cancers and mitotic activity quantification. Despite the use of whole slide imaging (WSI) for tissue evaluation, the field of hematology has lagged behind. While many hospitals rely on limited technologies for automated peripheral blood evaluation (e.g. CellavisionTM), the Scopio LabsTM X100 digital scanner provides high resolution oil-immersion level dynamic images of large scanned areas (https://scopiolabs.com/hematology/). With recent FDA-clearance and newly implemented AI capabilities, the Scopio Labs scanner allows for clear and accurate cytomorphologic characterization and cell quantification for peripheral blood smears (PBS). To this end, we aimed to be one of the few pioneering institutes in the United States to adopt early and implement this technology into our routine workflow as a 'hub and spoke' model for optimized case assessment, data sharing and result reporting across multiple satellite locations within our hospital health system. DESIGN A Scopio x100 digital scanner was deployed at our main hospital site, with an anticipated secondary scanner for installment at a satellite laboratory. PBS flagged for hematopathologist review from two satellite laboratories were scanned, and full-field digitalized slides were evaluated by hematopathologists following AI automated analyses. RESULTS 311 peripheral smears were scanned since April 2021 and representative slides were digitalized at 100x magnification (Figure 1, weblink: https://demo.scopiolabs.com/#/view_scan/9231acaf-f898-4649-950d-a41c26c2baaa) with rapid monolayer, monolayer, full-field, and full-field cytopenia scan options available. The automated AI capabilities classified cells into lineage-specific categories with quantification based on cytomorphologic features (Figure 2). Other AI features include additional cell assignment, cell annotation and comments accessible to all users, finalized report PDF generation, export, upload into our current PowerPath TM software with linkage to the corresponding flow cytometry and bone marrow biopsy reports; and the ability to share digitalized slides with clinicians, laboratory personnel and trainees using uniquely generated weblinks. Images can be used for lectures and tumor boards. Additionally, an 80-case study set for PBS was created for medical students, residents and fellow teaching purposes, including cases displaying acute B-cell lymphoblastic leukemia (B-ALL), acute myelomonocytic leukemia (AMML), hypersegmented neutrophils in COVID-19(+) patients, myelodysplastic syndrome (MDS), atypical lymphocytes, hemoglobinopathies, platelet disorders and various lymphomas. Overall improvements were made to the following areas: CLINICAL WORK/DIAGNOSIS 1. Time-saving due to pre-categorization of cells into lineage-specific groups for pathologist review 2. Minimizes subjectivity in cell counting and cellularity assessment EDUCATION 1. Case-based collection with flow and molecular being maintained here 2. Efficient case retrieval with retained annotations/comments for teaching purposes 3. Wide array of digitalized images for hematology atlas and publications ARCHIVING 1. Collection of reference images (intra/inter departmental) for an array of morphological entities for clinical reference and refined diagnosis (e.g. Bethesda reference images for pap by ASC) 2. Digital catalogue for long-term case follow-up and retrospective review CONCLUSION The Scopio Labs X100 digital system provides an efficient and cost-effective web-based tool to streamline clinical workflow and enhance PBS evaluation. With its recent AI capabilities of cell quantification, lineage-assignment and report-generation, we aim to continue our efforts to fully integrate Scopio Labs into our routine daily clinical workflow for reviewing PBS specimens. CONFLICT OF INTEREST STATEMENT The authors have nothing to disclose with regard to the submitted work Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

The root meristem is shaped by brassinosteroid control of cell geometry

Growth extent and direction determine cell and whole-organ architecture. How they are spatio-temporally modulated to control size and shape is not well known. Here we tackled this question by studying the effect of brassinosteroid (BR) signalling on the structure of the root meristem. Quantification of the three-dimensional geometry of thousands of individual meristematic cells across different tissue types showed that the modulation of BR signalling yields distinct changes in growth rate and anisotropy, which affects the time that cells spend in the meristem and has a strong impact on the final root form. By contrast, the hormone effect on cell volume was minor, establishing cell volume as invariant to the effect of BR. Thus, BR has the highest effect on cell shape and growth anisotropy, regulating the overall longitudinal and radial growth of the meristem, while maintaining a coherent distribution of cell sizes. Moving from single-cell quantification to the whole organ, we developed a computational model of radial growth. The simulation demonstrates how differential BR-regulated growth between the inner and outer tissues shapes the meristem and thus explains the non-intuitive outcomes of tissue-specific perturbation of BR signalling. The combined experimental data and simulation suggest that the inner and outer tissues have distinct but coordinated roles in growth regulation.

