News & Events
It’s always an honor to have our work and research mentioned by the media and in the press. We are especially pleased when our work reaches readers outside the scientific community. Read on to see some of the latest stories about our findings.
Bio-Lab Activities Annual Report: January 04th, 2024
An Year of Innovation, Collaboration, and Fun!
Generative AI and Ethics - The Heart of Our Story
Imagine a place where every idea is valued, where the coffee pot is always brewing, and a crossroads of brilliant collaborations lead to groundbreaking discoveries. That's our lab! 🗣️ Voices from the Lab We kicked off this report with two thought-provoking essays by lab members reflecting on Generative AI and Ethics in AI. These aren't just essays, folks – they're windows into the future, penned by our very own BioAI wizards! Members also shared personal insights, quirky anecdotes, and brainy opinions on our focal points. It's like a kaleidoscope of brainpower! 🌍✨Collaborative Success The past year has been a testament to the power of collaboration, creativity and diversity. This synergy has been pivotal in achieving our goals and propelling our lab to new heights. Our PhDs, postdocs, engineers, and interns have been the cornerstone of our progress, demonstrating an unwavering commitment to excellence. We didn't just stay cooped up in the lab! We hosted a summer school that felt more like a festival of learning, an interactive LLM workshop that turned into an idea explosion, and an ICCV workshop that was basically the Oscars of AI! 🏆🎉 🔮 Looking to the Future As we celebrate our achievements, we also look forward with anticipation to the year 2024. Our annual report (right here, and yes, it's as cool as it sounds!) is just a glimpse of our journey, encapsulating the hard work and success of the past year. Here's to 2024, where the next chapter of our story awaits!
🌟 Stay tuned for more stories, and don’t forget to check out our full annual report – it’s not just a read; it's an experience! 🌟
Arctic LLM Workshop: October 27th -28th, 2023
Speaker Lineup: Bio-AI Lab Members
Teknologibygget Tek 1.023, UiT Tromsø, Norway
Location: Auditorium 1.023 av Teknologibygget, UiT NORWAY
Join us for a two-day comprehensive workshop on LLMs under the enchanting northern lights. Investigate the evolution of LLM models, comprehend fine-tuning mechanisms, and be captivated by the power of prompts. Examine the complexities of alignment, interpretability, and robustness issues. Discover high-speed and parameter-efficient LLM variants while understanding the challenges of large-scale training. Learn the ins and outs of LLM-based application development to stay current in the world of LLMs.
GUEST LECTURE: Thursday October 13th, 2022
Prof. Ananda Shankar Chowdhury
Jadavpur University, India
Location: Auditorium 1.023 av Teknologibygget, UiT NORWAY
We thank Prof. Ananda Shankar Chowdhury for an invited guest talk at UiT The Arctic University of Norway while there on sabbatical about the challenges and solutions of cancer detection in computer vision. =================================== About the Speaker: Ananda Shankar Chowdhury is a Professor and former Head in the department of Electronics and Telecommunication Engineering (E.T.C.E.) at Jadavpur University, Kolkata, India where he leads the Imaging, Vision and Pattern Recognition (IVPR) group. He received a B.Sc. (with Hons.) in Physics from the Presidency College in 1996, a B.Tech. from the Institute of Radiophysics and Electronics in 1999, and an M.E. in Computer Engineering from Jadavpur University in 2001, all from India. He earned his Ph.D. in Computer Science from The University of Georgia, USA in July 2007. He then worked as a post-doctoral fellow in the department of Radiology and Imaging Sciences at National Institutes of Health, USA during the period August 2007 to December 2008. His research interests are broadly in the areas of Computer Vision and Pattern Recognition with an emphasis on problems arising in Biomedical, Multimedia and Surveillance domains. Dr. Chowdhury has published more than ninety papers in leading international journals and conferences in addition to a monograph from the Springer series of Advances in Computer Vision and Pattern Recognition. He has held invited academic visits to University of Münster in Germany, National University of Singapore in Singapore, University of Campinas and University of Sao Paulo in Brazil, University of Padova in Italy, University of Groningen in The Netherlands, and University of La Rochelle in France. He is a senior