Digital image analysis agrees with visual estimates of adult bone marrow trephine biopsy cellularity

Construction of an Automatic Quantification Method for Bone Marrow Cellularity Using Image Analysis Software

Although rapid, the evaluation of bone marrow (BM) cellularity is semi-quantitative and largely dependent upon visual estimates. We aimed to construct an automatic quantification method using image analysis software. We used hematoxylin and eosin (HE)-stained specimens of BM biopsies and clots from patients who underwent BM examination at Tottori University Hospital from 2020 to 2022. We compared image analysis (Methods A, B, and C) with visual estimates in pathology reports of 91 HE specimens in 54 cases (29 males, 25 females), including 38 biopsy and 53 clot specimens. Cellularity was visually scored as hypocellular (n = 17), normocellular (n = 44), or hypercellular (n = 30). Compared with the visual estimates, intraclass correlation coefficients for Methods A, B, and C were 0.80, 0.85, and 0.88, respectively. The most appropriate values were obtained with Method C which detected both non-fatty and cell nuclear areas.

Digital imaging software versus the "eyeball" method in quantifying steatosis in a liver biopsy

Steatotic livers represent a potentially underutilized resource to increase the donor graft pool; however, 1 barrier to the increased utilization of such grafts is the heterogeneity in the definition and the measurement of macrovesicular steatosis (MaS). Digital imaging software (DIS) may better standardize definitions to study posttransplant outcomes. Using HALO, a DIS, we analyzed 63 liver biopsies, from 3 transplant centers, transplanted between 2016 and 2018, and compared macrovesicular steatosis percentage (%MaS) as estimated by transplant center, donor hospital, and DIS. We also quantified the relationship between DIS characteristics and posttransplant outcomes using log-linear regression for peak aspartate aminotransferase, peak alanine aminotransferase, and total bilirubin on postoperative day 7, as well as logistic regression for early allograft dysfunction. Transplant centers and donor hospitals overestimated %MaS compared with DIS, with better agreement at lower %MaS and less agreement for higher %MaS. No DIS analyzed liver biopsies were calculated to be >20% %MaS; however, 40% of liver biopsies read by transplant center pathologists were read to be >30%. Percent MaS read by HALO was positively associated with peak aspartate aminotransferase (regression coefficient= 1.04 1.08 1.12 , p <0.001), peak alanine aminotransferase (regression coefficient = 1.04 1.08 1.12 , p <0.001), and early allograft dysfunction (OR= 1.10 1.40 1.78 , p =0.006). There was no association between HALO %MaS and total bilirubin on postoperative day 7 (regression coefficient = 0.99 1.01 1.04 , p =0.3). DIS provides reproducible quantification of steatosis that could standardize MaS definitions and identify phenotypes associated with good clinical outcomes to increase the utilization of steatite livers.

MarrowQuant 2.0: A Digital Pathology Workflow Assisting Bone Marrow Evaluation in Experimental and Clinical Hematology

Bone marrow (BM) cellularity assessment is a crucial step in the evaluation of BM trephine biopsies for hematologic and nonhematologic disorders. Clinical assessment is based on a semiquantitative visual estimation of the hematopoietic and adipocytic components by hematopathologists, which does not provide quantitative information on other stromal compartments. In this study, we developed and validated MarrowQuant 2.0, an efficient, user-friendly digital hematopathology workflow integrated within QuPath software, which serves as BM quantifier for 5 mutually exclusive compartments (bone, hematopoietic, adipocytic, and interstitial/microvasculature areas and other) and derives the cellularity of human BM trephine biopsies. Instance segmentation of individual adipocytes is realized through the adaptation of the machine-learning-based algorithm StarDist. We calculated BM compartments and adipocyte size distributions of hematoxylin and eosin images obtained from 250 bone specimens, from control subjects and patients with acute myeloid leukemia or myelodysplastic syndrome, at diagnosis and follow-up, and measured the agreement of cellularity estimates by MarrowQuant 2.0 against visual scores from 4 hematopathologists. The algorithm was capable of robust BM compartment segmentation with an average mask accuracy of 86%, maximal for bone (99%), hematopoietic (92%), and adipocyte (98%) areas. MarrowQuant 2.0 cellularity score and hematopathologist estimations were highly correlated (R² = 0.92-0.98, intraclass correlation coefficient [ICC] = 0.98; interobserver ICC = 0.96). BM compartment segmentation quantitatively confirmed the reciprocity of the hematopoietic and adipocytic compartments. MarrowQuant 2.0 performance was additionally tested for cellularity assessment of specimens prospectively collected from clinical routine diagnosis. After special consideration for the choice of the cellularity equation in specimens with expanded stroma, performance was similar in this setting (R² = 0.86, n = 42). Thus, we conclude that these validation experiments establish MarrowQuant 2.0 as a reliable tool for BM cellularity assessment. We expect this workflow will serve as a clinical research tool to explore novel biomarkers related to BM stromal components and may contribute to further validation of future digitalized diagnostic hematopathology workstreams.

