Bibliography

Summary: In this longitudinal COVID-19 cohort study, they used Sysmex DI-60 digital cell imaging analyzer with CellaVision DM Software to capture and analyze peripheral blood smear morphology over time. After slides were automatically made and Wright-Giemsa stained on an SP-10, the DI-60 scanned optimal smear areas and produced high-magnification images specifically for analyzing erythrocytes and platelets. Using the software’s built-in measuring rulers, the authors automatically quantified microcytes (<6.5 μm) and macrocytes (>8.5 μm) and visually assessed features such as anisocytosis, platelet aggregation, large platelets, and giant platelets. Two skilled reviewers independently verified the digital images, and these CellaVision-supported morphology results were tracked across four disease timepoints (admission, discharge, 1-year, 2-year follow-up) and compared between severe and non-severe cases.

Summary: In this research article, CellaVision instruments were used as the image-capture platform for building the red-blood-cell (RBC) morphology dataset. Peripheral blood smears from Massachusetts General Hospital and Brigham and Women’s Hospital were imaged on CellaVision systems (DM96 or DI60); the instruments automatically selected well-spread smear areas and captured large fields containing roughly 1,000–3,000 RBCs per image for routine clinical evaluation. The digital smear images were then fed into the authors’ open-source “RBC-diff” computer-vision pipeline to quantify nine RBC shape categories (including schistocytes, elliptocytes, sickle cells, spiculated cells, teardrops, micro/macrocytes). The main findings were that RBC-diff achieved high agreement with expert readers and clinical grading, and that objective schistocyte quantitation from CellaVision images improved diagnosis of thrombotic microangiopathy subtypes (iTTP/HUS) and predicted mortality risk, showing clear added diagnostic and prognostic value from digital RBC morphology at scale.

Summary: In this research study, CellaVision DM96 was used as the digital morphology platform for reviewing peripheral blood smears during a vaccine follow-up study. Alongside CBC measurements on a Sysmex XN-9000, the authors examined whole-blood smears with the CellaVision DM96 Digital Cell Morphology System to assess morphological changes over time after Pfizer BNT162b2 vaccination. The DM96 review supported their hematology findings by documenting low-frequency anomalous/atypical lymphocyte forms (including plasmacytoid variants) that increased after the first dose and then declined after the booster, helping connect routine blood-film morphology with the evolving immune response.

Summary: In this observational research article on hospitalized COVID-19 patients, the authors used a CellaVision DC-1 digital morphology analyzer to scan peripheral blood smears and generate automated 500-cell differentials with preclassification, which were then manually verified by hematopathologists; atypical plasmacytoid/immunoblastic lymphocytes (“covidocytes”) were counted as a distinct category. The smear morphology data from the DC-1 were combined with routine lab values to look for outcome-predictive patterns. The main findings were that COVID-19 patients showed reproducible peripheral-smear abnormalities, including frequent covidocytes, smudge cells, and platelet aggregates, and that a simple risk-stratification decision tree using RBC count, creatinine, urea, and covidocyte percentage predicted critical outcomes about as well as WHO clinical severity scoring or the neutrophil-to-lymphocyte ratio. The authors conclude that incorporating digital blood-morphology assessment (enabled by CellaVision DC-1) adds clinical value for triage and prognostication in hospitalized COVID-19.

Summary: In this article, CellaVision was used to digitally image and archive Wright–Giemsa–stained peripheral blood smears from patients before and after various CAR-T therapies. The smears were scanned on a CellaVision DM1200 and analyzed with CellaVision’s peripheral blood smear software, which automatically sorted WBCs into 18 categories; the authors then focused on the lymphocyte and atypical-lymphocyte image groups and manually reviewed the archived images over multiple days post-infusion. Using these serial CellaVision images, they discovered a distinct, time-ordered sequence of three atypical lymphocyte morphologies (early, mid, late) that appeared after CAR-T infusion in responders and tracked the expansion/contracting phases of CAR-T cells. This characteristic morphologic progression was largely absent in non-responders and in transplant/donor-lymphocyte controls, supporting that the digital morphology readout can serve as a practical surrogate for CAR-T expansion and early response assessment.

Summary: In this short communication, the authors used a CellaVision analyzer as one of three automated systems to perform differential counts of lymphocytes and neutrophils in mouse blood, and they compared those results against manual microscopic counting as the gold standard. Slides were prepared on an automated slide maker, then CellaVision analyzer was run per manufacturer instructions on small-volume mouse blood samples, enabling digital morphology–based counting alongside ADVIA 2120i and Horiba Yumizen H2500. The main finding was that CellaVision showed the closest agreement with manual counts for both lymphocytes (low bias ~4.75) and neutrophils (bias ~0.68), whereas ADVIA agreed only for lymphocytes and Horiba agreed for neither cell type. Based on these Bland–Altman and Passing–Bablok comparisons, the authors concluded that CellaVision can reliably be used for neutrophil and lymphocyte differentials in preclinical mouse studies, with the added practical advantage of requiring very small blood volumes.

Summary: In this paper, CellaVision is used as the source of the image dataset for training and evaluating the model. The authors worked with 10,299 individual white blood cell images acquired using a CellaVision DM96 analyzer in the core lab at Hospital Clinic of Barcelona, with cells annotated by expert clinical pathologists. They then trained a convolutional neural network with data augmentation on these CellaVision images to classify five WBC subtypes (neutrophils, eosinophils, basophils, lymphocytes, monocytes). Using this CellaVision-derived dataset, they achieved about 96% average classification accuracy, concluding that automated WBC classification from DM96 images can support faster, safer, and more consistent hematology diagnostics.

Summary: In this article the aim was to do a broad, high-quality comparison of routine CSF markers of neuroinflammation and blood–brain barrier (BBB) dysfunction in people with recent-onset psychotic disorders versus matched healthy controls. A CellaVision DM96 was used specifically for differential counting and classification of cells in cerebrospinal fluid (CSF). After CSF and blood cell counts were run on a Sysmex XN9000, the CSF cell type distribution was analyzed using a DM96, with lab staff blinded to patient/control status. This DM96-based CSF differential supported the study’s neuroinflammation/BBB biomarker outcomes by providing standardized, reviewable CSF cytology alongside the routine CSF WBC measures. They found no significant shift in overall CSF cell-type percentages between patients with recent-onset psychosis and healthy controls (e.g., lymphocytes and macrophages proportions were similar). However, alongside this stable cell-type distribution, the study still detected signs of a subgroup-level immune/BBB signal: more patients had mildly elevated CSF WBCs and patients showed higher BBB permeability markers.

