Live-cell imaging in the deep learning era

Pylvänäinen, Joanna W.; Gómez-de-Mariscal, Estibaliz; Henriques, Ricardo; Jacquemet, Guillaume

doi:10.1016/j.ceb.2023.102271

Cited by 20 publications

(10 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep learning is revolutionising microscopy through datadriven analysis and discovery (14). However, significant barriers persist in accessing these advanced techniques, including a lack of training data, computing resources, and expertise (4, 6, 14). Proprietary platforms create technological and cultural obstacles, while complex workflows im pede adoption by non-experts.…”

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…Deep learning enables the transformative analysis of large multidimensional microscopy datasets, but barriers remain in implementing these advanced techniques (3, 4). Many researchers lack access to sufficient annotated data, highperformance computing resources, and expertise to develop, train, and deploy complex deep-learning models.…”

Section: Introductionmentioning

confidence: 99%

“…Many researchers lack access to sufficient annotated data, highperformance computing resources, and expertise to develop, train, and deploy complex deep-learning models. In recent years, several approaches have been developed to democratise the usage of deep learning for microscopy (4). Multiple tools, such as BioImage.io, facilitate sharing and reusing broadly useful, previously trained deep learning models, distributing them as one-click image analysis solutions (1, 5).…”

Section: Introductionmentioning

confidence: 99%

“…Multiple tools, such as BioImage.io, facilitate sharing and reusing broadly useful, previously trained deep learning models, distributing them as one-click image analysis solutions (1, 5). Yet often, deep learning models need to be trained or finetuned on the end user dataset to perform well (1, 4, 6). We previously released ZeroCostDL4Mic (2), an online platform relying on Google Colab that helped democratise deep learning by providing a zero-code interface to train and evaluate models capable of performing various bioimage analysis tasks, such as segmentation, object detection, denoising, super-resolution microscopy, and image-to-image translation.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

DL4MicEverywhere: Deep learning for microscopy made flexible, shareable, and reproducible

Hidalgo-Cenalmor,

Pylvänäinen,

Ferreira

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Deep learning has revolutionised the analysis of extensive microscopy datasets, yet challenges persist in the widespread adoption of these techniques. Many lack access to training data, computing resources, and expertise to develop complex models. We introduce DL4MicEverywhere, advancing our previous ZeroCostDL4Mic platform, to make deep learning more accessible. DL4MicEverywhere uniquely allows flexible training and deployment across diverse computational environments by encapsulating methods in interactive Jupyter notebooks within Docker containers –a standalone virtualisation of required packages and code to reproduce a computational environment–. This enhances reproducibility and convenience. The platform includes twice as many techniques as originally provided by ZeroCostDL4Mic and enables community contributions via automated build pipelines. DL4MicEverywhere empowers participatory innovation and aims to democratise deep learning for bioimage analysis.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

DL4MicEverywhere: Deep learning for microscopy made flexible, shareable, and reproducible

Hidalgo-Cenalmor,

Pylvänäinen,

Ferreira

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Accurate Identification of Cancer Cells in Complex Pre‐Clinical Models Using a Deep‐Learning Neural Network: A Transfection‐Free Approach

Cortesi,

Liu,

Powell

et al. 2024

Advanced Biology

View full text Add to dashboard Cite

Abstract3D co‐cultures are key tools for in vitro biomedical research as they recapitulate more closely the in vivo environment while allowing a tighter control on the culture's composition and experimental conditions. The limited technologies available for the analysis of these models, however, hamper their widespread application. The separation of the contribution of the different cell types, in particular, is a fundamental challenge. In this work, ORACLE (OvaRiAn Cancer ceLl rEcognition) is presented, a deep neural network trained to distinguish between ovarian cancer and healthy cells based on the shape of their nucleus. The extensive validation that are conducted includes multiple cell lines and patient‐derived cultures to characterize the effect of all the major potential confounding factors. High accuracy and reliability are maintained throughout the analysis (F1score> 0.9 and Area under the ROC curve ‐ROC‐AUC‐ score = 0.99) demonstrating ORACLE's effectiveness with this detection and classification task. ORACLE is freely available (https://github.com/MarilisaCortesi/ORACLE/tree/main) and can be used to recognize both ovarian cancer cell lines and primary patient‐derived cells. This feature is unique to ORACLE and thus enables for the first time the analysis of in vitro co‐cultures comprised solely of patient‐derived cells.

show abstract

Fast and Long‐Term Super‐Resolution Imaging of Endoplasmic Reticulum Nano‐structural Dynamics in Living Cells Using a Neural Network

Rahm,

Balakrishnan,

Wehrheim

et al. 2024

Small Science

View full text Add to dashboard Cite

Stimulated emission depletion (STED) microscopy is a super‐resolution technique that surpasses the diffraction limit and has contributed to the study of dynamic processes in living cells. However, high laser intensities induce fluorophore photobleaching and sample phototoxicity, limiting the number of fluorescence images obtainable from a living cell. Herein, these challenges are addressed by using ultra‐low irradiation intensities and a neural network for image restoration, enabling extensive imaging of single living cells. The endoplasmic reticulum (ER) is chosen as the target structure due to its dynamic nature over short and long timescales. The reduced irradiation intensity combined with denoising permits continuous ER dynamics observation in living cells for up to 7 h with a temporal resolution of seconds. This allows for quantitative analysis of ER structural features over short (seconds) and long (hours) timescales within the same cell, and enabled fast 3D live‐cell STED microscopy. Overall, the combination of ultralow irradiation with image restoration enables comprehensive analysis of organelle dynamics over extended periods in living cells.

show abstract

Live-cell imaging in the deep learning era

Cited by 20 publications

References 69 publications

DL4MicEverywhere: Deep learning for microscopy made flexible, shareable, and reproducible

DL4MicEverywhere: Deep learning for microscopy made flexible, shareable, and reproducible

Accurate Identification of Cancer Cells in Complex Pre‐Clinical Models Using a Deep‐Learning Neural Network: A Transfection‐Free Approach

Fast and Long‐Term Super‐Resolution Imaging of Endoplasmic Reticulum Nano‐structural Dynamics in Living Cells Using a Neural Network

Contact Info

Product

Resources

About