Search

Philadelphia chromosome-negative myeloproliferative neoplasms (Ph-negative MPNs) such as polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis are characterized by abnormal proliferation of mature bone marrow cell lineages. Since various non-hematologic disorders can also cause leukocytosis, thrombocytosis and polycythemia, the detection of abnormal peripheral blood cells is essential for the diagnostic screening of Ph-negative MPNs. We sought to develop an automated diagnostic support system of Ph-negative MPNs. Our strategy was to combine the complete blood cell count and research parameters obtained by an automated hematology analyzer (Sysmex XN-9000) with morphological parameters that were extracted using a convolutional neural network deep learning system equipped with an Extreme Gradient Boosting (XGBoost)-based decision-making algorithm. The developed system showed promising performance in the differentiation of PV, ET, and MF with high accuracy when compared with those of the human diagnoses, namely: > 90% sensitivity and > 90% specificity. The calculated area under the curve of the ROC curves were 0.990, 0.967, and 0.974 for PV, ET, MF, respectively. This study is a step toward establishing a universal automated diagnostic system for all types of hematology disorders.

Background This study investigated the feasibility and accuracy of an automated hematology analyzer in the detection of schistocytes. Methods In total, 1,026 peripheral blood samples were collected. Schistocytes were morphologically diagnosed by manual examination of digital microscopic red blood cell images captured by a Sysmex DI-60. Automated diagnoses were performed using a Sysmex XN-3000. Result The accuracy of automated diagnosis using the XN-3000 with the default algorithm "fragments?" was determined through comparison with the findings of morphological examination. The comparison showed a sensitivity of 100% and a specificity of 41.6% for automated diagnosis. To improve the low specificity, a two-step analysis was performed. Use of the algorithm "fragments?" in XN-3000 followed by an off-line analysis using the cell parameter %FRC (percent fragmented red blood cells) yielded a sensitivity of 86.5% and a specificity of 70.3%. Conclusions Our study indicated that combined use of the %FRC parameter with the default algorithm of the Sysmex XN-3000 automated hematology analyzer can improve the low specificity of the default algorithm in rapid screening for schistocytes.

Detection of dysmorphic cells in peripheral blood (PB) smears is essential in diagnostic screening of hematological diseases. Myelodysplastic syndromes (MDS) are hematopoietic neoplasms characterized by dysplastic and ineffective hematopoiesis, which diagnosis is mainly based on morphological findings of PB and bone marrow. We developed an automated diagnostic support system of MDS by combining an automated blood cell image-recognition system using a deep learning system (DLS) powered by convolutional neural networks (CNNs) with a decision-making system using extreme gradient boosting (XGBoost). The DLS of blood cell image-recognition has been trained using datasets consisting of 695,030 blood cell images taken from 3,261 PB smears including hematopoietic malignancies. The DLS simultaneously classified 17 blood cell types and 97 morphological features of such cells with >93.5% sensitivity and >96.0% specificity. The automated MDS diagnostic system successfully differentiated MDS from aplastic anemia (AA) with high accuracy; 96.2% of sensitivity and 100% of specificity (AUC 0.990). This is the first CNN-based automated initial diagnostic system for MDS using PB smears, which is applicable to develop new automated diagnostic systems for various hematological disorders.

Morphological microscopic examinations of nucleated cells in body fluid (BF) samples are performed to screen malignancy. However, the morphological differentiation is time-consuming and labor-intensive. This study aimed to develop a new flowcytometry-based gating analysis mode "XN-BF gating algorithm" to detect malignant cells using an automated hematology analyzer, Sysmex XN-1000. XN-BF mode was equipped with WDF white blood cell (WBC) differential channel. We added two algorithms to the WDF channel: Rule 1 detects larger and clumped cell signals compared to the leukocytes, targeting the clustered malignant cells; Rule 2 detects middle sized mononuclear cells containing less granules than neutrophils with similar fluorescence signal to monocytes, targeting hematological malignant cells and solid tumor cells. BF samples that meet, at least, one rule were detected as malignant. To evaluate this novel gating algorithm, 92 various BF samples were collected. Manual microscopic differentiation with the May-Grunwald Giemsa stain and WBC count with hemocytometer were also performed. The performance of these three methods were evaluated by comparing with the cytological diagnosis. The XN-BF gating algorithm achieved sensitivity of 63.0% and specificity of 87.8% with 68.0% for positive predictive value and 85.1% for negative predictive value in detecting malignant-cell positive samples. Manual microscopic WBC differentiation and WBC count demonstrated 70.4% and 66.7% of sensitivities, and 96.9% and 92.3% of specificities, respectively. The XN-BF gating algorithm can be a feasible tool in hematology laboratories for prompt screening of malignant cells in various BF samples.

