MMV_Im2Im: an open-source microscopy machine vision toolbox for image-to-image transformation

GigaScience, Journal Year: 2024, Volume and Issue: 13

Published: Jan. 1, 2024

Over the past decade, deep learning (DL) research in computer vision has been growing rapidly, with many advances DL-based image analysis methods for biomedical problems. In this work, we introduce MMV_Im2Im, a new open-source Python package image-to-image transformation bioimaging applications. MMV_Im2Im is designed generic framework that can be used wide range of tasks, including semantic segmentation, instance restoration, generation, and so on. Our implementation takes advantage state-of-the-art machine engineering techniques, allowing researchers to focus on their without worrying about details. We demonstrate effectiveness more than 10 different problems, showcasing its general potentials applicabilities. For computational researchers, provides starting point developing or algorithms, where they either reuse code fork extend facilitate development methods. Experimental benefit from work by gaining comprehensive view concept through diversified examples use cases. hope give community inspirations how integrated into assay process, enabling studies cannot done only traditional experimental assays. To help get started, have provided source code, documentation, tutorials at [https://github.com/MMV-Lab/mmv_im2im] under MIT license.

Language: Английский

---

EfficientBioAI: making bioimaging AI models efficient in energy and latency

Nature Methods, Journal Year: 2024, Volume and Issue: 21(3), P. 368 - 369

Published: Jan. 24, 2024

Language: Английский

---

Self‐Driving Microscopes: AI Meets Super‐Resolution Microscopy

Small Methods, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 10, 2025

Abstract The integration of Machine Learning (ML) with super‐resolution microscopy represents a transformative advancement in biomedical research. Recent advances ML, particularly deep learning (DL), have significantly enhanced image processing tasks, such as denoising and reconstruction. This review explores the growing potential automation microscopy, focusing on how DL can enable autonomous imaging tasks. Overcoming challenges automation, adapting to dynamic biological processes minimizing manual intervention, is crucial for future microscopy. Whilst still its infancy, revolutionize drug discovery disease phenotyping leading similar breakthroughs been recognized this year's Nobel Prizes Physics Chemistry.

Language: Английский

---

Deep-learning-based image compression for microscopy images: An empirical study

Biological Imaging, Journal Year: 2024, Volume and Issue: 4

Published: Jan. 1, 2024

Abstract With the fast development of modern microscopes and bioimaging techniques, an unprecedentedly large amount imaging data is being generated, stored, analyzed, shared through networks. The size poses great challenges for current infrastructure. One common way to reduce by image compression. This study analyzes multiple classic deep-learning-based compression methods, as well empirical on their impact downstream processing models. We used label-free prediction models (i.e., predicting fluorescent images from bright-field images) example task comparison analysis Different techniques are compared in ratio, similarity, and, most importantly, accuracy original compressed images. found that artificial intelligence (AI)-based largely outperform ones with minimal influence 2D tasks. In end, we hope this could shed light potential raise awareness impacts deep-learning analysis.

Language: Английский

---

2D Label-free Prediction of Multiple Organelles Across Different Transmitted-light Microscopy Images with Bag-of-Experts

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: May 24, 2024

ABSTRACT Label-free prediction has emerged as a significant application of artificial intelligence (AI) in the field bioimaging, which aims to predict localization specific organelles directly from readily-accessible transmitted-light images, thereby alleviating need for acquiring fluorescent images. Despite existence numerous research, practice, high variability imaging conditions, modalities, and resolutions poses challenge final prediction. In this study, we propose “Bag-of-Experts” strategy, targeting at different organelles, with self-supervised pre-training. The comprehensive experimentation showcases that our model is agnostic image modalities certain extent, indicating considerable generalizability. code released at: https://github.com/MMV-Lab/LightMyCells

Language: Английский

---

2D Label-Free Prediction of Multiple Organelles Across Different Transmitted-Light Microscopy Images with Bag-of-Experts

Published: May 27, 2024

Language: Английский

---

Research on the Application of Deep Learning in Biomedical Image Processing

Published: March 4, 2024

Language: Английский

---