Summary: In this article the authors investigated whether COVID-19 patients with reactive lymphocytes (COVID-19 RL) show clinical or biological characteristics linked to disease outcome. In addition, they also aimed to develop an automatic, objective system to recognize these COVID-19 reactive lymphocytes on blood smears, and to study their immunophenotype to understand their role in patient prognosis. Peripheral blood smears from COVID-19 patients were stained and then digitally imaged using the CellaVision DM96 system. The DM96 provided standardized cell images that pathologists reviewed to identify atypical “COVID-19 reactive lymphocytes,” and those same DM96 images formed the dataset for training and testing their convolutional-neural-network classifier. In short, CellaVision was the image-capture platform enabling both morphological assessment and the AI-based automatic recognition used in the study.

Summary: The study aimed to evaluate whether high-resolution peripheral blood smear images could be used to detect COVID-19 on a per-patient basis using a novel multiple-instance learning approach. All blood smear images in the study were acquired with the CellaVision DM9600 digital slide scanner. The large, consistent image sets produced by the DM9600 enabled the authors to apply multiple-instance deep learning, aggregating evidence across many cells per patient for COVID-19 screening. The systems uniform imaging geometry also supported interpretability experiments—such as masking cell centers versus edges—to probe where diagnostic signal originated within the scans. The authors concluded that their MIL/DOBA pipeline can screen for COVID-19 with high diagnostic performance (about 79% accuracy and ROC-AUC ≈ 0.90), indicating a genuine morphological association between COVID-19 and blood cells; they also found that performance improved when combining information from both red and white blood cells.

Summary: In this research study, CellaVision DM96 was used as the digital morphology platform for reviewing peripheral blood smears during a vaccine follow-up study. Alongside CBC measurements on a Sysmex XN-9000, the authors examined whole-blood smears with the CellaVision DM96 Digital Cell Morphology System to assess morphological changes over time after Pfizer BNT162b2 vaccination. The DM96 review supported their hematology findings by documenting low-frequency anomalous/atypical lymphocyte forms (including plasmacytoid variants) that increased after the first dose and then declined after the booster, helping connect routine blood-film morphology with the evolving immune response.

Summary: In this observational research article on hospitalized COVID-19 patients, the authors used a CellaVision DC-1 digital morphology analyzer to scan peripheral blood smears and generate automated 500-cell differentials with preclassification, which were then manually verified by hematopathologists; atypical plasmacytoid/immunoblastic lymphocytes (“covidocytes”) were counted as a distinct category. The smear morphology data from the DC-1 were combined with routine lab values to look for outcome-predictive patterns. The main findings were that COVID-19 patients showed reproducible peripheral-smear abnormalities, including frequent covidocytes, smudge cells, and platelet aggregates, and that a simple risk-stratification decision tree using RBC count, creatinine, urea, and covidocyte percentage predicted critical outcomes about as well as WHO clinical severity scoring or the neutrophil-to-lymphocyte ratio. The authors conclude that incorporating digital blood-morphology assessment (enabled by CellaVision DC-1) adds clinical value for triage and prognostication in hospitalized COVID-19.

Summary: In this article, CellaVision was used to digitally image and archive Wright–Giemsa–stained peripheral blood smears from patients before and after various CAR-T therapies. The smears were scanned on a CellaVision DM1200 and analyzed with CellaVision’s peripheral blood smear software, which automatically sorted WBCs into 18 categories; the authors then focused on the lymphocyte and atypical-lymphocyte image groups and manually reviewed the archived images over multiple days post-infusion. Using these serial CellaVision images, they discovered a distinct, time-ordered sequence of three atypical lymphocyte morphologies (early, mid, late) that appeared after CAR-T infusion in responders and tracked the expansion/contracting phases of CAR-T cells. This characteristic morphologic progression was largely absent in non-responders and in transplant/donor-lymphocyte controls, supporting that the digital morphology readout can serve as a practical surrogate for CAR-T expansion and early response assessment.

Summary: In this short communication, the authors used a CellaVision analyzer as one of three automated systems to perform differential counts of lymphocytes and neutrophils in mouse blood, and they compared those results against manual microscopic counting as the gold standard. Slides were prepared on an automated slide maker, then CellaVision analyzer was run per manufacturer instructions on small-volume mouse blood samples, enabling digital morphology–based counting alongside ADVIA 2120i and Horiba Yumizen H2500. The main finding was that CellaVision showed the closest agreement with manual counts for both lymphocytes (low bias ~4.75) and neutrophils (bias ~0.68), whereas ADVIA agreed only for lymphocytes and Horiba agreed for neither cell type. Based on these Bland–Altman and Passing–Bablok comparisons, the authors concluded that CellaVision can reliably be used for neutrophil and lymphocyte differentials in preclinical mouse studies, with the added practical advantage of requiring very small blood volumes.

Summary: In this paper, CellaVision is used as the source of the image dataset for training and evaluating the model. The authors worked with 10,299 individual white blood cell images acquired using a CellaVision DM96 analyzer in the core lab at Hospital Clinic of Barcelona, with cells annotated by expert clinical pathologists. They then trained a convolutional neural network with data augmentation on these CellaVision images to classify five WBC subtypes (neutrophils, eosinophils, basophils, lymphocytes, monocytes). Using this CellaVision-derived dataset, they achieved about 96% average classification accuracy, concluding that automated WBC classification from DM96 images can support faster, safer, and more consistent hematology diagnostics.

Summary: In this article the aim was to do a broad, high-quality comparison of routine CSF markers of neuroinflammation and blood–brain barrier (BBB) dysfunction in people with recent-onset psychotic disorders versus matched healthy controls. A CellaVision DM96 was used specifically for differential counting and classification of cells in cerebrospinal fluid (CSF). After CSF and blood cell counts were run on a Sysmex XN9000, the CSF cell type distribution was analyzed using a DM96, with lab staff blinded to patient/control status. This DM96-based CSF differential supported the study’s neuroinflammation/BBB biomarker outcomes by providing standardized, reviewable CSF cytology alongside the routine CSF WBC measures. They found no significant shift in overall CSF cell-type percentages between patients with recent-onset psychosis and healthy controls (e.g., lymphocytes and macrophages proportions were similar). However, alongside this stable cell-type distribution, the study still detected signs of a subgroup-level immune/BBB signal: more patients had mildly elevated CSF WBCs and patients showed higher BBB permeability markers.

Summary: In this study, CellaVision was used for the manual white-blood-cell differential: experienced technicians reviewed May-Grünwald-Giemsa–stained neonatal blood smears on a CellaVision DM96 analyser, which automatically imaged and presented cells until 100 leukocytes were classified per smear. The DM96-based counts were then compared with automated Sysmex XN-9000 results to evaluate how well automated immature-granulocyte metrics diagnose neonatal infection.