Detection of dysmorphic cells in peripheral blood (PB) smears is essential for diagnosis of hematological malignancies. Myelodysplastic syndromes (MDS) are heterogeneous hematopoietic stem cell disorders that can lead to acute leukemia. Although examinations of bone marrow aspiration and biopsy as well as chromosomal and genetic tests are essential to diagnose these disorders, conventional tests such as complete blood count (CBC) and peripheral blood (PB) smear examinations remain to be initial diagnostic work ups. Detection of dysplastic cells in PB smears and evaluation of CBCs are particularly useful for rapid screening. Thanks to the recent advancement of computational and laboratory technologies, routine manual microscopic examinations have been replaced by automated hematology analyzers in many hematology laboratories. However, detection of dysmorphic cells is still challenging. Therefore, more sophisticated image recognition systems need to be developed. In this study, we developed a novel MDS diagnostic system using PB smears. The system consists of a convolutional neural network (CNN)-based image recognition deep learning system (DLS) and an EGB-based decision-making algorithm (XGBoost). All PB smears were prepared at Juntendo University Hospital. The slides were stained with May Grunwald-Giemsa using a fully automated slide-maker. A total of 703,970 digitalized cell images were collected. First, we trained the CNN-based image-recognition system using 695,030 blood cell images taken from 3,261 PB smears of which 1,165 were obtained from patients with hematological disorders. The hematological disorders included MDS (n=94 cases), myeloproliferative neoplasms (n=127), acute myeloid leukemia (n= 38), acute lymphoblastic leukemia (ALL, n=27), malignant lymphoma (n=324), multiple myeloma (n=82) and AA (n=42). Of all images, these 695,030 images were used to train the CNN-based image-recognition system (Fig 1), and rest of the images (n=8,940) were used for validation. The datasets were classified into 17 cell types and 97 abnormal morphological features by two board-certified laboratory technologists and one senior hematopathologist using the morphological criteria of the Clinical and Laboratory Standards Institute (CLSI) H20-A2 guideline and the 2017 WHO classification of myeloid neoplasms and acute leukemia. After accumulating the image patterns using the training datasets, the performance of the DLS was evaluated using the validation datasets that were prepared for testing the DLS. The internal features learnt by the DLS using t-distributed Stochastic Neighbor Embedding (t-SNE)(Fig 2). In this context, our CNN-based image-recognition system exhibited a sensitivity of >93.5% and a specificity of >96.0% when classifying cells in subsets of the validation datasets. We then created an automated MDS diagnostic system by combining the CNN-based image-recognition system with a form of XGBoost. To establish diagnostic algorithm, the training datasets obtained from 75 MDS and 36 Aplastic Anemia (AA) cases were used for learning of the cell image pattern for each disease. The diagnosis of all datasets was validated by independent hematopathologists using clinical information, laboratory, flow cytometric, and genetic data, and bone marrow aspiration and biopsy findings. The accuracy of the system was tested using validation datasets (26 MDS and 11 AA cases). The system differentiated MDS from AA with the sensitivity and specificity of 96.2% and 100%, respectively (AUC 0.990). In conclusion, this is the first CNN-based automated initial diagnostic system for MDS using PB smears, which is applicable to develop new automated image diagnostic systems for various hematological disorders. Currently, we are collecting more data to improve the accuracy. Disclosures Kimura: Sysmex Corporation: Employment. Takehara:Sysmex Corporation: Employment. Uchihashi:Sysmex Corporation: Employment.

