AI4Life

AI4Life_Logo
Menu

Category: News

Categories
News

AI4Life’s Highlights at I2K 2024

AI4Life’s highlights at I2K 2024

by Beatriz Serrano-Solano

AI4Life played an important role at the 2024 Images to Knowledge (I2K) conference, held from October 23-25 at the Human Technopole in Milan. As a major event for experts at the intersection of life sciences and quantitative image analysis, I2K offered AI4Life partners an ideal platform to showcase their latest advancements and community contributions.

AI4Life’s impact began at the organizational level, with the AI4Life scientific coordinators Florian Jug and Anna Kreshuk serving on the scientific organizers committee. The conference also welcomed Ricardo Henriques from Instituto Gulbenkian de Ciência (IGC) – University College London (UCL) as an invited speaker, highlighting AI4Life’s expertise in superresolution microscopy.

IMG_20241025_235937
Hands-on learning with specialised workshops

A series of workshops allowed participants to dive into AI4Life’s practical tools and methods:

  • Building Your Own Chatbot for BioImage Analysis: Opening the series on Wednesday.
  • Accelerating Microscopy Image Annotation with SAMJ Annotator and DL4MicEverywhere: Making your deep learning pipelines flexible, shareable, and reproducible workshops on Thursday, enhancing image annotation speed and flexibility.
  • Friday’s sessions included 4 workshops on BiaPy: deep learning based Bioimage Analysis for all audiences, Recent Advances in ilastik, AI4Life: Empowering BioImaging through the BioEngine, and an Introduction to AI4Life and the BioImage Model Zoo
Poster presentations

AI4Life partners presented two posters:
– BioImage.IO Chatbot: A community-driven AI assistant advancing integrative computational Bioimaging, presented on Wednesday.
– Building Foundations for AI-Driven Bioimage Analysis: Discussing infrastructure and annotation platforms critical for bioimage analysis, shared on Thursday. 

 

We look forward to the next I2K and are excited to contribute further to this inspiring community!

Categories
News

Summary of the AI4Life 3rd General Assembly

Summary of the AI4Life 3rd General Assembly

by Beatriz Serrano-Solano

The AI4Life 3rd General Assembly, held at Human Technopole in Milan on October 21-22, 2024, brought together project partners and collaborators to discuss advancements, challenges, and future directions in AI-driven image analysis.

Key Highlights

The event started with the welcoming remarks by Dorothea Dörr (AI4Life Project Manager) and John Eriksson (Euro-BioImaging Director General and coordinator of the project), followed by an introductory reflection on AI4Life’s achievements over the past two years by Anna Kreshuk and Florian Jug, scientific coordinators of the project.

 
Session 1 – AI4Life services

The first session focused on demonstrations for creating, uploading, and consuming models.
* Demo 1: Showcased how users can create and upload models via a web interface or programmatically through an API.
* Demo 2: Demonstrated model consumption from bioimage.io, community partner tools, and code environments.
* Demo 3: Emphasized the importance of FAIR principles, highlighting AI-ready datasets, FAIR models, and documentation enhancements.

Session 2 – Work Package updates

Work package leads updated the attendees on infrastructure and user services, with WP6 announcing that AI4Life’s 3rd Open Call will be launched soon.

Session 3 – Additional Work Package Presentations

Presentations from WPs 1, 3, 5, and 7 covering project coordination, direct support, data standards, and outreach.

Research Infrastructure Updates

Brief updates from the partner Research Infrastructures: INSTRUCT-ERIC, EU-OPENSCREEN ERIC, EMPHASIS, and EMBRC-ERIC.

Session 4 – Planning the Future

A forward-looking session exploring upcoming project collaborations with BigPicture and AI4EOSC, along with a panel discussion on project sustainability and post-project impact.


The assembly concluded with plans for year three’s demos, events, and outreach initiatives, with final remarks from Anna Kreshuk and Florian Jug.

Categories
News

AI4Life in Chemical & Engineering News

AI4Life in Chemical & Engineering News

by Beatriz Serrano-Solano

In October 2024, Chemical & Engineering News explored how artificial intelligence is revolutionizing microscopy and featured AI4Life’s role in bridging the gap between life scientists and AI experts. AI4Life was recognized for making advanced AI tools more accessible to researchers, supporting their diverse needs, and fostering collaboration across disciplines.

Read the full article here.