Summary: In this article the authors investigated whether COVID-19 patients with reactive lymphocytes (COVID-19 RL) show clinical or biological characteristics linked to disease outcome. In addition, they also aimed to develop an automatic, objective system to recognize these COVID-19 reactive lymphocytes on blood smears, and to study their immunophenotype to understand their role in patient prognosis. Peripheral blood smears from COVID-19 patients were stained and then digitally imaged using the CellaVision DM96 system. The DM96 provided standardized cell images that pathologists reviewed to identify atypical “COVID-19 reactive lymphocytes,” and those same DM96 images formed the dataset for training and testing their convolutional-neural-network classifier. In short, CellaVision was the image-capture platform enabling both morphological assessment and the AI-based automatic recognition used in the study.

Summary: The study aimed to evaluate whether high-resolution peripheral blood smear images could be used to detect COVID-19 on a per-patient basis using a novel multiple-instance learning approach. All blood smear images in the study were acquired with the CellaVision DM9600 digital slide scanner. The large, consistent image sets produced by the DM9600 enabled the authors to apply multiple-instance deep learning, aggregating evidence across many cells per patient for COVID-19 screening. The systems uniform imaging geometry also supported interpretability experiments—such as masking cell centers versus edges—to probe where diagnostic signal originated within the scans. The authors concluded that their MIL/DOBA pipeline can screen for COVID-19 with high diagnostic performance (about 79% accuracy and ROC-AUC ≈ 0.90), indicating a genuine morphological association between COVID-19 and blood cells; they also found that performance improved when combining information from both red and white blood cells.

Summary: In this method-comparison study, the CellaVision DC-1 digital cell imaging analyzer was evaluated against both the Sysmex DI-60 and manual microscopy using 200 May–Grünwald–Giemsa blood smears (100 normal, 100 abnormal). The DC-1 scanned each slide, captured high-resolution images, and automatically preclassified 200 white blood cells into standard categories before a trained morphologist performed postclassification review. Its performance in WBC differential counting and RBC morphological characterization was assessed through accuracy, within-run precision, clinical sensitivity/specificity, and agreement analyses. The DC-1 showed good correlation with DI-60 and manual microscopy for major WBC classes, high reproducibility, and strong agreement for several RBC morphology categories. Overall, the study concluded that DC-1 is an accurate, compact, and user-friendly digital morphology system suitable for low-volume or satellite laboratories, offering reliable digital smear evaluation with remote-consultation capability.

Summary: In this study, the CellaVision DC-1 was evaluated as a compact digital morphology analyzer for small–medium labs. The authors used DC-1 to scan peripheral blood smears and automatically preclassify ~210 cells into 18 classes (12 WBC + 6 non-WBC), and then had hematology experts verify/reclassify the pre-calls; these results were compared with 200-cell manual differentials and turnaround time measurements. DC-1 showed excellent precision (%CV 0.0–3.5%) and very high specificity/NPV across classes, and its WBC differential counts matched manual microscopy closely (largest mean difference ~2.74%). However, the positive predictive value was 0% for seven rare/pathologic classes (e.g., blasts, promyelocytes, NRBCs), meaning DC-1 tended to mislabel these and still required expert verification even on normal slides. Total turnaround time was the same as manual counting (~8:55), so the conclusion is that DC-1 is reliable for routine assistive PBS review in smaller labs but doesn’t inherently shorten TAT and may add verification workload for some cell categories.

Summary: In the veterinary method-comparison study, CellaVision was used to produce the manual reference leukocyte differential against which the new point-of-care analyzer (scil vCell 5) was judged. For every canine and feline blood sample, the lab made May–Grünwald/Giemsa-stained smears and then ran a 200-cell differential on a CellaVision DM1200 using the CellaVision DM veterinary software. The DM1200 scanned the smear, located the monolayer, captured WBC images at high magnification, and pre-classified cells automatically based on morphology; an expert reviewer then validated and reclassified any misidentified cells before finalizing the manual differential. These CellaVision-supported manual counts (NEU, LYM, MON, EOS, BASO) served as a key gold-standard comparator alongside the ADVIA 2120, letting the authors quantify correlation, bias, and total error for the scil vCell 5. In short, CellaVision’s role was not to generate the test results, but to provide a standardized, digital, expert-verified reference differential for robust evaluation of analyzer performance in dogs and cats.

Summary: In this study, the Sysmex DI-60 digital morphology analyzer was used to perform body-fluid (CSF, pleural fluid, ascites) cell differentials on cytospin slides. The DI-60 scanned the entire smear digitally, pre-classified about 250 nucleated cells per slide into leukocyte categories (neutrophils, lymphocytes, eosinophils, macrophages/monocytes, and “other” cells), and also sorted non-leukocyte items like smudge cells and artifacts; a hematology expert then verified and reclassified the pre-classification. These DI-60 verified results were compared against manual microscopy and the Sysmex XN-350 to evaluate accuracy, agreement, and workflow suitability. The main takeaway was that DI-60 delivered reliable differentials for the common BF cell types, with performance improving after human verification, supporting its use as an efficient screening/assistive digital morphology method in BF analysis.

Summary: In this research article, the authors used the Sysmex DI-60 digital cell imaging analyzer to automatically scan peripheral blood smears, capture cell images, and perform two tasks: pre-characterization of RBC morphology and pre-classification of WBC differentials (200 cells/slide), with technologists then verifying/reclassifying results before reporting. They evaluated DI-60 on 822 routine smears, directly comparing its outputs to manual light-microscope review and also measuring total assay time versus hands-on verification time. The main findings were that the DI-60 showed acceptable performance on normal samples and good accuracy for some RBC features (polychromasia, target cells, ovalocytes), but it struggled with several abnormalities—most notably very low specificity for schistocytes (many false positives) and weaker performance for atypical lymphocytes, basophils, and some immature cells unless carefully verified. Even though DI-60’s total run time was longer than manual review, the hands-on time was ~144 seconds per slide shorter, indicating a real workflow-efficiency benefit when used with expert verification, especially in high-volume labs.

