CellposeCellCounter wins Rep. Sarah Elfreth’s 2025 Congressional App Challenge in Maryland’s Third District

When asked what inspired the creation of CellposeCellCounter, Matthew Yang said, “I was inspired to create CellposeCellCounter during my time interning in Dr. Liang’s lab at the University of Maryland School of Medicine. In the wet lab, I saw with my own eyes how crucial cell count, confluency, and viability are for reproducible research—yet how tedious they were to obtain. Most scientists still count cells manually with a hemocytometer, a time-consuming process that everyone in the lab agreed was frustrating. Automated counters existed, but they were too expensive for many labs to afford.

“Having read research demonstrating the capability of deep-learning segmentation models to identify and detect cells in images, I had the idea to build an accessible, AI-driven solution to replace both the tedious and costly current methods. Using my coding and machine learning skills, I built CellposeCellCounter to automatically measure cell metrics from a single uploaded image. I built the app in hopes that it would replace the need for expensive cell counters, which can cost upwards of $5,000. I knew that it was a trying time for science research, which meant that underresourced labs would greatly benefit from my app.

“In fact, my lab didn’t have an automated cell counter, and everyone relied on one single hemocytometer. When I presented my prototype, every hand in the room went up after I asked who found manual counting annoying—confirming the need for my project. Given this response, I was further inspired to make my app open-source on Hugging Face Spaces, allowing other researchers to use it whenever they needed.

“My app is also inspired by the many innovative integrations of AI in biology. Projects I’ve heard of and admire, such as SYBIL, the AI model that can predict lung cancer given a single CT scan, motivate me to pursue similar innovation. Seeing how technology can transform biomedical research inspired me to apply those same principles to microscopy—using machine learning not just to analyze images, but to make powerful research tools accessible to every lab, regardless of resources.”