Accurate cell nuclei segmentation is critical for downstream tasks in kidney pathology and remains a major challenge due to the morphological diversity and imaging variability of renal tissues. While our prior work has evaluated early-generation AI cell foundation models in this domain, the effectiveness of recent cell foundation models remains unclear. In this study, we benchmark advanced AI cell foundation models (2025), including CellViT++ variants and Cellpose-SAM, against three widely used cell foundation models developed prior to 2024, using a diverse large-scale set of kidney image patches within a human-in-the-loop rating framework. We further performed fusion-based ensemble evaluation and model agreement analysis to assess the segmentation capabilities of the different models. Our results show that CellViT++ [Virchow] yields the highest standalone performance with 40.3% of predictions rated as "Good" on a curated set of 2,091 challenging samples, outperforming all prior models. In addition, our fused model achieves 62.2% "Good" predictions and only 0.4% "Bad", substantially reducing segmentation errors. Notably, the fusion model (2025) successfully resolved the majority of challenging cases that remained unaddressed in our previous study. These findings demonstrate the potential of AI cell foundation model development in renal pathology and provide a curated dataset of challenging samples to support future kidney-specific model refinement.

The use of artificial intelligence to enable precision medicine and decision support systems through the analysis of pathology images has the potential to revolutionize the diagnosis and treatment of cancer. Such applications will depend on models' abilities to capture the diverse patterns observed in pathology images. To address this challenge, we present Virchow, a foundation model for computational pathology. Using self-supervised learning empowered by the DINOv2 algorithm, Virchow is a vision transformer model with 632 million parameters trained on 1.5 million hematoxylin and eosin stained whole slide images from diverse tissue and specimen types, which is orders of magnitude more data than previous works. The Virchow model enables the development of a pan-cancer detection system with 0.949 overall specimen-level AUC across 17 different cancer types, while also achieving 0.937 AUC on 7 rare cancer types. The Virchow model sets the state-of-the-art on the internal and external image tile level benchmarks and slide level biomarker prediction tasks. The gains in performance highlight the importance of training on massive pathology image datasets, suggesting scaling up the data and network architecture can improve the accuracy for many high-impact computational pathology applications where limited amounts of training data are available.

Several years ago, I attended a Buddhist retreat in which I was introduced to the idea of the "retinue," a constellation of influential and supportive people whom one imagines in an enveloping cloud as one meditates. I took the concept one step further and decided to create an actual photo montage that I could hang on the wall above my desk: my childhood piano teacher, my high school English teacher, my rabbi, mentors in science, writers who encouraged me--in all, 20 people who had profoundly influenced me. Some members of my retinue were still living, some not. In some cases I could find the photographs myself. In others, I had to contact the mentors. When I finally tracked down William Gerace, who introduced me to physics nearly 50 years ago, he was puzzled as to why I should desire such a montage. We had not spoken for decades. Reluctantly, he sent me an old, out-of-focus photo of himself, dating back to the days when I knew him. Now, Gerace is a professor of science education at the University of North Carolina at Greensboro, after a 30-year career as a professor of physics at the University of Massachusetts Amherst, during which time he made the transition from theoretical nuclear physicist to leader in science education and co-founder of the Scientific Reasoning Research Institute at the University of Massachusetts, Amherst. When I knew him, in the late 1960s, he was a lowly instructor in physics at Princeton, where he had recently received his Ph.D. I was an undergraduate. The photo shows a man in his late 20s, about 5 feet 6, slight in build, dark hair beginning to thin, dressed in a button-down shirt and blue sweater, and a Mona Lisa smile. Each new mathematical technique Bill taught us was offered with the enthusiasm of a 12-year-old boy showing his friend a strange new butterfly. I first met Bill Gerace during a physics lab my sophomore year.