Summary: In this article, CellaVision instruments were evaluated as part of a multicenter assessment of digital morphology performance. Seven Italian labs received standardized May–Grünwald–Giemsa peripheral blood smears and analyzed them both by manual optical microscopy (the reference method) and by automated image-processing systems, specifically either the Sysmex DI-60 or the CellaVision DM96 running CellaVision DM Software. The CellaVision analyzers performed automated slide scanning and pre-classification of leukocytes into standard WBC categories, after which skilled morphologists reviewed and reclassified cells as needed; results from pre-classification and post-reclassification were compared to manual differentials. The main findings were that digital morphology systems can improve efficiency and objectivity, especially when extended counts are needed for rare cells, but variability remains for low-frequency cell types; neutrophil counts met desirable imprecision targets, while several other classes (e.g., blasts, basophils, NRBCs) showed poorer reproducibility across labs. Importantly, post-reclassification digital review was faster than manual microscopy (about 191 s vs 326 s per slide), so the authors conclude that digital systems such as CellaVision analyzers are valuable for saving time and enabling larger cell counts, which in turn can reduce statistical imprecision in routine morphology work.

Summary: In this study, the authors used CellaVision DM96 in digital slide mode as the image-capture platform for bone marrow smears. They stained 43 bone-marrow films with May–Grünwald Giemsa, scanned them on DM96, and then selected well-focused microscopic fields (901×823 px) from the digital slides; these fields were cropped into 704×704 images and further segmented into single-cell crops to build a large dataset for deep-learning training and testing. The main finding was that semi-supervised learning can greatly expand labeled training data for bone-marrow cell classification, and that the combined method (confirmed self-training + active learning) performed best. After 25 rounds, CST+AL increased training images to ~47,843 and achieved the highest test accuracy (~0.976, with mean recall/precision ~0.976), outperforming either method alone. Errors mainly occurred among closely related granulocyte maturation stages (metamyelocytes, band neutrophils, segmented neutrophils), suggesting these are inherently harder to separate for both humans and AI.

Summary: In this article, the authors describes an solution for implementing a costumized middelware. They used a DI-60 analyzer as part of their routine CBC/CBCD workflow together with Sysmex XN9000 analyzers, all connected through Data Innovations Instrument Manager middleware. The DI-60 supplied automated digital differentials and morphology images when slides were generated or when middleware rules flagged samples for microscopic review, effectively acting as the lab’s digital microscopy step within a largely paperless, rules-driven process. While the articles main focus is middleware, the DI-60 is the CellaVision instrument enabling the digital slide review that the rules either trigger, suppress, or streamline. The authors report that this integrated system achieved very high autoverification rates and reduced unnecessary slide work, showing how DI-60-based digital morphology fits into a modern, efficiency-optimized hematology workflow.

Summary: In this study, CellaVision images were used as a morphological reference method for body-fluid differentials. Cytospin slides from cerebrospinal and ascitic fluids were scanned on a DM96 slide scanning unit running the CellaVision DM1200 Body Fluid Application, which preclassified 200 nucleated cells per sample. Two experienced technologists then manually reviewed and corrected these preclassifications, and the resulting mononuclear (MN%) and polymorphonuclear (PMN%) proportions were calculated. The aim was to determine how well instrumental body-fluid differentials correlate with morphology by comparing Sysmex XN BF-mode results against conventional optical microscopy, multicolor flow cytometry, and CellaVision digital differentials in tumor-free fluids. The main finding was strong agreement across methods: CellaVision MN% and PMN% closely matched manual microscopy and flow cytometry with high correlations and only small biases, leading the authors to conclude that instrumental BF differentials are reliable when malignant cells are absent and that CellaVision can effectively support or replace manual microscopy in routine practice, including for remote validation and as a reflex tool when analyzer flags suggest atypical cells.

Summary: In this article, they are using CellaVision digital morphology software to evaluate a UKNEQAS external quality assessment (EQA) peripheral blood film, alongside conventional microscopy. They emphasize that unlike manual microscopy—where operators choose fields subjectively—CellaVision automatically selects key areas and requires the reviewer to systematically examine WBC, RBC, and platelet image sets, which can reduce “attention drift” toward only one abnormality. In the featured EQA case (acute leukemia plus sickle cell disease), most microscope users correctly saw blasts but missed sickle cells; by contrast, the CellaVision RBC screen made sickle and boat-shaped cells easy to spot, so their lab detected the additional diagnosis before the official UKNEQAS report. The main point is that CellaVision added diagnostic value in EQA review by preventing important RBC morphology from being masked by striking WBC findings, supporting broader, more reliable smear assessment.

Summary: In this research article, the authors used the Sysmex DI-60 digital cell imaging analyzer to automatically scan peripheral blood smears, capture cell images, and perform two tasks: pre-characterization of RBC morphology and pre-classification of WBC differentials (200 cells/slide), with technologists then verifying/reclassifying results before reporting. They evaluated DI-60 on 822 routine smears, directly comparing its outputs to manual light-microscope review and also measuring total assay time versus hands-on verification time. The main findings were that the DI-60 showed acceptable performance on normal samples and good accuracy for some RBC features (polychromasia, target cells, ovalocytes), but it struggled with several abnormalities—most notably very low specificity for schistocytes (many false positives) and weaker performance for atypical lymphocytes, basophils, and some immature cells unless carefully verified. Even though DI-60’s total run time was longer than manual review, the hands-on time was ~144 seconds per slide shorter, indicating a real workflow-efficiency benefit when used with expert verification, especially in high-volume labs.

Summary: In this article, CellaVision instruments were evaluated as part of a multicenter assessment of digital morphology performance. Seven Italian labs received standardized May–Grünwald–Giemsa peripheral blood smears and analyzed them both by manual optical microscopy (the reference method) and by automated image-processing systems, specifically either the Sysmex DI-60 or the CellaVision DM96 running CellaVision DM Software. The CellaVision analyzers performed automated slide scanning and pre-classification of leukocytes into standard WBC categories, after which skilled morphologists reviewed and reclassified cells as needed; results from pre-classification and post-reclassification were compared to manual differentials. The main findings were that digital morphology systems can improve efficiency and objectivity, especially when extended counts are needed for rare cells, but variability remains for low-frequency cell types; neutrophil counts met desirable imprecision targets, while several other classes (e.g., blasts, basophils, NRBCs) showed poorer reproducibility across labs. Importantly, post-reclassification digital review was faster than manual microscopy (about 191 s vs 326 s per slide), so the authors conclude that digital systems such as CellaVision analyzers are valuable for saving time and enabling larger cell counts, which in turn can reduce statistical imprecision in routine morphology work.