Categories
News

AI4Life Denoising Challenge 2024: Results

AI4Life Denoising Challenge 2024: Results

by Vera Galinova, Beatriz Serrano-Solano

The AI4Life Denoising Challenge, which aimed to improve the denoising of microscopy images using deep learning, has been successfully completed. With 104 registrations from 27 countries, the challenge gathered participation from the global scientific community, generating 151 total submissions across four distinct leaderboards featuring different datasets containing one of two noise types. 

What was the challenge about?

Microscopy images are indispensable for biological and medical research, but noise introduced during acquisition can impair image quality and complicate interpretation. The challenge’s focus was the unsupervised denoising task, which, unlike supervised learning, does not require pairs of noisy and clean images, simplifying the real-world application of the algorithms. Researchers applied advanced deep learning algorithms to real-world datasets, aiming to improve image quality while preserving essential features such as edges, textures, and fine details. Find all the information on the challenge at https://ai4life.eurobioimaging.eu/denoising-challenge/

Key statistics

  • Total Participants: 104 registrants from 27 countries
  • Total Submissions: 151
  • Leaderboard Submissions:
  • Structured Noise 1: 17 submissions from 7 participants
  • Structured Noise 2: 7 submissions from 5 participants
  • Unstructured Noise 1: 16 submissions from 7 participants
  • Unstructured Noise 2: 13 submissions from 8 participants

Results by category

Structured Noise 1: Hagen et al (50 total submissions)
 

Name

Affiliation

Algorithm

Result

SI-PSNR

Result

SSIM

Code

1

bensalmon

University of Birmingham

COSDD

31.4847

0.5523

Link to code

2

mcroft

Human Technopole

  N2V

30.8962

0.5534

Link to code

3

shanetoy

Seoul National University

  N2V

30.8406

0.5525

Link to code 
Structured Noise 2: SUPPORT (20  total submissions)
 

Name

Affiliation

Algorithm

Result

SI-PSNR

Result

SSIM

Code

1

bensalmon

University of Birmingham

COSDD

30.4415

0.6278

Link to code

2

a897574323

University of Shanghai for Science and Technology

  N2V

29.7744

0.6112

Link to code

3

edoardogiacomello

Human Technopole

  N2V

28.8547

0.5272

Link to code
Unstructured Noise 1: JUMP (47  total submissions)
 

Name

Affiliation

Algorithm

Result

SI-PSNR

Result

SSIM

Code

1

bensalmon

University of Birmingham

COSDD

35.6282

0.9507

Link to code

2

mcroft

Human Technopole

N2V

35.4957 

0.9413

Link to code

3

edoardogiacomello

Human Technopole

N2V

35.4957

0.9412

Link to code
Unstructured Noise 2: W2S (34 total submissions)
 

Name

Affiliation

Algorithm

Result

SI-PSNR

Result

SSIM

Code

1

bensalmon

University of Birmingham

COSDD

35.6855

0.9163

Link to code

2

edoardogiacomello

Human Technopole

N2V2

35.0505 

0.9025

Link to code

3

edoardogiacomello

Human Technopole

N2V

35.0319 

0.9027

Link to code

What’s next?

The AI4Life Denoising Challenge provided a platform for benchmarking denoising methods across various datasets. The challenge showed how unsupervised deep learning algorithms, like COSDD and N2V, can significantly enhance image quality. The leaderboard results asserted the robustness of these methods across different datasets and noise types. 

We encourage researchers to continue to participate in the challenge through late submissions. Please contact us if you want to participate by submitting your algorithm, and the challenge team will assist you. 

Looking ahead, AI4Life plans to host more challenges soon. Stay tuned for upcoming announcements!

 

Categories
News

AI4Life joins the AI4Europe Community

AI4Life joins the AI4Europe Community

by Beatriz Serrano-Solano

We are excited to announce that AI4Life has officially joined the AI4Europe community, a collaborative platform dedicated to advancing Artificial Intelligence research and innovation across Europe.

The AI-on-Demand (AIoD) platform serves as a community-driven channel designed to ensure quality, trustworthiness, and explainability in AI solutions. Here are some key benefits:

  • Enhanced collaboration: AI4Life members can engage with peers and experts across various disciplines.
  • Access to resources: The AIoD platform provides a wealth of resources, including datasets, tools, services, and educational courses.
  • Dissemination: the platform provides channels to share research outputs, news updates or forthcoming events.

Explore the AI4Life record at https://www.ai4europe.eu/ai-community/projects/ai4life

Categories
News

AI4Life joins FAIRsharing

AI4Life joins FAIRsharing

by Beatriz Serrano-Solano

AI4Life has officially joined the FAIRsharing community. By registering our records on FAIRsharing, we are reaffirming our commitment to the FAIR principles, which are crucial for the sustainable management and sharing of data and AI models.

