GPU acceleration partner
Upload a digital fine-needle aspiration slide. CytoLens returns a Bethesda classification with a calibrated confidence score as a second read for your cytopathologists.
[ workflow ]
Three steps inside your existing digital-pathology workflow. CytoLens assists the cytopathologist; it does not replace the sign-out.
Submit a digital FNA whole-slide image through the secure web platform. No installation, no local hardware.
The NVIDIA-optimized model detects follicular cells, scores architectural and nuclear features, and maps them to the Bethesda system in under five minutes.
Receive a category with a calibrated confidence score and the regions that drove it, so your pathologist can confirm, adjust, or escalate.
[ platform walkthrough ]
[ platform ]
A first-pass read in under five minutes, freeing pathologist time for the cases that need it.
Outputs map to the categories your team already signs out against, with confidence on every call.
Runs in the browser. Nothing to install, no GPU to provision at the lab.
Developed on synthetic imagery, so no patient data was needed to build the model.
Built to connect with LIS and digital-pathology systems as validation progresses.
Expanding from current categories to the complete Bethesda system.
[ current status ]
We are direct about validation status, because the people evaluating CytoLens are too.
// Intended as a second-read assistant for qualified cytopathologists. Results support, and do not replace, professional diagnosis and sign-out.
[ the team ]
Software engineer and Visual Effects Society member. A decade in synthetic 3D data and computer vision across Hollywood, Microsoft HoloLens, and Meta Reality Labs, now applied to medical imaging.
Head & Neck surgeon with 30+ years in endoscopy, laser microsurgery, and reconstruction. Academic leader and clinical researcher, directing CytoLens medical validation.
from binary core
Binary Core generates synthetic training data for medical imaging AI: pathology, radiology, and surgical planning. If patient-data access, privacy review, or rare-case scarcity is the bottleneck for your model, that is exactly the problem we solve.
[ partner with us ]
We work with a small number of hospital and laboratory partners during validation. Request access and we'll walk your team through the platform.
we reply within 24 hours · San Diego, CA
GPU acceleration partner