Deep aerenchyma: a transformer-based pipeline for scalable phenotyping of rice root aerenchyma lacunae across environments
Hani Atef, Laura Fierro-Dominguez, Paula Andrea Lozano-Montaña, Sergi Navarro Sanz, Julie Bals, Benoît Clerget, Christophe Périn, Maria Camila Rebolledo, Romain Fernandez · 2026
Read Deep aerenchyma: a transformer-based pipeline for scalable phenotyping of rice root aerenchyma lacunae across environments by Hani Atef, Laura Fierro-Dominguez, Paula Andrea Lozano-Montaña, Sergi Navarro Sanz, Julie Bals, Benoît Clerget, Christophe Périn, Maria Camila Rebolledo, Romain Fernandez (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13595614.txt.
Description
Hani Atef and co-authors present Deep Aerenchyma, an AI pipeline that measures the air channels inside rice roots automatically, replacing slow hand tracing of root slices. These lacunae move oxygen through flooded roots. They shape both methane emissions from paddies and the plant's water flow. That makes them a breeding target. Yet they have been hard to phenotype at scale, as the 2026 report notes.
A vision transformer model learned from 1,760 annotated images gathered across countries, growth stages, and farming systems. It segmented root tissues and air spaces with high overlap against human markings, and its air-to-cortex ratios on a held-out set matched manual measurement almost exactly. Expert reviewers rated the machine's output at least as consistent as human annotation. Released as open software with an online demonstrator, the pipeline revealed reproducible differences across varieties, water regimes, and environments in six test cases.
Catalog record
| CREATOR | Hani Atef, Laura Fierro-Dominguez, Paula Andrea Lozano-Montaña, Sergi Navarro Sanz, Julie Bals, Benoît Clerget, Christophe Périn, Maria Camila Rebolledo, Romain Fernandez |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67420 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |