Roche adds PD-L1-spotting AI to its non-small cell lung cancer pathology platform

Roche digital pathology
Roche's automatic algorithm is designed to determine whether tumors are positive for the PD-L1 biomarker and eligible for particular therapies. (Roche)

Roche has launched a new version of its digital pathology platform in Europe, with a CE-IVD mark for an artificial-intelligence-powered image analysis algorithm aimed at non-small cell lung cancer.

The uPath PD-L1 software provides assessments of scanned slides to assist in the diagnosis of patients and selection of targeted treatment options, connected with the company’s Ventana PD-L1 assay.

The automatic algorithm is designed to determine whether tumors are positive for the PD-L1 biomarker and eligible for particular therapies while highlighting positively and negatively stained tumor cells with a visual overlay.

The algorithm received a CE mark as an in vitro diagnostic to review and classify tissues and cells based on their morphology, color, intensity, size, pattern and shape.

RELATED: Roche launches uPath multislide digital pathology software

“Improving diagnostic consistency and certainty is crucial in providing faster, higher-quality and more accurate diagnoses to cancer patients,” said Thomas Schinecker, CEO of Roche Diagnostics. “Our uPath PD-L1 (SP263) image analysis for non-small cell lung cancer is the first next-generation CE-IVD PD-L1 algorithm to the clinical market.”

Roche’s uPath enterprise software was first launched in January 2019, designed to present pathologists with all available slides a single view, alongside the patient’s health information as part of an end-to-end system.