Lisen Imprinting Diagnostics is working with ChristianaCare to clinically validate its lung cancer detection assay, built around quantifying “loss of imprinting” in ambiguous biopsy cases. The collaboration aims to validate the company’s machine-learning model on real-world samples using its QCIGISH in-situ hybridization workflow. In a previously reported blinded set of 155 patients, QCIGISH-based modeling showed 99.1% sensitivity and 92.1% specificity for lung malignancy classification. LisenID said the ChristianaCare study focuses on validating the model rather than the underlying biomarker biology. The test is positioned as a decision-support tool for pathologists, particularly when standard analysis leaves uncertainty. LisenID’s use of liquid lavage samples and image recognition tied to imprinting expression thresholds is designed to deliver an actionable malignant versus benign call for follow-up decisions.