Background Storing K(x)EDTA-conjugated blood samples at room temperature or under insufficient cooling conditions results in various morphological changes such as swelling of the blood cells. These changes are reproducible and have already been described well. However, they can lead to incorrect flagging when using automated hematology analyzers for complete blood counts and white blood cell differentials. The aim of this study was to determine if those changes can be detected automatically and used to prevent false positive flagging. Methods 150 blood samples were aged under controlled conditions and the impact on the "Aged sample" software was checked retrospectively. The results were verified in a second retrospective study including 6288 routine samples. Results When tested in a routine laboratory, the "Aged sample" software was able to reduce overall flagging by 23% without increasing false negative flagging. Conclusions The "Aged sample" software of XN-Series analyzers does not only detect and flag samples that are aging or were stored under suboptimal conditions but also prevents false positive flagging.

Introduction: Nucleated cells differential analysis of body fluid (BF) samples is important diagnostic tool for several diseases including cancer metastasis. Detection of tumor cells in BF requires the manual morphological scanning of slides by the cytopathologists, which is time-consuming, labor-intensive and not always reliable because of a relatively low overall sensitivity rates (ranging 40-90%) with the higher false-negative rates for lymphomas and mesotheliomas. This study aimed to develop the scattergram gating analysis for detection of tumor cells in BF using the automated hematology analyzer Sysmex XN-1000 (Sysmex, Kobe, Japan). Methods: We used a total of 220 BF samples (53 cerebrospinal fluids, 73 pleural or ascitic fluids, and 94 peritoneal dialysis effluent) obtained from patients with cytological diagnoses (papanicolaou stain) including negative and positive of tumor cells, and chronic inflammation with an elevated lymphocyte and histiocyte fractions. As a reference method, morphological manual differential (200 cells counts) was performed by two experienced technologists using cytospin slides stained with the May-Grunwald Giemsa. The BF mode of XN-1000 (XN-BF) determines the differential cell counts of BF samples using side scatter and fluorescence intensity in WDF channel after the nuclear DNA/RNA stain by nucleic acid dye. The polymorphonuclear cells, mononuclear cells and high fluorescence cells (HF-BF), corresponding with a high amount of nucleic acids, are differentiated. Mesothelial cells and macrophages are theoretically categorized as HF-BF cells and included in the total nucleated cell count but not in the WBC count. We selected the tumor cells positive (n=18) and chronic inflammation positive samples (n=108) by morphological manual differential, and reviewed their scattergram patterns determined by XN-BF. Then the novel scattergram gating strategy targeting the tumor cells was evaluated. The gating criteria were based on the WDF scatter plots; #1: detect the cells with larger size and higher fluorescence signal in comparison with general leukocytes, which mainly derived from clustered tumor cells, #2: detect the middle sized mononuclear cells with less granules rather than neutrophils and similar fluorescence signal to monocytes, which targeting hematological malignant cells and solid tumor cells. BF samples that meet at least one criterion were interpreted as positive for tumor cells. Results: The malignant BF samples containing tumor cells showed the different scattergram patterns from the benign ones with chronic inflammation; the malignant BF formed the isolated cellular clusters in our gating system, and the inflammatory BF showed the continuous expansion into the HF-BF area. Our scattergram gating analysis achieved an overall sensitivity of 72.2% and specificity of 98.0% in detecting tumor cells positive samples when screening against all samples outcomes. The positive predictive value was 76.5% and the negative predictive value was 97.5%. For the samples with absence of tumor cells and inflammatory observations (n=94), no false positive was detected. Of notes, each of our gating criterion detected the different type of tumor cells. For example, the scattergram gating analysis #1 achieved an overall sensitivity of 72.7% and specificity of 99.0% in detecting adenocarcinoma with the positive predictive value of 80.0% and the negative predictive value of 98.6%. Conclusions: A simple measurement of BF by automated hematology analyzer in which cells are minimally handled has a potential to reduce costs and allow routine cell screening in clinical applications. For the BF malignancy diagnostics, the scattergram gating analysis is promising to (i) augment diagnostic routines without requiring additional sample preparation procedure, (ii) limit operator bias, and (iii) provide a standardized measurement. Disclosures No relevant conflicts of interest to declare.