Summary: In this study, the authors used CellaVision DM96 in digital slide mode as the image-capture platform for bone marrow smears. They stained 43 bone-marrow films with May–Grünwald Giemsa, scanned them on DM96, and then selected well-focused microscopic fields (901×823 px) from the digital slides; these fields were cropped into 704×704 images and further segmented into single-cell crops to build a large dataset for deep-learning training and testing. The main finding was that semi-supervised learning can greatly expand labeled training data for bone-marrow cell classification, and that the combined method (confirmed self-training + active learning) performed best. After 25 rounds, CST+AL increased training images to ~47,843 and achieved the highest test accuracy (~0.976, with mean recall/precision ~0.976), outperforming either method alone. Errors mainly occurred among closely related granulocyte maturation stages (metamyelocytes, band neutrophils, segmented neutrophils), suggesting these are inherently harder to separate for both humans and AI.

Summary: In this article, the authors describes an solution for implementing a costumized middelware. They used a DI-60 analyzer as part of their routine CBC/CBCD workflow together with Sysmex XN9000 analyzers, all connected through Data Innovations Instrument Manager middleware. The DI-60 supplied automated digital differentials and morphology images when slides were generated or when middleware rules flagged samples for microscopic review, effectively acting as the lab’s digital microscopy step within a largely paperless, rules-driven process. While the articles main focus is middleware, the DI-60 is the CellaVision instrument enabling the digital slide review that the rules either trigger, suppress, or streamline. The authors report that this integrated system achieved very high autoverification rates and reduced unnecessary slide work, showing how DI-60-based digital morphology fits into a modern, efficiency-optimized hematology workflow.

Summary: In this study, CellaVision images were used as a morphological reference method for body-fluid differentials. Cytospin slides from cerebrospinal and ascitic fluids were scanned on a DM96 slide scanning unit running the CellaVision DM1200 Body Fluid Application, which preclassified 200 nucleated cells per sample. Two experienced technologists then manually reviewed and corrected these preclassifications, and the resulting mononuclear (MN%) and polymorphonuclear (PMN%) proportions were calculated. The aim was to determine how well instrumental body-fluid differentials correlate with morphology by comparing Sysmex XN BF-mode results against conventional optical microscopy, multicolor flow cytometry, and CellaVision digital differentials in tumor-free fluids. The main finding was strong agreement across methods: CellaVision MN% and PMN% closely matched manual microscopy and flow cytometry with high correlations and only small biases, leading the authors to conclude that instrumental BF differentials are reliable when malignant cells are absent and that CellaVision can effectively support or replace manual microscopy in routine practice, including for remote validation and as a reflex tool when analyzer flags suggest atypical cells.

Summary: In this study, Sysmex DI-60 was used as an automated digital morphology analyzer to support and potentially replace some manual smear reviews once new review rules were optimized for a women’s and children’s hospital. After CBCs were run on a Sysmex XN-9000 and slides were made with an SP-10 stainer, 200 validation samples that triggered the new rules were scanned on the DI-60; its preclassification of WBCs, RBC morphology, and platelet findings was then compared with expert manual review/reclassification as the reference. The authors found high consistency (>~85–88%) between DI-60 preclassification and manual microscopy for RBC- and platelet-related rules, meaning DI-60 results could often be reported directly for those triggers, and DI-60 also correlated well with manual counts for common mature WBC types. However, when flags for blasts/abnormal lymphocytes or atypical lymphocytes were triggered, DI-60 preclassification was not reliable enough to ensure leukemia wouldn’t be missed, so direct microscopic review remained necessary. Overall, the paper positions DI-60 as a strong labor-saving digital alternative for many rule-triggered smear reviews in these specific patient populations, while still requiring expert microscopy for high-risk abnormal WBC flags.

Summary: In this Letter to the Editor, they used Sysmex DI-60 digital morphology system, which was part of an integrated XN-9000 line (XN analyzer + SP-10 slide maker/stainer + DI-60). For both reported patients, the XN-9000 CBC showed marked monocytosis, and the DI-60 pre-classified a high proportion of cells as monocytes, in agreement with the analyzer differential. Manual smear review, however, showed these “monocytes” were actually dysplastic macropolycytes (giant neutrophils) with neutrophil-lineage features, and myeloperoxidase staining on DI-60 images confirmed strong MPO positivity, proving they were granulocytes rather than monocytes. So in this article, DI-60 is used as the digital imaging and preclassification tool whose misclassification exposes a pitfall: in severe granulocytic dysplasia, both the XN analyzer and DI-60 can report pseudo-monocytosis, meaning expert visual reclassification and, when needed, confirmatory stains remain essential.

Summary: In this SARS-CoV-2 / Plasmodium falciparum coinfection case report, the lab used a CellaVision DM1200 to examine the peripheral blood smear and perform the white blood cell differential. The DM1200’s digital images clearly showed erythrocytic inclusions of P. falciparum (signet-ring forms) along with atypical lymphocytes compatible with COVID-19. These findings supported the rapid confirmation of malaria alongside COVID-19 and helped guide prompt combined therapy.

Summary: In this Kimura disease case report, CellaVision was used to confirm and document the patient’s persistent eosinophilia on a peripheral blood smear. After the ADVIA 2120i CBC showed very high eosinophil counts, the lab performed a smear review with CellaVision DM96, which digitally captured and displayed eosinophils for verification. The DM96 images showed marked eosinophilia without morphological abnormalities. This smear confirmation helped rule out dysplastic or malignant eosinophil features and supported the broader diagnostic workup that eventually led to Kimura disease.

Summary: In this Letter to the Editor, CellaVision was used to review the peripheral blood smear after an automated analyzer (ADVIA 2120) reported apparent basophilia. The CellaVision smear imaging showed that the “basophils” flagged by ADVIA were actually flower cells (atypical, lobulated lymphoma cells) and that true basophils were absent. This digital morphology finding helped explain the pseudobasophilia and supported the diagnosis of HTLV-1–associated adult T-cell leukemia/lymphoma.

Summary: In this case report, the CellaVision Digital Morphology System was used during peripheral blood smear review to capture rare neutrophils containing yeast-like inclusions that were partly obscured by toxic granulation. The system initially preclassified these cells into non-specific buckets (“unidentified/other”) and even labeled a cluster of free yeasts as “artifact,” but crucially it still flagged and presented them for hematopathologist review. On the CellaVision images, about 2–4% of neutrophils showed ovoid organisms with crescentic nuclei and capsule-like halos, morphologically suggestive of Histoplasma capsulatum, which prompted confirmatory antigen/serology testing and culture. The main finding is that CellaVision-enabled digital review helped detect disseminated histoplasmosis directly from blood smear morphology, a pattern that can be missed on manual review, enabling timely antifungal treatment.