AI4Life’s contributions to FAIRsharing include a set of guidelines and resources designed to operationalize the FAIR principles within the context of AI and bioimaging. Some of the key resources and guidelines now available on FAIRsharing include:

BioImage Model Zoo Model Specification

A standard format for documenting and disseminating AI models, ensuring cross-compatibility with bioimaging tools and easy integration into research workflows. Explore the record here.

Guidelines for AI-Ready Datasets (MIFA)

Detailed guidance on metadata, formats, and accessibility to make datasets AI-ready and compliant with FAIR standards. Explore the record here.

BioImage.IO Repository

The BioImage Model Zoo is community-driven repository of pre-trained AI models. Explore the record here.

AI4Life Collection

All these records are collected under the AI4Life umbrella record. https://fairsharing.org/collection/AI4Life

For more details on our FAIRsharing records and how they can benefit your research, take a look at our recent deliverable:

 https://zenodo.org/records/13618545 

Categories
News

BioImage.IO Chatbot featured in Nature Methods

BioImage.IO Chatbot featured in Nature Methods

by Beatriz Serrano-Solano

The BioImage.IO Chatbot has been featured in Nature Methods as of August 2024. This tool is designed to make advanced bioimage analysis accessible to researchers at all levels using Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) to offer a user-friendly interface to perform complex image analysis tasks.

To explore its capabilities and learn how it can accelerate your research, check out the full article published on FocalPlane:

BioImage.IO Chatbot: Ready to use and discover!

Read the full article here.

Categories
News

New AI4Life section: services

New Website Section: AI4Life Services

by Beatriz Serrano-Solano

We’re launching a new section on the AI4Life website dedicated to our service offering to the bioimaging community, making AI and deep learning techniques more accessible to researchers worldwide.

From the BioImage Model Zoo, a community-driven AI model repository, to the BioImage Archive, which ensures long-term data preservation, BioEngine, our cloud infrastructure that simplifies bioimage analysis, and the containers that we offer for reproducible deep learning.

Visit our website to learn more about these services and how they can support your work in bioimaging: https://ai4life.eurobioimaging.eu/services/ 

Categories
News

AI4Life launches digital badges

AI4Life launches digital badges

by Beatriz Serrano-Solano

AI4Life is now using Badgr/Canvas Badges to issue and share digital badges with contributors to the project. Here you’ll find how to use them.

Step 1. Create an account in Badgr

If you’ve received an email from Badgr (noreply@badgr.com) with the subject “You earned <name of the badge> badge!”, congratulations, you can now add it to your social media channels!

To access your new badge, you need to create an account in Badgr if you don’t have one already:

https://badgr.com/

The badge will appear in your Badgr backpack (Figure 1).

Figure 1. Backpack view at Badgr
Step 2. Share your badge on LinkedIn

Now you’re ready to share the badge in your networks using the provided personalised URL, here’s how!

Click on “share” on the badge that appears on your backpack and then on “Add to profile” (Figure 2).

Figure 2. Pop-up window to share the Badge in social media

A new pop-up window will appear with the name of the badge (e.g. “Reviewer of the 1st/2nd AI4Life Open Call” or “Awardee project of the AI4Life Open Call“) and the issuing organization (Figure 3). At this stage, you need to change the issuing organization to “AI4Life EU project” (should autocomplete when you start typing) as you can see in Figure 4.

Figure 3. Pop-up window with the original issuing organization
Figure 4. Updated issuing organization

Click on save and you’re good to go!

Step 3. Check your profile and share with your contacts

Now you can share on other platforms too!

Categories
News

Opportunity alert: HT launches a call for access to national facilities for data handling and analysis

Opportunity alert: HT launches a call for access to national facilities for data handling and analysis

by Joran Deschamps & Damian Edward Dalle Nogare

As part of the National Facility for Data Handling and Analysis at Human Technopole, the Bioimage Analysis infrastructure unit aims at providing state-of-the-art analysis of biomedical imaging data to researchers working at publicly-funded Italian institutions. As part of this mission, we are pleased to announce our 2024 open call for bioimage analysis projects. This open call allows users to submit analysis projects to our facility based on data generated at their own institutes. Successful applicants will collaborate with our analysis team to define the scope and goals of the project, and to develop analysis pipelines to meet their scientific objectives.

This call is currently open, with the first application review period scheduled for September. More information, including a description of the program, eligibility criteria, deadlines, and instructions to apply can be found at this link (pdf).