
CYTOREADER is a cloud based deep learning application supporting cervical pre-cancer screening based on dual stain (p16/Ki67) liquid cytology (CINtec® plus). CYTOREADER comprises a fully automated workflow to digitize and classify glass slides for patient referral to colposcopy. It provides a web interface for rapid AI-assisted diagnostics of cervical pre-cancer screening.
The heart of the application is CYTOREADER AI EVALUATOR, the CYTOREADER dual stain detection algorithm, which is a deep neural network for the detection of dual stain positive cells. As CYTOREADER runs in the cloud, it has technically unlimited scalability in throughput and unlimited geographical outreach. It increases efficiency of slide interpretation, and provides diagnostic procedures with objectivity and high accuracy globally, especially also in settings that lack personnel, expertise or infrastructure. In full automation, CYTOREADER has shown superior accuracy compared to manual DS reading with less colposcopy referrals.
CYTOREADER emerged from a scientific project between the Steinbeis Center for Medical Systems Biology (STCMED) in cooperation with the Division of Cancer Epidemiology & Genetics of the US National Cancer Institute (NCI) and Kaiser Permanente Northern California (KPNC). CYTOREADER is currently used as research-only (RUO) in epidemiological studies.
References
|