Summary: In this correspondence, the lab describes how they use the Sysmex DI-60 with CellaVision software for routine digital blood-smear review across a hub-and-spoke network, with remote assessment enabled. The DI-60/CellaVision system was set to automatic preclassification of 200 WBCs into 14 leukocyte subtypes plus 5 “non-leukocyte” categories (erythroblasts, giant platelets, platelet aggregates, smudge cells, artefacts). The key point is that the authors systematically review not only WBC categories but also the “non-leukocyte” buckets, because important pathology can be hidden there. By carefully checking sections like “thrombocyte aggregation” and “artefact,” they unexpectedly detected clinically crucial findings—e.g., Plasmodium-parasitized RBCs, bacteria phagocytosed by neutrophils and free cocci (sepsis), microfilaria, fungal elements including Fusarium and Candida spores, and cryoglobulins mixed with platelet clumps—often enabling earlier diagnosis and faster clinical action. Their conclusion is that digital morphology can surface rare patterns outside the main WBC sort, but expert review of all CellaVision categories remains essential to avoid missing these “hidden” signals.

Summary: In this Letter to the Editor on Southeast Asian ovalocytosis (SAO) in a critically ill COVID-19 patient, the authors used CellaVision DM96 to perform the peripheral blood smear morphology review. The DM96 digital analysis revealed the characteristic red-cell pattern—numerous macro-ovalocytes, stomato-ovalocytes, and stomatocytes with longitudinal/transverse “stomas,” sometimes multiple per cell, which was distinctive enough to raise the diagnosis of SAO. Finding SAO on the DM96 smear prompted confirmatory testing, including molecular demonstration of the SLC4A1 deletion and ektacytometry. The main takeaway is that DM96-enabled digital morphology was the key trigger for diagnosing an otherwise unsuspected congenital RBC membrane disorder, and the authors suggest SAO plus COVID-19 may synergistically worsen RBC deformability and potentially increase thrombotic risk.

Summary: In this case report, the authors used a CellaVision DC-1 to obtain and document digital peripheral-blood images of abnormal lymphocytes that appeared after cord-blood transplantation in patients with Adult T-cell leukemia-lymphoma (ATL). The CellaVision images provided clear morphology evidence of ATL-like lymphocytes in the blood and helped trigger further diagnostic work-up when relapse was suspected. On deeper analysis (flow cytometry, chimerism testing, and inverse PCR), they found the abnormal cells were donor-derived and polyclonal HTLV-1–infected T cells, not true recurrent adult T-cell leukemia/lymphoma. The main conclusion is that this represented pseudo-progression after transplant, and that when abnormal lymphocytes are detected post-HCT, clinicians should assess chimerism/clonality before assuming relapse—even if morphology on CellaVision looks ATL-like.

Summary: In this article on a locally acquired babesiosis case, CellaVision was used to capture and display abnormal cells on the peripheral blood smear. The key diagnostic images of Babesia microti on the Wright-stained smear were shown in Figure 1; panels A and B are conventional Wright’s-stain views, while panel C was specifically captured and displayed using the CellaVision DM96 system together with CellaVision Remote Review Software v6.0.1. So, CellaVision’s role here was as the digital microscopy/remote review platform that documented the characteristic intra-erythrocytic ring forms and merozoites (including the Maltese-cross tetrad), supporting rapid recognition of babesiosis before PCR confirmation.

Summary: In this IgE multiple myeloma case report, CellaVision was used for peripheral blood morphology review as part of the initial hematology workup. After the CBC was run on a Sysmex XN-9100, the authors assessed the blood smear with CellaVision digital microscopy, which documented that the white cell differential had no significant abnormalities or circulating plasma cells, while the red-cell series showed rouleaux formation. That smear finding supported the suspicion of a monoclonal gammopathy and complemented the protein studies (capillary electrophoresis and immunofixation) that ultimately identified an IgE-kappa monoclonal component. In short, CellaVision served as the digital morphology tool to verify blood-film features (notably rouleaux and absence of plasma cells) within the diagnostic pathway of this rare myeloma subtype.

Summary: In this Letter to the Editor, CellaVision DM9600 was used as the digital morphology step that enabled early recognition of candidemia. The lab’s routine CBC on a Beckman Coulter DxH 800 showed abnormal scatterplot “clouds” suggestive of fungi, so a peripheral blood smear was made and then reviewed on a CellaVision DM9600 (alongside a manual Nikon microscope). On the DM9600 images, they directly observed yeast with budding forms and pseudohyphae, both extracellularly and phagocytosed inside neutrophils, which allowed them to alert clinicians immediately and trigger confirmatory fungal blood cultures. Culture later identified Candida parapsilosis, and targeted antifungal therapy was started days earlier than it otherwise would have been, leading to clinical improvement and discharge. The main message is that DM9600-based smear review can catch circulating fungi very early when automated analyzers raise suspicion, supporting faster diagnosis and treatment of severe nosocomial fungal infections.

Summary: In this case report, the authors used a DI-60 analyzer to capture and document high-magnification digital images of the patient’s peripheral blood smear. The DI-60 images clearly showed numerous yeast-like organisms with central septa both inside neutrophils and extracellularly, which allowed the laboratory to make a rapid presumptive diagnosis of disseminated Talaromyces marneffei infection directly from the smear. This DI-60-supported diagnosis was then confirmed by blood culture. The main takeaway is that digital smear review on DI-60 enabled early recognition of a life-threatening opportunistic fungal infection in an AIDS patient, highlighting the value of careful peripheral smear examination for fungi in septic HIV cases.

Summary: In this case report, the authors used the DI-60 to digitally identify and document peripheral blood smear morphology in a patient with suspected hemorrhagic fever with renal syndrome (HFRS). The DI-60 images showed left-shifted neutrophils with reactive changes, reactive plasmacytoid lymphocytes (some binucleated), and increased smudge cells, which are all identified by the DI-60. The main finding was that this morphology—together with CBC abnormalities (high WBC, severe thrombocytopenia) and a characteristic cell counter scattergram—strongly pointed to hantavirus infection/HFRS, later confirmed by positive hantavirus IgM/IgG.

Summary: In this case report, the CellaVision Digital Morphology System was used during peripheral blood smear review to capture rare neutrophils containing yeast-like inclusions that were partly obscured by toxic granulation. The system initially preclassified these cells into non-specific buckets (“unidentified/other”) and even labeled a cluster of free yeasts as “artifact,” but crucially it still flagged and presented them for hematopathologist review. On the CellaVision images, about 2–4% of neutrophils showed ovoid organisms with crescentic nuclei and capsule-like halos, morphologically suggestive of Histoplasma capsulatum, which prompted confirmatory antigen/serology testing and culture. The main finding is that CellaVision-enabled digital review helped detect disseminated histoplasmosis directly from blood smear morphology, a pattern that can be missed on manual review, enabling timely antifungal treatment.

