Bright 4B: Scaling Hyperspherical Learning for Segmentation in 3D Brightfield Microscopy

TL;DR

Bright-4B is a large hyperspherical model that enables accurate, label-free 3D segmentation of cellular structures in brightfield microscopy, outperforming existing methods without requiring fluorescence or extensive post-processing.

Contribution

We introduce Bright-4B, a novel hyperspherical foundation model with advanced attention and stabilization mechanisms for direct 3D segmentation from brightfield images.

Findings

Outperforms CNN and Transformer baselines in preserving fine structural detail

Achieves accurate segmentation of nuclei, mitochondria, and organelles without fluorescence

Handles diverse cell types and depths effectively

Abstract

Label-free 3D brightfield microscopy offers a fast and noninvasive way to visualize cellular morphology, yet robust volumetric segmentation still typically depends on fluorescence or heavy post-processing. We address this gap by introducing Bright-4B, a 4 billion parameter foundation model that learns on the unit hypersphere to segment subcellular structures directly from 3D brightfield volumes. Bright-4B combines a hardware-aligned Native Sparse Attention mechanism (capturing local, coarse, and selected global context), depth-width residual HyperConnections that stabilize representation flow, and a soft Mixture-of-Experts for adaptive capacity. A plug-and-play anisotropic patch embed further respects confocal point-spread and axial thinning, enabling geometry-faithful 3D tokenization. The resulting model produces morphology-accurate segmentations of nuclei, mitochondria, and other organelles from brightfield stacks alone--without fluorescence, auxiliary channels, or handcrafted post-processing. Across multiple confocal datasets, Bright-4B preserves fine structural detail across depth and cell types, outperforming contemporary CNN and Transformer baselines. All code, pretrained weights, and models for downstream finetuning will be released to advance large-scale, label-free 3D cell mapping.