DNA Dyes—Highly Sensitive Reporters of Cell Quantification: Comparison with Other Cell Quantification Methods

Cell quantification is widely used both in basic and applied research. A typical example of its use is drug discovery research. Presently, plenty of methods for cell quantification are available. In this review, the basic techniques used for cell quantification, with a special emphasis on techniques based on fluorescent DNA dyes, are described. The main aim of this review is to guide readers through the possibilities of cell quantification with various methods and to show the strengths and weaknesses of these methods, especially with respect to their sensitivity, accuracy, and length. As these methods are frequently accompanied by an analysis of cell proliferation and cell viability, some of these approaches are also described.

miR-9a regulates levels of both rhomboid mRNA and protein in the early Drosophila melanogaster embryo

MicroRNAs have subtle and combinatorial effects on the expression levels of their targets. Studying the consequences of a single microRNA knockout often proves difficult as many such knockouts exhibit phenotypes only under stress conditions. This has led to the hypothesis that microRNAs frequently act as buffers of noise in gene expression. Observing and understanding buffering effects requires quantitative analysis of microRNA and target expression in single cells. To this end, we have employed single molecule fluorescence in situ hybridization, immunofluorescence, and high-resolution confocal microscopy to investigate the effects of miR-9a loss on the expression of the serine-protease rhomboid in Drosophila melanogaster early embryos. Our single-cell quantitative approach shows that rhomboid mRNA exhibits the same spatial expression pattern in WT and miR-9a knockout embryos, although the number of mRNA molecules per cell is higher when miR-9a is absent. However, the level of rhomboid protein shows a much more dramatic increase in the miR-9a> knockout. Specifically, we see accumulation of rhomboid protein in miR-9a mutants by stage 5, much earlier than in WT. The data therefore show that miR-9a functions in the regulation of rhomboid activity by both inducing mRNA degradation and inhibiting translation in the blastoderm embryo. Temporal regulation of neural proliferation and differentiation in vertebrates by miR-9a is well-established. We suggest that miR-9a family microRNAs are conserved regulators of timing in neurogenic processes. This work shows the power of single-cell quantification as an experimental tool to study phenotypic consequences of microRNA mis-regulation.

Quantitative and Qualitative Image Analysis of In Vitro Co-Culture 3D Tumor Spheroid Model by Employing Image-Processing Techniques

This work proposes a novel region-estimation (RE) algorithm using the quantification of colon-cancer (HCT-8) and fibroblasts (NIH3T3) cells to estimate the densest region of colon-cancer cells in in vitro 3D co-cultured spheroids. Cells were labelled with different cell tracker dyes to track the cells. The technique involves staining cells with cell trackers The quantification of HCT-8 and NIH3T3 cells by the RE algorithm leads to distribution pattern analysis of cells from the core to the periphery, which ultimately estimates the densest region of HCT-8 cells in an in vitro 3D cell spheroid. Cell quantification by the RE algorithm was compared with the results of cell quantification by ImageJ software. Results demonstrated the distribution patterns of cells from the core to the peripheral region of the in vitro 3D cell spheroid. The overall experimentation showed that the proposed methodology outperformed state-of-the-art approaches in terms of segmentation, quantification, and reducing biasing error.

Sparcle: assigning transcripts to cells in multiplexed images

BackgroundImaging-based spatial transcriptomics has the power to reveal patterns of single-cell gene expression by detecting mRNA transcripts as individually resolved spots in multiplexed images. However, molecular quantification has been severely limited by the computational challenges of segmenting poorly outlined, overlapping cells, and of overcoming technical noise; the majority of transcripts are routinely discarded because they fall outside the segmentation boundaries. This lost information leads to less accurate gene count matrices and weakens downstream analyses, such as cell type or gene program identification.ResultsHere, we present Sparcle, a probabilistic model that reassigns transcripts to cells based on gene covariation patterns and incorporates spatial features such as distance to nucleus. We demonstrate its utility on both multiplexed error-robust fluorescence in situ hybridization (MERFISH) and single-molecule FISH (smFISH) data.ConclusionsSparcle improves transcript assignment, providing more realistic per-cell quantification of each gene, better delineation of cell boundaries, and improved cluster assignments. Critically, our approach does not require an accurate segmentation and is agnostic to technological platform.