Summary: In this correspondence, the lab describes how they use the Sysmex DI-60 with CellaVision software for routine digital blood-smear review across a hub-and-spoke network, with remote assessment enabled. The DI-60/CellaVision system was set to automatic preclassification of 200 WBCs into 14 leukocyte subtypes plus 5 “non-leukocyte” categories (erythroblasts, giant platelets, platelet aggregates, smudge cells, artefacts). The key point is that the authors systematically review not only WBC categories but also the “non-leukocyte” buckets, because important pathology can be hidden there. By carefully checking sections like “thrombocyte aggregation” and “artefact,” they unexpectedly detected clinically crucial findings—e.g., Plasmodium-parasitized RBCs, bacteria phagocytosed by neutrophils and free cocci (sepsis), microfilaria, fungal elements including Fusarium and Candida spores, and cryoglobulins mixed with platelet clumps—often enabling earlier diagnosis and faster clinical action. Their conclusion is that digital morphology can surface rare patterns outside the main WBC sort, but expert review of all CellaVision categories remains essential to avoid missing these “hidden” signals.

Summary: In this Letter to the Editor on Southeast Asian ovalocytosis (SAO) in a critically ill COVID-19 patient, the authors used CellaVision DM96 to perform the peripheral blood smear morphology review. The DM96 digital analysis revealed the characteristic red-cell pattern—numerous macro-ovalocytes, stomato-ovalocytes, and stomatocytes with longitudinal/transverse “stomas,” sometimes multiple per cell, which was distinctive enough to raise the diagnosis of SAO. Finding SAO on the DM96 smear prompted confirmatory testing, including molecular demonstration of the SLC4A1 deletion and ektacytometry. The main takeaway is that DM96-enabled digital morphology was the key trigger for diagnosing an otherwise unsuspected congenital RBC membrane disorder, and the authors suggest SAO plus COVID-19 may synergistically worsen RBC deformability and potentially increase thrombotic risk.

Summary: In this case report, the authors used a CellaVision DC-1 to obtain and document digital peripheral-blood images of abnormal lymphocytes that appeared after cord-blood transplantation in patients with Adult T-cell leukemia-lymphoma (ATL). The CellaVision images provided clear morphology evidence of ATL-like lymphocytes in the blood and helped trigger further diagnostic work-up when relapse was suspected. On deeper analysis (flow cytometry, chimerism testing, and inverse PCR), they found the abnormal cells were donor-derived and polyclonal HTLV-1–infected T cells, not true recurrent adult T-cell leukemia/lymphoma. The main conclusion is that this represented pseudo-progression after transplant, and that when abnormal lymphocytes are detected post-HCT, clinicians should assess chimerism/clonality before assuming relapse—even if morphology on CellaVision looks ATL-like.

Summary: In this article on a locally acquired babesiosis case, CellaVision was used to capture and display abnormal cells on the peripheral blood smear. The key diagnostic images of Babesia microti on the Wright-stained smear were shown in Figure 1; panels A and B are conventional Wright’s-stain views, while panel C was specifically captured and displayed using the CellaVision DM96 system together with CellaVision Remote Review Software v6.0.1. So, CellaVision’s role here was as the digital microscopy/remote review platform that documented the characteristic intra-erythrocytic ring forms and merozoites (including the Maltese-cross tetrad), supporting rapid recognition of babesiosis before PCR confirmation.

Summary: In this IgE multiple myeloma case report, CellaVision was used for peripheral blood morphology review as part of the initial hematology workup. After the CBC was run on a Sysmex XN-9100, the authors assessed the blood smear with CellaVision digital microscopy, which documented that the white cell differential had no significant abnormalities or circulating plasma cells, while the red-cell series showed rouleaux formation. That smear finding supported the suspicion of a monoclonal gammopathy and complemented the protein studies (capillary electrophoresis and immunofixation) that ultimately identified an IgE-kappa monoclonal component. In short, CellaVision served as the digital morphology tool to verify blood-film features (notably rouleaux and absence of plasma cells) within the diagnostic pathway of this rare myeloma subtype.

Summary: In this Letter to the Editor, CellaVision DM9600 was used as the digital morphology step that enabled early recognition of candidemia. The lab’s routine CBC on a Beckman Coulter DxH 800 showed abnormal scatterplot “clouds” suggestive of fungi, so a peripheral blood smear was made and then reviewed on a CellaVision DM9600 (alongside a manual Nikon microscope). On the DM9600 images, they directly observed yeast with budding forms and pseudohyphae, both extracellularly and phagocytosed inside neutrophils, which allowed them to alert clinicians immediately and trigger confirmatory fungal blood cultures. Culture later identified Candida parapsilosis, and targeted antifungal therapy was started days earlier than it otherwise would have been, leading to clinical improvement and discharge. The main message is that DM9600-based smear review can catch circulating fungi very early when automated analyzers raise suspicion, supporting faster diagnosis and treatment of severe nosocomial fungal infections.

Summary: In this case report, the authors used a DI-60 analyzer to capture and document high-magnification digital images of the patient’s peripheral blood smear. The DI-60 images clearly showed numerous yeast-like organisms with central septa both inside neutrophils and extracellularly, which allowed the laboratory to make a rapid presumptive diagnosis of disseminated Talaromyces marneffei infection directly from the smear. This DI-60-supported diagnosis was then confirmed by blood culture. The main takeaway is that digital smear review on DI-60 enabled early recognition of a life-threatening opportunistic fungal infection in an AIDS patient, highlighting the value of careful peripheral smear examination for fungi in septic HIV cases.

Summary: In this case report, the authors used the DI-60 to digitally identify and document peripheral blood smear morphology in a patient with suspected hemorrhagic fever with renal syndrome (HFRS). The DI-60 images showed left-shifted neutrophils with reactive changes, reactive plasmacytoid lymphocytes (some binucleated), and increased smudge cells, which are all identified by the DI-60. The main finding was that this morphology—together with CBC abnormalities (high WBC, severe thrombocytopenia) and a characteristic cell counter scattergram—strongly pointed to hantavirus infection/HFRS, later confirmed by positive hantavirus IgM/IgG.

Summary: In the article, the authors used a CellaVision DM1200 analyzer to digitally capture peripheral blood smear images. These digital images helped confirm abnormal erytrocyte morphology when earlier functional tests were equivocal, prompting further targeted genetic workup. The main finding was a rare case of acquired spherocytosis due to a somatic SPTB mutation occurring alongside an MDS-associated U2AF1 mutation, with the latter likely reflecting clonal hematopoiesis rather than causing the hemolysis. Clinically, the patient improved rapidly after splenectomy, with hemoglobin and bilirubin normalizing, supporting that the spherocytosis (not MDS) drove the anemia.