Automated recognition of glomerular lesions in the kidneys of mice by using deep learning

Background

In recent years, digital pathology has been rapidly developing and applied throughout the world. Especially in clinical settings, it has been utilized in a variety of situations, including automated cancer diagnosis. Conversely, in non-clinical research, it has not yet been utilized as much as in clinical settings. We have been performing automated recognition of various pathological animal tissues and quantitative analysis of pathological findings, including liver and lung. In this study, we attempted to construct an artificial intelligence (AI)-based trained model that can automatedly recognize glomerular lesions in mouse kidneys that are characterized by complex structures.

Materials and methods

By using hematoxylin and eosin (HE)-stained whole slide images (WSI) from Col4a3 KO mice as variation data, normal glomeruli and glomerular lesions were annotated, and deep learning (DL) was performed with the use of the neural network classifier DenseNet system in HALO AI. The trained model was refined by correcting the annotation of misrecognized tissue area and reperforming DL. The accuracy of the trained model was confirmed by comparing the AI-obtained results with the pathological grades evaluated by pathologists. The generality of the trained model was also confirmed by analyzing the WSI of adriamycin (ADR)-induced nephropathy mice, which is a different disease model.

Results

Glomerular lesions (including mesangial proliferation, crescent formation, and sclerosis) observed in Col4a3 KO mice and ADR mice were detected by our trained model. The number of glomerular lesions detected by our trained model were also highly correlated with that of counted by pathologists.

Conclusion

In this study, we constructed a trained model allowing us to automatedly recognize glomerular lesions in the mouse kidney with the use of the HALO AI system. The findings and insights of this study will facilitate the development of digital pathology in non-clinical research and improve the probability of success in drug discovery research.

Using deep learning for quantification of cellularity and cell lineages in bone marrow biopsies and comparison to normal age-related variation

Cellularity estimation forms an important aspect of the visual examination of bone marrow biopsies. In clinical practice, cellularity is estimated by eye under a microscope, which is rapid, but subjective and subject to inter- and intraobserver variability. In addition, there is little consensus in the literature on the normal variation of cellularity with age. Digital image analysis may be used for more objective quantification of cellularity. As such, we developed a deep neural network for the segmentation of six major cell and tissue types in digitized bone marrow trephine biopsies. Using this segmentation, we calculated the overall bone marrow cellularity in a series of biopsies from 130 patients across a wide age range. Using intraclass correlation coefficients (ICC), we measured the agreement between the quantification by the neural network and visual estimation by two pathologists and compared it to baseline human performance. We also examined the age-related changes of cellularity and cell lineages in bone marrow and compared our results to those found in the literature. The network was capable of accurate segmentation (average accuracy and dice score of 0.95 and 0.76, respectively). There was good neural network-pathologist agreement on cellularity measurements (ICC=0.78, 95% CI 0.58-0.85). We found a statistically significant downward trend for cellularity, myelopoiesis and megakaryocytes with age in our cohort. The mean cellularity began at approximately 50% in the third decade of life and then decreased ±2% per decade to 40% in the seventh and eighth decade, but the normal range was very wide (30-70%).

