Mermod M, Jourdan EF, Gupta R, et al. - Researchers sought to develop, compare, and validate several machine learning models to predict occult lymph node metastases in clinically N0 oral squamous cell carcinoma (OSCC). On a training cohort (n = 56), they tested the combination of the biomarkers, CD31 and PROX1, with relevant histological parameters using four different state-of-the-art machine learning models. Then, they tested the optimized models on an external validation cohort (n = 112) of early-stage (T1-2 N0) OSCC. With the best overall performance and accuracy, the random forest model maintained a negative predictive value > 95%. This suggests the utility of this new clinical decision algorithm for significantly improving the management of patients with early-stage OSCC.