FlowCat

B-cell neoplasm is a hematological disorder involving the malignant growth of B lymphocytes. Multi-channel flow cytometry (MFC) is an integral tool in the diagnosis of neoplasms. MFC data analysis requires trained experts to manually gate cell populations of interest in two-dimensional plots, which is a time-consuming and subjective process.

FlowCat aims to automatically classify MFC data into lymphoma diagnosis labels without the need for any manual gating.

We developed an artificial intelligence (AI) model to automatically classify MFC data into lymphoma diagnosis labels. We transformed the MFC data into self-organizing maps (SOMs) that were then analyzed using a convolution neural network (CNN). We were able to achieve an expert-level accuracy (Weighted F1 score of 0.94) in classifying the MFC samples into eight classes: chronic lymphocytic leukemia and its predecessor monoclonal B-cell lymphocytosis (CLL/MBL), marginal zone lymphoma (MZL), mantle cell lymphoma (MCL), prolymphocytic leukemia (PL), follicular lymphoma (FL), hairy cell leukemia (HCL), and lymphoplasmacytic lymphoma (LPL) and healthy controls. This level of classification accuracy was possible by deep learning on a big data set from a single lab comprising more than 20,000 samples that were measured with the same protocol.

However, MFC is not standardized and the protocol with which a sample is acquired is subject to inter-laboratory variability and thus the MFC panel changes over time in terms of the number of tubes per sample, markers measured, marker-fluorochromes conjugates, as well as the cytometer used. This limits the AI model to a specific panel with which it was trained.

We extended our AI model (base model) to four additional MFC panels (target data sets) with much smaller data sets using transfer learning. We developed a workflow, Merge_TL, that combines transfer learning with FCS file merging to handle differences across MFC panels. We trained models for each of the four target data sets by transferring the features learned from our base model. Our workflow enabled models to be adapted to multiple MFC panels and improve the performance of these models making it possible to achieve expert-level accuracy that was otherwise not possible.

We provide an easy-to-use web service at https://hema.to. If you are interested in customizing FlowCat for your diagnostic workflow, get in touch with us!

Press release: https://www.uni-bonn.de/en/news/221-2021

Collaborators: res mechanica (https://resmechanica.com/)

Publications:

1.     Zhao, M., Mallesh, N., Höllein, A., Schabath, R., Haferlach, C., Haferlach, T., Elsner, F., Lüling, H., Krawitz, P. and Kern, W. (2020), Hematologist-Level Classification of Mature B-Cell Neoplasm Using Deep Learning on Multiparameter Flow Cytometry Data. Cytometry, 97: 1073-1080. 

2.      Mallesh, N., Zhao, M., Meintker, L., Höllein, A., Elsner, F., Lüling, H., Haferlach, T., Kern, W., Westermann, J., Brossart, P., et al. (2021). Knowledge transfer to enhance the performance of deep learning models for automated classification of B cell neoplasms. Patterns, 100351.