Highlights from ISBI 2025 in Houston, Texas

This year’s International Symposium on Biomedical Imaging (ISBI) was held in Houston, Texas—home to NASA’s Johnson Space Center and the sprawling Texas Medical Center. With sunny skies and wide Texan roads welcoming us, we arrived two days early to make the most of the trip before diving into 4 days of cutting-edge research. We kicked off our trip with an NBA game at the Toyota Center and followed it up with a tour of the NASA Space Center. Walking through the rocket park and seeing the Saturn V rocket and the SpaceX Falcon 9 Booster up close was an unforgettable reminder of how science can propel us (literally!) to new frontiers.

The conference officially began on Monday and brought together researchers from around the world to share and discuss breakthroughs in biomedical imaging. Out of 631 accepted papers, 190 were selected for oral presentation, covering a diverse range of topics, from image segmentation and multimodal fusion to foundation models, each sparking new questions and ideas. Another highlight for us was the series of keynote talks from leaders in the field. Topics ranged from precision imaging for early cancer detection to bias in AI and domain adaptation, offering insights into both current trends and where the field may be heading next. The speakers not only shared their research but also reflected candidly on the challenges ahead, making the sessions both inspiring and thought-provoking.

Beyond the presentations and posters, ISBI 2025 was a great place to network. We had the chance to exchange ideas with fellow researchers, hear about the hurdles they’re tackling in their work, and find common ground that might lead to future collaborations. There’s always something uniquely energising about being surrounded by people who are just as excited about pixels and pathology as you are.

Contributions from our lab:

Jiaqi Lv [Poster Presentation]: Deep Learning Based Segmentation of Blood Vessels from H&E Stained Oesophageal Adenocarcinoma Whole-Slide Images .

Esha Nasir [Abstract Poster Presentation]: OMReg-Optimized Multimodal Whole Slide Image Registration.

Yijie Zhu [Oral Presentation]: Gland Segmentation Using SAM With Cancer Grade as a Prompt.

We’d like to thank our collaborators and the broader ISBI community for a fantastic week in Houston—full of discovery, dialogue and inspiration.

Yijie Zhu and Jiaqi Lv