Artificial Intelligence helps diagnose Leukemia

November 04, 2021

Prof. Dr. med. Dipl. Phys. Peter Krawitz (picture: University Clinics Bonn)


Software trained with more than 30.000 data sets from B-Cell Lymphoma patients

Already in 2020, Cluster of Excellence ImmunoSensation2 Member Prof. Dr. Peter Krawitz and his team showed, how artificial Intelligence can help in the diagnosis of lymphomas and leukemias. The machine learning method developed by the scientists has since been further developed. It is made freely accessible and may be utilized also by smaller laboratories. The respective study has now been published in "Patterns".


Leukemia diagnostics relies on the analysis of blood- or bone marrow samples by flow cytometry. Large amounts of data are generated, as various markers are needed in addition to parameters like cell-size and -shape. “With 20 markers, the physician would already have to compare about 150 two-dimensional images," says Prof. Dr. Peter Krawitz from the Institute for Genomic Statistics and Bioinformatics at Bonn University Hospital. "That's why it's usually too costly to sift through the entire data set thoroughly."


In order to train the artificial intelligence (AI), Prof. Krawitz and the bioinformaticians Nanditha Mallesh and Max Zhao analyzed more than 30,000 data sets from patients with B-cell lymphomas. "The AI takes full advantage of the data and increases the speed and objectivity of the diagnoses," says Nanditha Mallesh. Still, the result presented by the AI can only be considered as suggestion and has to be reviewed by the physician. "The point of using AI is not to replace physicians, but to make the best use of the information contained in the data in the best possible way." Prof. Krawitz states.


The big step that brings the method closer to a broad clinical application is the free availability of the AI. The knowledge transfer offered also enables small laboratories to benefit from the development. Only a short training period is needed for the AI to internalize the specifics of the new lab. Subsequently, the knowledge derived from many thousands of data sets is available. All raw data and the complete software are open source and thus freely accessible. "With, we want to enable the exchange of anonymized flow cytometry data between laboratories and thus create the conditions for even higher quality in diagnostics," says Dr. Hannes Lüling from res mechanica.


The team sees great potential in this technology. For the diagnosis of B-cell lymphomas, also genetic and cytomorphological data are collected. "If we can succeed in using AI for these methods as well, then we would have an even more powerful tool," says Prof. Krawitz.



The study was funded by the German Research Foundation.



Nanditha Mallesh, Max Zhao, Lisa Meintker, Alexander Höllein, Franz Elsner, Hannes Lüling, Torsten Haferlach, Wolfgang Kern, Jörg Westermann, Peter Brossart, Stefan W. Krause, Peter M. Krawitz: Knowledge transfer to enhance the performance of deep learning models for automated classification of B-cell neoplasms, Patterns, DOI: 10.1016/j.patter.2021.100351




Prof. Dr. med. Dipl. Phys. Peter Krawitz
Institute for Genomic Statistics and Bioinformatics
University Hospital Bonn Phone +49 228 287 14799
E-mail: pkrawitz(at)