Experiment Overview

Repository ID: FR-FCM-Z2JW Experiment name: Hematologist-level classification of mature B-cell neoplasm using deep learning on multiparameter flow cytometry data MIFlowCyt score: 31.50%
Primary researcher: Max Zhao PI/manager: Peter Krawitz Uploaded by: Max Zhao
Experiment dates: 2016-01-01 - 2018-12-31 Dataset uploaded: Apr 2020 Last updated: Oct 2020
Keywords: [multicolor flow cytometry] [Deep Learning] [neural networks] [Non-Hodgkin Lymphoma] Manuscripts:
Organizations: Münchner Leukämielabor (MLL) GmbH, München, (Germany)
Purpose: Classify B-NHL subtypes from blood and bone marrow samples
Conclusion: We developed an automated model to directly classify FCS data into multiclass diagnosis label without the need for human supervision or any manual gating. Our model achieved a weighted F1-score of 0.94 for an eight-class classification: CLL/MBL, PL, FL, HCL, LPL, MCL, MZL and healthy controls.
Comments: None
Funding: Not disclosed
Quality control: Navios cytometer was calibrated according to the manufacturer’s recommendations
Download FCS Files or login and see the dataset in your inbox for further annotation details.