A Single-cell Morphological Dataset of Leukocytes from AML Patients and Non-malignant Controls

<p>The Munich AML Morphology Dataset contains 18,365 expert-labeled single-cell images taken from peripheral blood smears of 100 patients diagnosed with Acute Myeloid Leukemia at Munich University Hospital between 2014 and 2017, as well as 100 patients without signs of hematological malignancy. Image acquisition was done using a M8 digital microscope / scanner (Precipoint GmbH, Freising, Germany) at 100-fold optical magnification and oil immersion. Pathological and non-pathological leukocytes were classified into a standard morphological classification scheme derived from clinical practice by trained experts. To quantify inter- and intra-rater variability of examiners, a subset of images was re-annotated up to two times. The dataset has been used by the authors to train a convolutional neural network for single-cell morphology classification.</p>

慕尼黑急性髓系白血病形态学数据集收录了慕尼黑大学医院在2014年至2017年间，对100名被诊断为急性髓系白血病的患者以及100名无血液系统恶性肿瘤迹象的患者的外周血涂片中的18,365张由专家标注的单细胞图像。图像采集采用M8数字显微镜/扫描仪（制造商：Precipoint GmbH，德国弗赖辛）在100倍光学放大和油浸条件下进行。病理学和非病理学白细胞被训练有素的专家根据临床实践提炼的标准形态学分类方案进行分类。为了量化评审员之间的以及评审员内部的差异，部分图像被重新标注，最多两次。该数据集已被作者用于训练用于单细胞形态学分类的卷积神经网络。