A system for generating surgical procedure training media draws upon the realistic data of an actual surgical procedure for realistic training without the risks. A 3D capturing component records three-dimensional model plus imaging data over time of a portion of a patient's body undergoing a surgical procedure. A spatial detection system detects an orientation of a surgical instrument relative to the patient's body during the surgical procedure. A modeling component creates a four-dimensional model (3D model+time) of the portion of the patient's body. Animation such as contingent events, trainee prompts, a virtual surgical instrument, etc., can be added to the model to expand upon the training potential. A user interface processes and edits training media for playback of the four-dimensional model including defining triggers responsive to a trainee simulated surgical inputs to pace sequencing of playback. An interactive player responds to pacing the playback of the editing training media or to a spatially detected simulated surgical instrument held by the student for direct tissue interaction.