BIA seminar series
BIA seminar series
The primary goal of this seminar series is to provide a dynamic platform for bioimage analysts, enabling the community to stay up to date with the latest developments in the field and foster community interactions. The seminars are designed to cater to intermediate and advanced analysts, focusing on practical, high-level content that extends beyond basic instruction.
Target Audience
The seminars are tailored for bioimage analysts who have intermediate to advanced knowledge in the field. This includes professionals and academics who are currently practising or researching within the domains of bioimage analysis, microscopy, and related life sciences. Themes are designed to cover relevant aspects and latest “news” of bioimage analysis, ensuring a comprehensive learning experience for intermediate to advanced analysts. Each session may feature one or two talks, depending on the complexity of the topic and the expertise of the speakers.
The seminar series will concentrate on a topic for three sessions in a row. The seminar will usually be in the last week of the month, but to help engagement of the global community we will not have a fixed day and time.
Planned topics so far include:
Large Language Models
Infrastructure for deploying image analysis workflows
Current and future plans for essential BIA tools
Best practices to build a bioimaging project starting from a user request: what to include, what to ask, how to deliver it?
to be continued.....
Registration
Please register for the seminar series here: BIA seminar series.
The registration will be used to send you the link to the seminar sessions. You will be registered for all upcoming GloBIAS seminar sessions in the future.
Second Topic
Infrastructure for deploying image analysis workflows
Please find below the information for the second set of GloBIAS seminars starting in November.
Designing and implementing systems to process petabyte scale imaging datasets: Lessons learned
Ankur Kumar, Computational Associate, Broad Institute of MIT and Harvard, Cambridge, USA
Wednesday, 27th of November 2 - 4 pm EST/ 8 - 10 pm CET
Abstract: Biological dataset sizes have increased exponentially in the last decade with no signs of slowing down. Processing these datasets is no longer trivial and requires careful planning and system design.
I will give an overview of the fundamental building blocks for creating such compute pipelines, drawing examples and lessons from my experience managing and processing datasets from the CellPainting Gallery.
NextFlow for Bioimage analysis
Christian Tischer, Team Leader, Data Science Center, EMBL, Heidelberg, Germany
Wednesday, 22nd of January 2025 4pm CET
Abstract: Ideally, bioimage analysis workflows should be reproducible, scalable and executable on various compute infrastructures. To this end, a number of bioimage analysts from different institutions started using Nextflow to deploy and share bioimage analysis workflows both locally and on high performance computing platforms. In this seminar, I will introduce Nextflow and show examples for how it can be used in the context of bioimage analysis.
Image analysis using Galaxy
Beatrix Serrano-Solano, Euro-BioImaging ERIC Bio-Hub, EMBL Heidelberg, Germany
& Anne Fouilloux , Simula Research Laboratory, Oslo, Norway
& Leonid Kostrykin, Biomedical Computer Vision Group, Heidelberg University, BioQuant, IPMB, Heidelberg, Germany
Wednesday, 26th of February 2025 2pm to 3:30 pm CET
Abstract: This webinar will introduce the Galaxy Image Analysis Community and highlight our mission to advance the development of FAIR and reproducible image analysis workflows. As part of our commitment to making image data analysis more accessible and collaborative, we will showcase how Galaxy can serve the imaging community. The session will explore Galaxy’s capabilities for integrating popular image analysis tools, interactive environments, and notebooks, making it a versatile platform for researchers across various scientific domains. We will also present how Galaxy facilitates the creation and sharing of reusable workflows, promoting open science and fostering collaboration. To give participants hands-on insight, we’ll provide a live demonstration on designing and running image analysis workflows within Galaxy.
Previous events:
Large Language Models for Bioimage Analysis
Exploring how emerging AI technologies can be applied to enhance image analysis tasks.
Large Language Models – an introduction for life scientists
Robert Haase, Image Data Scientist, Lecturer, Training Coordinator, ScaDS.AI / University of Leipzig, Germany
Tuesday 27th August 2024, 3pm - 4pm CEST
Large Language Models (LLMs) are changing the way how humans interact with computers. This has impact on all scientific fields by enabling new ways to achieve for example data analysis goals. In this talk we will go through an introduction to LLMs with respect to applications in the life sciences, softly focusing on bio-imaging. We will learn how to generate text and images using LLMs and how LLMs can extract information from text and images. We will go through selected prompt engineering techniques enabling scientists to tune the output of LLMs towards their scientific goal and how to do quality assurance in this context.
Please check out the follow-up discussion on Image.sc
BioImage.IO - introduction and application
Caterina Fuster-Barceló, Postdoctoral Researcher, Universidad Carlos III de Madrid, Spain
Friday 27th September 2024, 2pm - 3pm CEST
The dynamic field of bioimage analysis continually seeks innovative tools to democratize access to analysis tools and its documentation. The BioImage.IO Chatbot, leveraging state-of-the-art AI technologies including Large Language Models (LLMs) and Retrieval Augmented Generation (RAG), provides an interactive platform that significantly integrates the exploration and application of bioimage analysis tools and models. This seminar will introduce the BioImage.IO Chatbot's capabilities, focusing on how it facilitates access to advanced analysis tools and documentation, allows for the execution of complex models, and enables users to create customized extensions adjusted to specific research needs. In a live demo, attendees will see how to interact with the chatbot and all its assistants and capabilities. Join us to explore how the BioImage.IO Chatbot ca transform your research by making sophisticated analysis more intuitive and accessible.
Please check out the follow-up discussion on Image.sc
Exploring how emerging AI technologies can be applied to enhance image analysis tasks.
Harnessing Large Language Models for Bioimage Analysis with Omega
Dr. Loïc Alain Royer, Senior Group Leader & Director of Imaging AI, Chan Zuckerberg Biohub, USA
Friday 18th of October, 5:30pm - 7pm CEST
Abstract: This seminar will introduce Omega, a tool that utilizes large language models (LLMs) to enhance bioimage analysis. Integrated with the napari viewer, Omega offers a comprehensive range of capabilities including image processing, segmentation, and analysis. Key features include executing Python code, performing web searches, creating custom widgets, and interacting with segmentation algorithms like Cellpose and StarDist. Additionally, Omega can perform image denoising, generate detailed image processing strategies, and visually interpret the napari viewer's contents using LLMs. The tool also includes an AI-augmented code editor for code management and repair, an interactive didactic mode for educational purposes, and support for various coding libraries such as NumPy, scikit-image, and OpenCV. We will discuss challenges like potential errors and the need for careful prompt engineering. Practical examples will demonstrate how to use Omega to make widgets and accelerate the building of bioimage analysis workflows.