In this paper we propose a new scheme for measuring regional ventilation from tagged hyperpolarized helium-3 MR images. A new registration cost function that incorporates both the intensity information (SSD) and the shape feature (SSBMD) from the images is proposed for registering end inspiration to the end expiration image. The smoothness of the displacement field is maintained by incorporating the Laplacian regularization constraint (LAP) in the total cost function. The ventilation is quantified using the Jacobian determinant of the resulting displacement field from the proposed registration algorithm. Tags are automatically segmented from the images to evaluate the registration accuracy. The average tag positioning error is on the order of 2 mm after registration for all three subjects. These results may provide new method for assessing regional lung ventilation and may be used to track regional function changes of lung cancer patients following radiation therapy.