Deep generative modeling of annotated bacterial biofilm images
Abstract Biofilms are critical for understanding environmental processes, developing biotechnology applications, and progressing in medical treatments of various infections. Nowadays, a key limiting factor for biofilm analysis is the difficulty in obtaining large datasets with fully annotated images...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | npj Biofilms and Microbiomes |
| Online Access: | https://doi.org/10.1038/s41522-025-00647-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832595010475786240 |
|---|---|
| author | Angelina A. Holicheva Konstantin S. Kozlov Daniil A. Boiko Maxim S. Kamanin Daria V. Provotorova Nikita I. Kolomoets Valentine P. Ananikov |
| author_facet | Angelina A. Holicheva Konstantin S. Kozlov Daniil A. Boiko Maxim S. Kamanin Daria V. Provotorova Nikita I. Kolomoets Valentine P. Ananikov |
| author_sort | Angelina A. Holicheva |
| collection | DOAJ |
| description | Abstract Biofilms are critical for understanding environmental processes, developing biotechnology applications, and progressing in medical treatments of various infections. Nowadays, a key limiting factor for biofilm analysis is the difficulty in obtaining large datasets with fully annotated images. This study introduces a versatile approach for creating synthetic datasets of annotated biofilm images with employing deep generative modeling techniques, including VAEs, GANs, diffusion models, and CycleGAN. Synthetic datasets can significantly improve the training of computer vision models for automated biofilm analysis, as demonstrated with the application of Mask R-CNN detection model. The approach represents a key advance in the field of biofilm research, offering a scalable solution for generating high-quality training data and working with different strains of microorganisms at different stages of formation. Terabyte-scale datasets can be easily generated on personal computers. A web application is provided for the on-demand generation of biofilm images. |
| format | Article |
| id | doaj-art-e70b0f4ee9a74bffa39a3356dd6eccbe |
| institution | Kabale University |
| issn | 2055-5008 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Biofilms and Microbiomes |
| spelling | doaj-art-e70b0f4ee9a74bffa39a3356dd6eccbe2025-01-19T12:12:19ZengNature Portfolionpj Biofilms and Microbiomes2055-50082025-01-0111111510.1038/s41522-025-00647-4Deep generative modeling of annotated bacterial biofilm imagesAngelina A. Holicheva0Konstantin S. Kozlov1Daniil A. Boiko2Maxim S. Kamanin3Daria V. Provotorova4Nikita I. Kolomoets5Valentine P. Ananikov6Tula State University, Lenin pr. 92Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospekt 47Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospekt 47Tula State University, Lenin pr. 92Tula State University, Lenin pr. 92Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospekt 47Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospekt 47Abstract Biofilms are critical for understanding environmental processes, developing biotechnology applications, and progressing in medical treatments of various infections. Nowadays, a key limiting factor for biofilm analysis is the difficulty in obtaining large datasets with fully annotated images. This study introduces a versatile approach for creating synthetic datasets of annotated biofilm images with employing deep generative modeling techniques, including VAEs, GANs, diffusion models, and CycleGAN. Synthetic datasets can significantly improve the training of computer vision models for automated biofilm analysis, as demonstrated with the application of Mask R-CNN detection model. The approach represents a key advance in the field of biofilm research, offering a scalable solution for generating high-quality training data and working with different strains of microorganisms at different stages of formation. Terabyte-scale datasets can be easily generated on personal computers. A web application is provided for the on-demand generation of biofilm images.https://doi.org/10.1038/s41522-025-00647-4 |
| spellingShingle | Angelina A. Holicheva Konstantin S. Kozlov Daniil A. Boiko Maxim S. Kamanin Daria V. Provotorova Nikita I. Kolomoets Valentine P. Ananikov Deep generative modeling of annotated bacterial biofilm images npj Biofilms and Microbiomes |
| title | Deep generative modeling of annotated bacterial biofilm images |
| title_full | Deep generative modeling of annotated bacterial biofilm images |
| title_fullStr | Deep generative modeling of annotated bacterial biofilm images |
| title_full_unstemmed | Deep generative modeling of annotated bacterial biofilm images |
| title_short | Deep generative modeling of annotated bacterial biofilm images |
| title_sort | deep generative modeling of annotated bacterial biofilm images |
| url | https://doi.org/10.1038/s41522-025-00647-4 |
| work_keys_str_mv | AT angelinaaholicheva deepgenerativemodelingofannotatedbacterialbiofilmimages AT konstantinskozlov deepgenerativemodelingofannotatedbacterialbiofilmimages AT daniilaboiko deepgenerativemodelingofannotatedbacterialbiofilmimages AT maximskamanin deepgenerativemodelingofannotatedbacterialbiofilmimages AT dariavprovotorova deepgenerativemodelingofannotatedbacterialbiofilmimages AT nikitaikolomoets deepgenerativemodelingofannotatedbacterialbiofilmimages AT valentinepananikov deepgenerativemodelingofannotatedbacterialbiofilmimages |