DeepRF-Grad: Simultaneous design of RF pulse and slice selective gradient using self-learning machine

A deep reinforcement learning method referred to as DeepRF-Grad, is newly developed to design an RF pulse and a slice selective gradient waveform. The method is demonstrated for slice-selective inversion and compared with SLR and VERSE-designed pulses. The DeepRF-Grad designed pulse showed lower SAR (SLR: 13.2mG²s, VERSE: 6.37mG²s, DeepRF-Grad: 5.00mG²s). When designed for off-resonance robustness, the DeepRF-Grad generated enhanced off-resonance characteristics compared to that of VERSE-designed pulse, while showing similar SAR.