image: This is an AFM image of 10x10 micron2 cell surface. view more
Credit: Image courtesy of Igor Sokolov and Maxim E. Dokukin
Researchers report a noninvasive method to detect bladder cancer by using machine learning to analyze nanoscale-resolution images of cells collected from urine; analysis of urine samples from 25 patients with bladder cancer and 43 cancer-free individuals revealed that the method can provide a diagnosis with 94% accuracy, a statistically significant improvement over cystoscopy, the clinical standard.
Article #18-16459: "Noninvasive diagnostic imaging using machinelearning analysis of nanoresolution images of cell surfaces: Detection of bladder cancer," by Igor Sokolov et al.
MEDIA CONTACT: Igor Sokolov, Tufts University, Medford, MA; 617-627-2548; email: igor.sokolov@tufts.edu
###
Journal
Proceedings of the National Academy of Sciences