QPath-OSCC: result tables and predictions for a controlled hybrid quantum-classical benchmark on oral squamous cell carcinoma pathology (v12)

Derived result tables and model predictions (test AUCs, statistical tests, calibration, uncertainty, decision-curve data) accompanying the revised manuscript "Task-Selective Hybrid Quantum-Classical Multi-Task Learning for Oral Squamous Cell Carcinoma Pathology Assessment: A Controlled Benchmark Against Matched Classical Comparators" (Discover Artificial Intelligence). Five backbone seeds; a classically simulated variational quantum circuit head is compared with parameter-matched and larger classical heads for lymph-node-metastasis prediction. The Multi-OSCC images and labels (Guan et al., Scientific Data, 2026) are third-party and not included; source code is not included. No physical quantum computer was used.