FlowLOT: Linearized Optimal Transport for Flow Cytometry Analysis
Abstract
Multiparameter flow cytometry generates high-dimensional, unordered single-cell mea- surement data for disease diagnosis and monitoring, yet analysis often remains dependent on manual gating, limiting scalability and reproducibility. Existing machine-learning ap- proaches can reduce annotation burden but frequently require large training cohorts and offer limited interpretability. To address these challenges, we introduce FlowLOT , an optimal-transport-based framework that models the single-cell measurement data of a pa- tient sample as an empirical cellular distribution and maps it directly into a fixed-length feature vector. Within a single transparent architecture, FlowLOT unifies high-dimensional classification, interpretable visualization, and continuous quantitative inference. In few-shot regimes, using as few as 16 patients per class on FlowCAP-II and 8 patients per class on BLAST110, it accurately distinguishes healthy from acute myeloid leukemia (AML) sam- ples, reaching 94.3% and 98.0% balanced accuracy, respectively. The underlying embedding exposes marker-level variation driving disease-associated population shifts and enables quantitative measurable residual disease (MRD) estimation, achieving a Pearson correlation of 0.82 on held-out samples and 0.79 under cross-dataset transfer. Furthermore, at the clinically relevant 0.1% threshold for leukemia-associated immunophenotype (LAIP) residual disease, FlowLOT detects positivity with 72% sensitivity at 100% specificity. By replacing subjective manual gating and black-box deep learning with a distribution-aware framework, FlowLOT offers a sample-efficient, scalable, and interpretable solution for high- dimensional cytometry under realistic clinical and experimental constraints.
Related
- Evaluation of Methods for Identifying Human Monocytes by Flow Cytometry: TLR2 May be an Optimal Marker
- TempoBridge: Source-Conditioned Flow Matching with Optimal Transport Couplings for Single-Cell Population Transitions
- Super-resolution imaging flow cytometry
- Whole Blood Deep Immunophenotyping and Platelet-Leukocyte Aggregate Quantification by Spectral Flow Cytometry
- A 21-Color Spectral Flow Cytometry Panel for Comprehensive Immunophenotyping of Rat Tissues and a Companion 9-Color T-Cell Functional Assay
BibTeX
@misc{pathan2026flowlot,
title = {{FlowLOT: Linearized Optimal Transport for Flow Cytometry Analysis}},
author = {Naqib Sad Pathan and Mohammad Shifat-E-Rabbi and Kristofor E. Pas and Ivan Medri and Bartek Rajwa and Gustavo K. Rohde},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.17906},
eprint = {2609.17906},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.17906},
}
RIS
TY - UNPB
TI - FlowLOT: Linearized Optimal Transport for Flow Cytometry Analysis
AU - Naqib Sad Pathan
AU - Mohammad Shifat-E-Rabbi
AU - Kristofor E. Pas
AU - Ivan Medri
AU - Bartek Rajwa
AU - Gustavo K. Rohde
PY - 2026
DA - 2026/09/15
PB - arXiv
DO - 10.48550/arxiv.2609.17906
UR - https://arxiv.org/abs/2609.17906
AB - Multiparameter flow cytometry generates high-dimensional, unordered single-cell mea- surement data for disease diagnosis and monitoring, yet analysis often remains dependent on manual gating, limiting scalability and reproducibility. Existing machine-learning ap- proaches can reduce annotation burden but frequently require large training cohorts and offer limited interpretability. To address these challenges, we introduce FlowLOT , an optimal-transport-based framework that models the single-cell measurement data of a pa- tient sample as an empirical cellular distribution and maps it directly into a fixed-length feature vector. Within a single transparent architecture, FlowLOT unifies high-dimensional classification, interpretable visualization, and continuous quantitative inference. In few-shot regimes, using as few as 16 patients per class on FlowCAP-II and 8 patients per class on BLAST110, it accurately distinguishes healthy from acute myeloid leukemia (AML) sam- ples, reaching 94.3% and 98.0% balanced accuracy, respectively. The underlying embedding exposes marker-level variation driving disease-associated population shifts and enables quantitative measurable residual disease (MRD) estimation, achieving a Pearson correlation of 0.82 on held-out samples and 0.79 under cross-dataset transfer. Furthermore, at the clinically relevant 0.1% threshold for leukemia-associated immunophenotype (LAIP) residual disease, FlowLOT detects positivity with 72% sensitivity at 100% specificity. By replacing subjective manual gating and black-box deep learning with a distribution-aware framework, FlowLOT offers a sample-efficient, scalable, and interpretable solution for high- dimensional cytometry under realistic clinical and experimental constraints.
ER -