Learning 3D biophysical cell properties from 2D images and cell-population statistics

Santiago Hernández-Orozco; Hector Zenil

cs.AISep 18, 2026

Learning 3D biophysical cell properties from 2D images and cell-population statistics

Authors: Santiago Hernández-Orozco, Hector Zenil

Organizations: Algocyte | Oxford Immune Algorithmics, Oxford University Innovation & London Institute for Healthcare Engineering, UK · Algorithmic Dynamics Lab, Department of Biomedical Computing, School of Biomedical Engineering and Imaging Sciences, King’s Institute for AI, King’s College London, UK

Abstract

Inferring 3D cellular properties from 2D microscopy is difficult when a reference instrument reports only population statistics rather than labels for individual cells. Here we develop a population-supervised framework that maps single 2D red-cell images to latent biophysical quantities and aggregates them to mean corpuscular volume, red-cell distribution width and mean corpuscular haemoglobin. The model combines shared local inference, a biophysically structured decoder for volume and haemoglobin, learned instance weighting and device-specific calibration. We formalise conditions under which aggregate observations identify restricted instance predictors, show why population agreement does not by itself identify single-cell properties or 3D geometry, and derive the dispersion penalty induced by subset mean matching. The development dataset comprises 390 specimens and 1,105 acquisitions across six devices, with reported Pearson correlations of 0.86--0.98 against a Sysmex analyser. The framework provides a testable route from 2D images and population supervision to 3D cellular biophysics without claiming explicit 3D reconstruction.

Figures & tables

Figure 1: Representative $100\times 100$ masked RGB input crops illustrating variation in apparent morphology, focus and segmentation. Images are reproduced without intensity adjustment and enlarged by nearest-neighbour replication.

Index	Reported Pearson correlation	Reference
MCV	0.94–0.96	Sysmex analyser
RDW	0.86–0.92	Sysmex analyser
MCH	0.94–0.98	Sysmex analyser

Table 1: Reported Pearson correlations against a Sysmex analyser in the development data [ 11 ] .

Stage	Operations	Output shape
Input	Masked RGB crop	$B\times 3\times 100\times 100$
Block 1	Two $3\times 3$ convolutions, BN, ReLU; pool 2	$B\times 32\times 50\times 50$
Block 2	Two $3\times 3$ convolutions, BN, ReLU; pool 2	$B\times 64\times 25\times 25$
Block 3	Two $3\times 3$ convolutions, BN, ReLU; pool 5	$B\times 128\times 5\times 5$
Block 4	$3\times 3$ convolution, BN, ReLU; adaptive pool	$B\times 256\times 1\times 1$
Embedding	Flatten; linear $256\to D$ ; ReLU	$B\times D$

Table 2: Convolutional encoder and scalar-head architecture.

B

denotes the crop minibatch size and

D

the embedding dimension.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Figure D.1: Synthetic subsampling diagnostic. All size-two subsets of

(70,80,100,110)

fL are enumerated after scaling deviations about

90

fL by

a

. Expected mean-target loss equals

(250/3)a^{2}

. The vertical reference marks the generating dispersion.

Microscopy Image Analysis

Microscopy image analysis uses computational methods to detect, segment, track, reconstruct, and quantify structures in scientific images.

See all 198 papers

Sampling-Aware 3D Spatial Analysis in Multiplexed Imaging
Apr 9, 2026Ido Harlev, Tamar Oukhanov, Raz Ben-Uri +2Microscopy Image Analysis 3D Reconstruction
Weizmann Institute of Science, Israel
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
3D Masked Autoencoders are Robust Learners of Volumetric and Multimodal Cellular Representations for Microscopy
Jun 22, 2026Amirhossein Kardoost, Lion Gleiter, Tingying Peng +1Cross-Modal Representation Learning Masked Autoencoders
Institute of AI for Health & Helmholtz AI, Computational Health Center, Helmholtz Munich – German Research Center for Environmental Health, Neuherberg, Germany · Munich Center for Machine Learning (MCML), Munich, Germany · Department of Medicine III, Ludwig-Maximilian-University Hospital, Munich, Germany +2

Apr 9, 2026J/K move · Enter open · S save

Sampling-Aware 3D Spatial Analysis in Multiplexed Imaging

Ido Harlev, Tamar Oukhanov, Raz Ben-Uri, Leeat Keren+1

Weizmann Institute of Science, Israel

Microscopy Image Analysis 3D Reconstruction Sparse-View 3D Reconstruction

Open paper AskPDF arXiv

Highly multiplexed microscopy enables rich spatial characterization of tissues at single-cell resolution, yet most analyses rely on two-dimensional sections despite inherently three-dimensional tissue organization. Acquiring dense volumetric data in spatial proteomics remains costly and technically challenging, leaving practitioners to choose between 2D sections or 3D serial sections under limited imaging budgets. In this work, we study how sampling geometry impacts the stability of commonly used spatial statistics, and we introduce a geometry-aware reconstruction module that enables sparse yet consistent 3D analysis from serial sections. Using controlled simulations, we show that planar sampling reliably recovers global cell-type abundance but exhibits high variance for local statistics such as cell clustering and cell-cell interactions, particularly for rare or spatially localized populations. We observe consistent behavior in real multiplexed datasets, where interaction metrics and neighborhood relationships fluctuate substantially across individual sections. To support sparse 3D analysis in practice, we present a reconstruction approach that links cell projections across adjacent sections using phenotype and proximity constraints and recovers single-cell 3D centroids using cell-type-specific shape priors. We further analyze the trade-off between section spacing, coverage, and redundancy, identifying acquisition regimes that maximize reconstruction utility under fixed imaging budgets. We validate the reconstruction module on a public imaging mass cytometry dataset with dense axial sampling and demonstrate its downstream utility on an in-house CODEX dataset by enabling structure-level 3D analyses that are unreliable in 2D. Together, our results provide diagnostic tools and practical guidance for deciding when 2D sampling suffices and when sparse 3D reconstruction is warranted.

Abstract

Microscopy Image Analysis

Explore similar work

Sampling-Aware 3D Spatial Analysis in Multiplexed Imaging

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

3D Masked Autoencoders are Robust Learners of Volumetric and Multimodal Cellular Representations for Microscopy