Spinal cord injuries can result in substantial loss of trunk control, imposing significant challenges for affected individuals. Body-machine interfaces driven by trunk motion offer the potential for controlling assistive devices while promoting trunk rehabilitation. This work introduces a novel EMG/IMU-based interface for monitoring muscle activity and optimizing trunk rehabilitation outcomes. Our preliminary results show that the interface can seamlessly map trunk motion to a 2D visual interface, using EMG or IMU signals as control inputs. Despite achieving precise EMG control through trunk motions, new control strategies must be investigated to reduce user fatigue and enable targeted muscle rehabilitation.