BiaPy, an accessible AI tool for analysing biomedical images

Used to study cellular structures, tissue, and organs across a range of disciplines, image analysis
is an essential tool in biomedicine. However, applying AI to analyse these images has
traditionally been the preserve of experts in programming and data science. BiaPy breaks down
that barrier by offering an easy-to-use platform that allows advanced AI models to be applied
without the need for specialised technical knowledge.
"BiaPy aims to democratise access to artificial intelligence in bioimaging by enabling more
scientists and healthcare professionals to harness its potential without the need for advanced
programming or machine learning skills,” explained Daniel Franco, lead author of the study and
currently a postdoctoral researcher at the MRC Laboratory of Molecular Biology and Cambridge
University (United Kingdom).
BiaPy allows different types of analysis to be performed on scientific images, such as
automatically identifying cells or other biological structures, counting elements, classifying
samples according to their appearance, or improving image quality to see the finer details. All
this can be done with two-dimensional images as well as with three-dimensional images
obtained by means of various microscopy techniques. What is more, BiaPy has been designed
to be efficient and scalable: it can work with a broad variety of data volumes, from a few small
images to terabytes of information, such as those generated when tissue or entire organs are
scanned.
The tool is based on the use of “AI models”, which are algorithms trained to recognise patterns
in images, similar to the way the human eye can identify shapes or colours. Examples are used
to create a model: for example, images in which cells have already been tagged manually. With
sufficient training, the model learns to perform these tasks automatically, even on new images
it has never seen before.
“BiaPy has also been integrated into the BioImage Model Zoo (bioimage.io), a database in which
researchers from around the world share pre-trained models. Thanks to this integration, BiaPy
users can reuse existing models for new images or train their own models easily,” explained
Arrate Muñoz, senior co-author of the paper and member of the European consortium AI4Life
that developed the BioImage Model Zoo.
This tool is already being used in advanced scientific projects. One example is CartoCell, a
software solution developed in collaboration with the lab coordinated by Luis M. Escudero
(Institute of Biomedicine of Seville [Virgen del Rocío University Hospital/CSIC/University of
Seville]). CartoCell analyses microscopy images to reveal hidden patterns in the shape and
distribution of cells within 3D epithelial tissue from different organisms.
Another case worthy of note is its application in collaboration with the laboratories of Emmanuel
Beaurepaire (École Polytechnique, France) and Jean Livet (Institut de la Vision, Paris). These
groups have developed the ChroMS microscopy technique, which allows huge threedimensional
images of entire brains to be obtained using fluorescent colours generated by
proteins from jellyfish and corals. BiaPy is used to automatically detect each cell in these largescale
images, even in densely populated areas of the brain, allowing brain development to be
studied by reconstructing the lineage of cells based on their colours and three-dimensional
positions.
As an open-access tool, BiaPy is available free of charge to the scientific community, thereby
promoting collaboration and the ongoing improvement of the software. It can be used on PCs
or servers with multiple graphics cards, as well as in the cloud. It is easy to install and ensures
that experiments can be easily repeated in various environments, thus promoting open,
reproducible science.
As Ignacio Arganda, the senior author of the paper, pointed out, “the development of BiaPy
represents an important step towards the democratisation of advanced artificial computer
vision in microscopy. Its accessible design and focus on open collaboration reduce technical
barriers, making it easier for more researchers and healthcare professionals to apply artificial
vision to their studies. Its compatibility with various computing environments and its open-code
nature mean that it is a platform that offers huge potential in driving forward innovation and
speeding up scientific discovery.”
For further information about BiaPy, check out:
http://biapyx.github.io/
https://www.biorxiv.org/content/10.1101/2024.02.03.576026v3.abstract

Publication reference
Daniel Franco-Barranco, Jesús A. Andrés-San Román, Ivan Hidalgo-Cenalmor, Lenka
Backová, Aitor González-Marfil, Clément Caporal, Anatole Chessel, Pedro Gómez-
Gálvez, Luis M. Escudero, Donglai Wei, Arrate Muñoz-Barrutia & Ignacio Arganda-
Carreras
BiaPy: accessible deep learning on bioimages
Nature Methods Volume 22, No. 4, 2025.
DOI https://doi.org/10.1038/s41592-025-02699-y