Recent technical advances in whole slide imaging instrumentation

Microscopic observation of biological specimen smears is the mainstay of diagnostic pathology, as defined by the Digital Pathology Association. Though automated systems for this are commercially available, their bulky size and high cost renders them unusable for remote areas. The research community is investing much effort towards building equivalent but portable, low-cost systems. An overview of such research is presented here, including a comparative analysis of recent reports. This paper also reviews recently reported systems for automated staining and smear formation, including microfluidic devices; and optical and computational automated microscopy systems including smartphone-based devices. Image pre-processing and analysis methods for automated diagnosis are also briefly discussed. It concludes with a set of foreseeable research directions that could lead to affordable, integrated and accurate whole slide imaging systems.

Intraoperative fluorescence imaging with aminolevulinic acid detects grossly occult breast cancer: a phase II randomized controlled trial

BACKGROUND

Re-excision due to positive margins following breast-conserving surgery (BCS) negatively affects patient outcomes and healthcare costs. The inability to visualize margin involvement is a significant challenge in BCS. 5-Aminolevulinic acid hydrochloride (5-ALA HCl), a non-fluorescent oral prodrug, causes intracellular accumulation of fluorescent porphyrins in cancer cells. This single-center Phase II randomized controlled trial evaluated the safety, feasibility, and diagnostic accuracy of a prototype handheld fluorescence imaging device plus 5-ALA for intraoperative visualization of invasive breast carcinomas during BCS.

METHODS

Fifty-four patients were enrolled and randomized to receive no 5-ALA or oral 5-ALA HCl (15 or 30 mg/kg). Forty-five patients (n = 15/group) were included in the analysis. Fluorescence imaging of the excised surgical specimen was performed, and biopsies were collected from within and outside the clinically demarcated tumor border of the gross specimen for blinded histopathology.

RESULTS

In the absence of 5-ALA, tissue autofluorescence imaging lacked tumor-specific fluorescent contrast. Both 5-ALA doses caused bright red tumor fluorescence, with improved visualization of tumor contrasted against normal tissue autofluorescence. In the 15 mg/kg 5-ALA group, the positive predictive value (PPV) for detecting breast cancer inside and outside the grossly demarcated tumor border was 100.0% and 55.6%, respectively. In the 30 mg/kg 5-ALA group, the PPV was 100.0% and 50.0% inside and outside the demarcated tumor border, respectively. No adverse events were observed, and clinical feasibility of this imaging device-5-ALA combination approach was confirmed.

CONCLUSIONS

This is the first known clinical report of visualization of 5-ALA-induced fluorescence in invasive breast carcinoma using a real-time handheld intraoperative fluorescence imaging device.

TRIAL REGISTRATION

Clinicaltrials.gov identifier NCT01837225 . Registered 23 April 2013.

Application of Single-Cell Approaches to Study Myeloproliferative Neoplasm Biology

Philadelphia-negative myeloproliferative neoplasms (MPNs) are an excellent tractable disease model of a number of aspects of human cancer biology, including genetic evolution, tissue-associated fibrosis, and cancer stem cells. In this review, we discuss recent insights into MPN biology gained from the application of a number of new single-cell technologies to study human disease, with a specific focus on single-cell genomics, single-cell transcriptomics, and digital pathology.

Artificial Intelligence and Digital Microscopy Applications in Diagnostic Hematopathology

Digital Pathology is the process of converting histology glass slides to digital images using sophisticated computerized technology to facilitate acquisition, evaluation, storage, and portability of histologic information. By its nature, digitization of analog histology data renders it amenable to analysis using deep learning/artificial intelligence (DL/AI) techniques. The application of DL/AI to digital pathology data holds promise, even if the scope of use cases and regulatory framework for deploying such applications in the clinical environment remains in the early stages. Recent studies using whole-slide images and DL/AI to detect histologic abnormalities in general and cancer in particular have shown encouraging results. In this review, we focus on these emerging technologies intended for use in diagnostic hematology and the evaluation of lymphoproliferative diseases.