Summary: In this review article, CellaVision is discussed as a prime example of machine-learning–enabled digital morphology already in routine clinical hematology. The author explains that when automated analyzers still require smear review, digitized peripheral blood smears can be scanned and pre-classified by ML models before a human looks at cells, and the CellaVision system is specifically cited as doing this by classifying findings into ~17 WBC types, including mature and immature forms, with newer modules for RBC morphological grading. The review summarizes published evaluations showing that CellaVision’s WBC preclassification is generally very accurate for the 5-part differential (high specificity and good sensitivity), but performance drops for immature or atypical cells, so expert verification remains necessary. Overall, CellaVision is presented as a key current-state technology that improves standardization and efficiency in smear morphology while still operating in a “human-in-the-loop” model.

Summary: In this Letter to the Editor, they are using CellaVision digital morphology software to evaluate a UKNEQAS external quality assessment (EQA) peripheral blood film, alongside conventional microscopy. They emphasize that unlike manual microscopy—where operators choose fields subjectively—CellaVision automatically selects key areas and requires the reviewer to systematically examine WBC, RBC, and platelet image sets, which can reduce “attention drift” toward only one abnormality. In the featured EQA case (acute leukemia plus sickle cell disease), most microscope users correctly saw blasts but missed sickle cells; by contrast, the CellaVision RBC screen made sickle and boat-shaped cells easy to spot, so their lab detected the additional diagnosis before the official UKNEQAS report. The main point is that CellaVision added diagnostic value in EQA review by preventing important RBC morphology from being masked by striking WBC findings, supporting broader, more reliable smear assessment.

Summary: In this study, CellaVision was used for the manual white-blood-cell differential: experienced technicians reviewed May-Grünwald-Giemsa–stained neonatal blood smears on a CellaVision DM96 analyser, which automatically imaged and presented cells until 100 leukocytes were classified per smear. The DM96-based counts were then compared with automated Sysmex XN-9000 results to evaluate how well automated immature-granulocyte metrics diagnose neonatal infection.

Summary: In this multicenter study, the authors compared the performance of two automated cell-morphology analyzers: the Mindray MC-100i and the Sysmex DI-60. The DI-60 and MC-100i arms were run in parallel: for each of the 432 venous blood samples, paired smears were prepared (using either Mindray’s SC-120 stainer or the Sysmex SP-10 slide stainer), and 200 WBCs per smear were collected and preclassified into standard WBC categories (plus NRBCs/unclassified). Two senior morphologists then reviewed and reclassified the preclassification results from both systems to generate postclassification outputs. Both pre- and postclassification results from each analyzer were compared with manual microscopy as the reference method. The main finding was that the DI-60 performed well for common, mature WBC types, but its preclassification accuracy—especially its sensitivity for abnormal or immature cells—was clearly lower than that of the Mindray MC-100i, meaning the DI-60 required more expert correction for rare or pathological cells.

Summary: In this study, CellaVision was not the primary analyzer being evaluated but served as a key part of the reference workflow. After running CBC and WBC differentials on the HemoScreen point-of-care device and the Sysmex XN, the lab prepared May–Grünwald-Giemsa blood films and performed a 200-cell manual morphology review; this review was done either by light microscopy or by digital microscopy on a CellaVision DI-60, which provided the digital smear images used to confirm whether abnormal leukocytes were truly present. Using this CellaVision-supported microscopy confirmation as the ground truth, the authors found that HemoScreen’s CBC and five-part differential were highly comparable to Sysmex XN for most parameters and that both systems detected all samples with true leukocyte abnormalities, but HemoScreen produced fewer false abnormal-WBC flags—meaning fewer smears needed CellaVision/manual review.

Summary: In this study, CellaVision was not the primary analyzer being evaluated but served as a key part of the reference workflow. After running CBC and WBC differentials on the HemoScreen point-of-care device and the Sysmex XN, the lab prepared May–Grünwald-Giemsa blood films and performed a 200-cell manual morphology review; this review was done either by light microscopy or by digital microscopy on a CellaVision DI-60, which provided the digital smear images used to confirm whether abnormal leukocytes were truly present. Using this CellaVision-supported microscopy confirmation as the ground truth, the authors found that HemoScreen’s CBC and five-part differential were highly comparable to Sysmex XN for most parameters and that both systems detected all samples with true leukocyte abnormalities, but HemoScreen produced fewer false abnormal-WBC flags—meaning fewer smears needed CellaVision/manual review.

Summary: In this analytical validation study, the Sysmex DI-60 digital morphology analyzer were evaluated as a potential replacement for manual light microscopy in routine CBC smear review. After CBCs were run, blood smears were May–Grünwald–Giemsa stained and automatically scanned on the DI-60, where the CellaVision-based software pre-classified leukocytes for reviewer verification and possible post-reclassification. They used the DI-60 outputs to measure WBC differentials (mature cells and detection of immature/pathologic forms) and to support platelet assessment, including platelet clump detection and rough platelet count estimation. DI-60 results were then compared against the gold-standard manual light-microscope differential (100 cells) and against automated differentials from the Sysmex XN-10 to judge precision, accuracy, and agreement. Overall, the study positions the DI-60/CellaVision digital workflow as a reliable, time-saving digital alternative for mature WBC differentials and platelet review, while noting that difficult or doubtful cases still require confirmation by light microscopy.

Summary: In this study, the Sysmex DI-60 analyzer and the CellaVision Advanced RBC Software were used to digitally evaluate red-blood-cell size heterogeneity on Wright–Giemsa–stained peripheral blood smears. The DI-60 scanned each smear, generated an RBC overview, and automatically quantified anisocytosis, microcytes (MIC), and macrocytes (MAC) (along with other RBC morphology features), providing objective size-based parameters that were then compared with Sysmex XN-9000 indices such as RDW-CV and MCV. The authors conclude that the DI-60 Advanced RBC Software provides an accurate and objective digital assessment of RBC size abnormalities. In particular, its anisocytosis grading aligns well with Sysmex RDW-CV and can reliably identify clinically relevant anisocytosis using a 15% cutoff, while the microcyte and macrocyte percentages predict low or high MCV with strong diagnostic performance. They emphasize that the Advanced RBC software is consistent and time-saving compared with manual estimation, but should be used as a complement to automated CBC indices and expert review rather than a standalone replacement.