member of IEEE and an IAPR TC-member of Graph based Representations in Pattern Recognition. He currently serves as an Area Editor for Pattern Recognition Letters and as a Senior Area Editor for IEEE Signal Processing Letters. Earlier, he served Springer Nature Computer Science as an Associate Editor. =================================== Target group: Undergarduate and garduate computer science and physics students, PhD students in computer science and physics, employees at the CS and Physics and Technology department, external from industry. =================================== Abstract: In this talk, I will take you to an eventful journey through different seasons of computer vision with an aim of detecting cancer. In particular, we will segment and classify lung nodules and brain tumors to achieve this goal. In the first season, which I consider as “Autumn”, we will segment various types of lung nodules from CT images using only level sets and without applying any form of learning . It will be then discussed that some remarkable progress in machine learning is however necessary in general to ensure success of computer vision for achieving complex goals. A tentative period, required for this growth, can be thought of as a somewhat unproductive second season, considered to be a harsh “Winter”. Then, comes the third season, which I consider as “Spring”. Here, many flowers, techniques from deep learning and classical computer vision blossom together. In this season, we will accurately segment brain tumors in 3D from MRI using a combination of deep learning and graph cuts . The journey will come to an end with a fourth and final “Rainy” season, where purely deep learning rains or reigns. Here, we will first show how brain tumors can be classified from radiology as well as histopathology data using deep features and graph convolutional networks . We will finally demonstrate how attributes driven generative adversarial networks can synthesize and classify various types of lung nodules . References  R. Roy, P. Banerjee, A.S. Chowdhury, A Level Set based Unified Framework for Pulmonary Nodule Segmentation, IEEE Signal Process. Lett. 27 (2020), 1465-1469.  A. De, M. Tewari, E. Grisan, A.S. Chowdhury, A Deep Graph Cut Model for 3D Brain Tumor Segmentation, Proc. Forty-fourth International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), Glasgow, Scotland, UK (2022), 2105- 2109.  A. De, R. Mhatre, M. Tewari, A.S. Chowdhury, Brain tumor classification from Radiology and Histopathology using Deep Features and Graph Convolutional Network, Proc. Twenty-Sixth International Conference on Pattern Recognition (ICPR); Montreal, Canada (2022) (in press)  R. Roy, S. Mazumdar, A.S. Chowdhury, ADGAN: Attributes Driven Generative Adversarial Network for Synthesis and Multi-class Classification of Pulmonary Nodules, IEEE Trans. Neural Netw. Learn Syst. (DOI: 10.1109/TNNLS.2022.3190331)
Conference Paper Presentation: March 09th, 2021
Bildverarbeitung fuer die Medizin 2021
Location: Virtual, OTH Regensburg, Germany
Abstract: Examining specific sub-cellular structures while minimizing cell perturbation is important in the life sciences. Fluorescence labeling and imaging is widely used for introducing specificity despite its perturbative and photo-toxic nature. With the advancement of deep learning, digital staining routines for label-free analysis have emerged as a replacement for fluorescence imaging. Nonetheless, digital staining of sub-cellular structures such as mitochondria is sub-optimal. This is because the models designed for computer vision are directly applied instead of optimizing them for the nature of microscopy data. We propose a new loss function with multiple thresholding steps to promote more effective learning for microscopy data. Through this, we demonstrate a deep learning approach to translate the label-free brightfield images of living cells into equivalent fluorescence images of mitochondria with an average structural similarity of 0.77, thus surpassing the state-of-the-art of 0.7 with L1. Our results provide insightful examples of some unique opportunities generated by data-driven deep-learning enabled image translations.
To be announced
28th May 2021
Registration deadline: August 20th
Starts: August 30, 2021 at 09.00
Ends: 15 December 2021 at 09.00
Where: Campus Tromsø
10th MAY 2021
Sebastian Acuña and Mayank Roy paper titled 'Deriving high contrast fluorescence microscopy images through low contrast noisy image stacks'
Get Notified for Future Events
Subscibe to the Lab for future Events and Research Activites.