QuantEM: An optimized platform of vision transformer-based models for segmentation and analysis of electron microscopy data

Verdict. Here we present QuantEM, an open-source platform for segmentation and analysis of EM data across imaging modalities, tissues, and species. It intersects mitochondrial stress/dysfunction themes (aging).

What the authors report

Electron microscopy (EM) is essential for resolving cellular ultrastructure, yet quantitative analysis remains limited by labor-intensive segmentation and the scarcity of generalizable models. Across diverse naive datasets, QuantEM consistently matches or exceeds existing models on zero-shot segmentation while requiring less data for fine-tuning.

Key results stated in the abstract include the following. Here we present QuantEM, an open-source platform for segmentation and analysis of EM data across imaging modalities, tissues, and species. We assembled the largest curated collection of intracellular EM datasets to date, comprising over 15,000 two-dimensional images and 1,700 three-dimensional acquisitions from more than 600 datasets, including nearly 4,000 newly released acquisitions. Using this resource, we trained an EM-specific vision transformer foundation model and systematically optimized adaptation strategies for organelle segmentation.

Why it matters for mitochondrial biology

Within mitochondrial research, this work maps primarily to redox biology, aging, structural biology, computational. It is relevant to mitochondrial dysfunction discourse because the abstract invokes aging. That does not by itself establish a validated disease mechanism; it indicates thematic proximity. Server: biorxiv. Posted 2026-08-07. Synthesis confidence is bounded by abstract completeness.

Study design (abstract-level)

Electron microscopy (EM) is essential for resolving cellular ultrastructure, yet quantitative analysis remains limited by labor-intensive segmentation and the scarcity of generalizable models. Here we present QuantEM, an open-source platform for segmentation and analysis of EM data across imaging modalities, tissues, and species. Using this resource, we trained an EM-specific vision transformer foundation model and systematically optimized adaptation strategies for organelle segmentation.

Principal findings

1. Here we present QuantEM, an open-source platform for segmentation and analysis of EM data across imaging modalities, tissues, and species.
2. We assembled the largest curated collection of intracellular EM datasets to date, comprising over 15,000 two-dimensional images and 1,700 three-dimensional acquisitions from more than 600 datasets, including nearly 4,000 newly released acquisitions.
3. Using this resource, we trained an EM-specific vision transformer foundation model and systematically optimized adaptation strategies for organelle segmentation.
4. QuantEM provides pretrained models for mitochondria, endoplasmic reticulum, nuclei, and lipid droplets, integrated with interactive proofreading and downstream quantitative analyses through standalone and napari interfaces.
5. We further demonstrate its utility by revealing previously unrecognized subcellular compartmentation of hepatic glucokinase using immuno-electron microscopy.

Limitations of this brief

• This Mitos brief is an abstract-level synthesis of a preprint; it is not peer review and not a substitute for reading the full paper.
• Preprint status: findings may change with revision or journal review.
• Effect sizes, n numbers, statistics, and full experimental controls are typically incomplete at abstract resolution.
• Comparator/control language is weak or absent in the abstract, limiting causal inference from this brief alone.
• Primary source: biorxiv DOI 10.64898/2026.08.06.743293 (posted 2026-08-07).

Open scientific questions

• Which specific experimental panels in the full paper establish the strongest causal claim, and how robust are the controls?
• How do these findings sit relative to prior literature on the same pathway—replication, contradiction, or incremental extension?

Bottom line

For mitochondrial biologists focused on redox biology, aging, structural biology, this preprint is worth full-text review if the topic matches your program. Abstract-level takeaway: Here we present QuantEM, an open-source platform for segmentation and analysis of EM data across imaging modalities, tissues, and species. Confirm methods, effect sizes, and controls in the full PDF before citing the result as established.

Bibliographic record

Abstract-based editorial synthesis by Mitos. Not peer review.