3 ms·Merlin: A computed tomography vision–language foundation model and dataset2 points by bryanrasmussen 4mo ago