CytoBERT: A Foundation Model for Cytometry Data

Syed Abdul Haseeb Qadri; Bjarne C. Hiller; Felix Blanke; Vanja Sophie Cangalovic; Kutalmış Coşkun; Amin Mirzaei; Tom Siegl; Sebastian Bader

cs.LGAug 14, 2026

CytoBERT: A Foundation Model for Cytometry Data

Authors: Syed Abdul Haseeb Qadri, Bjarne C. Hiller, Felix Blanke, Vanja Sophie Cangalovic, Kutalmış Coşkun, Amin Mirzaei, Tom Siegl, Sebastian Bader,

Organizations: University of Rostock, Rostock, Germany · Marburg University, Marburg, Germany · Hessian Center for Artificial Intelligence, Darmstadt, Germany

Abstract

Cytometry measures the complex characteristics of single cells (e.g., counts and protein expression of immune cells) and is widely used across immunological research and clinical settings. However, cytometry data is highly heterogeneous and unstandardized due to experimental protocols and the choice of measured features. While machine learning methods hold the potential to gain deeper insights into cell biology, these challenges make them difficult to apply and transfer across studies. Recent advances in foundation models can alleviate these issues, but corresponding approaches are still scarce in this field. To address this, we provide CytoBERT, a publicly available, open-source, open-weight foundation model for single-cell cytometry data with variable marker panels. CytoBERT is pretrained in a self-supervised manner on a large-scale cytometry corpus (15 human datasets with heterogeneous marker panels and more than 50 million cells) curated through marker standardization, enabling it to learn transferable inter-marker relationships within cells. Fine-tuning CytoBERT for sample-level classification demonstrates that transfer learning across heterogeneous cytometry datasets is feasible, providing a starting point for scalable, generalizable cytometry analysis. Code is available at GitHub.

Figures & tables

Dataset	Subjects	Samples	Disease/condition	In pretrain?
SDY788	20	126	Transplant	Yes
SDY997	54	259	Autoimmune	No
SDY1535	50	100	HIV diagnosis	Yes
SDY1708	74	146	COVID severity	No
SDY1733	49	99	Liver cancer stage	No
SDY2011	113	185	COVID diagnosis	Yes

Table 1: Downstream CyTOF datasets. “Subjects” and “Samples” denote unique patients and sample files.

Figure 1: Sample classification results under grouped subject splitting. ROC-AUC is shown on the y-axis. The Friedman test found no significant differences between models (

p=0.093

), although CytoBERT achieved the best average rank.

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 2: Sample-level classification performance across CyTOF datasets under ungrouped splitting (top) and grouped subject splitting (bottom). The y-axis reports ROC-AUC. Grouped subject splitting ensures that samples from the same subject do not appear in both training and test folds.

Figure 3: Critical difference diagrams comparing average model ranks across CyTOF datasets under ungrouped splitting (top) and grouped subject splitting (bottom). Lower rank indicates better average performance.

Dataset	Logistic	CytoSet	CytoBERT Scratch	CytoBERT
SDY1708	0.560	0.568	0.556	0.723
SDY1733	0.918	0.932	0.937	0.932
SDY997	0.593	0.575	0.631	0.721
SDY2015	0.730	0.705	0.692	0.753
SDY1535	0.692	0.712	0.555	0.770
SDY788	0.487	0.412	0.123	0.246

Appendix

Table 2: Mean ROC-AUC of different models across downstream datasets.

Panel of Figure 4 — Figure 4: Overview of the CytoBERT pipeline, consisting of dataset curation and preprocessing, self-supervised masked expression-bin pretraining, and supervised fine-tuning for sample-level classification.

Figure 5: Shared marker vocabulary across all CyTOF datasets used in the project, consisting of 220 unique biomarker in total. Black = marker measured/available

Representation Learning

Representation learning develops data transformations that expose useful structure for prediction, transfer, generation, and control.

See all 804 papers

Retrieval-Augmented Vision Foundation Models for Robust Leukemia Cell Classification across Multiple Microscopy Datasets
Aug 11, 2026Carlos Zamora, Hiram Zuniga, Ulises Orozco-Rosas +1Retrieval-Augmented Classification Vision Foundation Model Adaptation
CETYS Universidad, Ave. CETYS Universidad No. 4, Fracc. El Lago, C.P. 22210, Tijuana, Baja California, Mexico
CytoFormer: A Molecularly Supervised Cell Foundation Model for Histopathology Cell Classification
Aug 17, 2026Jialu Yao, Songhao Li, Alina Yu +1Histopathology Image Classification Pathology Foundation Models
Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA, USA · Department of Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, PA, USA · Germantown Friends School, Philadelphia, PA, USA +1
Towards Label-Free Single-Cell Phenotyping Using Multi-Task Learning
May 14, 2026Saqib Nazir, Ardhendu BeheraMulti-Task Learning White Blood Cell Classification
Department of Computer Science, Edge Hill University, UK

Aug 11, 2026J/K move · Enter open · S save

Retrieval-Augmented Vision Foundation Models for Robust Leukemia Cell Classification across Multiple Microscopy Datasets

Carlos Zamora, Hiram Zuniga, Ulises Orozco-Rosas, Kenia Picos

CETYS Universidad, Ave. CETYS Universidad No. 4, Fracc. El Lago, C.P. 22210, Tijuana, Baja California, Mexico

Retrieval-Augmented Classification Vision Foundation Model Adaptation Medical Image Classification Cross-Domain Transfer Learning Oncology Cancer Subtype Classification

Open paper AskPDF arXiv

Leukemia cell image classification is challenged by real-world domain shifts from acquisition, staining, illumination, and site protocols, causing single-dataset models to generalize poorly in real clinical scenarios. This work presents a robust framework for leukemia classification across multiple heterogeneous datasets using a two-stage pipeline with a pretrained vision foundation model. Stage 1 performs binary classification (leukemia vs. non-leukemia) and is trained using 122,167 single-cell images. Stage 2 is conditionally applied to Stage 1 positives to perform subtype classification into Acute Lymphoblastic Leukemia (ALL) and Acute Myeloid Leukemia (AML), trained using 69,400 single-cell images. Labels are harmonized across five heterogeneous datasets to enable cross-dataset training, and performance is evaluated on a held-out dataset protocol to assess domain-shift generalization. Within this pipeline, three encoders are benchmarked (DinoBloom, pretrained on single-cell images; BiomedCLIP, pretrained on biomedical data; and CLIP as a general-purpose model) under linear probing, Low-Rank Adaptation (LoRA), and a Retrieval-Augmented Classification (RAC) module that retrieves the top-k most similar cell images to provide cytomorphological grounding. The objective is to quantify how much domain-specific pretraining contributes to performance under domain shift, and whether cost-effective adaptation and retrieval can be a viable alternative to expensive domain-specialized pretraining. The held-out protocol additionally serves as a diagnostic tool, revealing when classification performance is attributable to dataset-specific artifacts rather than to cytomorphological features.

Abstract

Representation Learning

Explore similar work

Retrieval-Augmented Vision Foundation Models for Robust Leukemia Cell Classification across Multiple Microscopy Datasets

CytoFormer: A Molecularly Supervised Cell Foundation Model for Histopathology Cell Classification

Towards Label-Free Single-Cell Phenotyping Using Multi-Task Learning