Automatic Bone Marrow Cellularity Estimation in H&E Stained Whole Slide Images

Bone marrow cellularity is an important measure in diagnostic hematopathology. Currently, the gold standard for bone marrow cellularity estimation is manual inspection of hematoxylin and eosin stained whole slide images (H&E WSI) by hematopathologists. However, these assessments are subjective and subject to interobserver and intraobserver variability. This may be reduced by using a computer-assisted estimate of bone marrow cellularity. The aim of this study was to develop a fully automated algorithm to estimate bone marrow cellularity in H&E WSI stains using bone marrow segmentation. Data consisted of eight bone marrow H&E WSIs extracted from eight subjects. An algorithm was developed to estimate the bone marrow cellularity consisting of biopsy segmentation, tissue classification, and bone marrow segmentation. Segmentations of the red and yellow bone marrow (YBM) were used to estimate the bone marrow cellularity within the WSI H&E stains. The DICE coefficient between automatic tissue segmentations and ground truth segmentations conducted by an experienced hematopathologist were used for validation. Furthermore, the agreement between the automatic and two manual cellularity estimates was assessed using Bland-Altman plots and intraclass correlation coefficients (ICC). The validation of the bone marrow segmentation demonstrated an average DICE of 0.901 and 0.920 for the red and YBM, respectively. A mean cellularity estimate difference of -0.552 and - 7.816 was obtained between the automatic cellularity estimates and two manual cellularity estimates, respectively. An ICC of 0.980 (95%CI: 0.925-0.995, P-value: 5.51 × 10^-7 ) was obtained between the automatic and manual cellularity estimates based on manual annotations. The study demonstrated that it was possible to obtain bone marrow cellularity estimates with a good agreement with bone marrow cellularity estimates obtained from an experienced hematopathologist. © 2019 International Society for Advancement of Cytometry.

Quantification of histopathological findings using a novel image analysis platform

Digital pathology, including image analysis and automatic diagnosis of pathological tissue, has been developed remarkably. HALO is an image analysis platform specialized for the study of pathological tissues, which enables tissue segmentation by using artificial intelligence. In this study, we used HALO to quantify various histopathological changes and findings that were difficult to analyze using conventional image processing software. Using the tissue classifier module, the morphological features of degeneration/necrosis of the hepatocytes and muscle fibers, bile duct in the liver, basophilic tubules and hyaline casts in the kidney, cortex in the thymus, and red pulp, white pulp, and marginal zone in the spleen were learned and separated, and areas of interest were quantified. Furthermore, using the cytonuclear module and vacuole module in combination with the tissue classifier module, the number of erythroblasts in the red pulp of the spleen and each area of acinar cells in the parotid gland were quantified. The results of quantitative analysis were correlated with the histopathological grades evaluated by pathologists. By using artificial intelligence and other functions of HALO, we recognized morphological features, analyzed histopathological changes, and quantified the histopathological grades of various findings. The analysis of histopathological changes using HALO is expected to support pathology evaluations.

Tumor Cell Subtypes Based on the Intracellular Hormonal Activity in KCNJ5-Mutated Aldosterone-Producing Adenoma

Aldosterone-producing adenomas (APAs) harbor marked intratumoral heterogeneity in terms of morphology, steroidogenesis, and genetics. However, an association of biological significance of morphologically identified tumor cell subtypes and genotypes is virtually unknown. KCNJ5 mutation is most frequently detected and generally considered a curable phenotype by adrenalectomy. Therefore, to explore the biological significance of KCNJ5 mutation in APA based on intracellular hormonal activities, 35 consecutively selected APAs (n=18; KCNJ5 mutated, n=17; wild type) were quantitatively examined in the whole tumor areas by newly developed digital image analysis incorporating their histological and ultrastructural features (14 cells from 2 KCNJ5-mutated APAs and 15 cells from 1 wild type) and CYP11B2 immunoreactivity. Results demonstrated that KCNJ5-mutated APAs had significantly lower nuclear/cytoplasm ratio and more abundant clear cells than wild type. CYP11B2 immunoreactivity was not significantly different between these genotypes, but a significant correlation was detected between the proportion of clear cells and CYP11B2 immunoreactivity in all of the APAs examined. CYP11B2 was predominantly immunolocalized in clear cells in KCNJ5-mutated APAs. Quantitative ultrastructural analysis revealed that KCNJ5-mutated APAs had significantly more abundant and smaller-sized mitochondria with well-developed cristae than wild type, whereas wild type had more abundant lipid droplets per unit area despite the small number of the cases examined. Our results did provide the novel insights into the morphological features of APA based on their biological significance. KCNJ5-mutated APAs were characterized by predominance of enlarged lipid-rich clear cells possibly resulting in increased neoplastic aldosterone biosynthesis.