To investigate the use of concentrated (lower duty cycle) high power pulses during saturation preparation to achieve an increase of the inhomogeneous magnetization transfer (ihMT) signal in vivo.ihMT imaging [1-2] shows promise for characterization of myelinated tissues but the signal difference it measures is relatively small, especially at powers achievable in humans. ihMT and MT typically employ RF irradiation spread out over time with duty cycles of 50% or more. In this work, the effect of changing the duty cycle of the RF irradiation on the ihMT signal is explored. This change is compared by taking an initial 50% duty cycle, i.e. distributed case, and, for the same average power over the saturation period, the concentrated case of less frequent, higher B₁ pulses.The experiments required for ihMT data (i.e. saturation at single, and dual offset frequencies) were conducted on 3 healthy volunteers using a GE 3T scanner and 8-channel head coil. RF preparation was achieved with 5ms pulses (cosine modulated for dual offset frequency irradiation) for a saturation duration of 2s, prior to single shot spin echo EPI (FOV=25x25cm²; matrix=128x128; slice=6mm; TE/TR=24ms/5s). For the distributed case, the 5ms pulses were initially applied every 10ms for a 50% duty cycle (Fig.1a). The effect of duty cycle, absolute offset frequency, |Δ| and B_1,RMS (over the saturation duration) was explored. The duty cycle was reduced by increasing the space between pulses whilst maintaining B_1,RMS (Fig.1b). The effect of a reduced duty cycle of 5% on the ihMT ratio (ihMTR) was examined for a range of |Δ| and B_1,RMS values. Also calculated was the MT ratio (MTR) difference between the saturated signals for duty cycles <50% and 50%, such that MTR difference = MTR(50% duty cycle) - MTR(<50% duty cycle).There were 2 to 5 fold increases in the ihMTR using the concentrated versus distributed preparation (i.e. data acquired at a lower duty cycle) from white matter (WM) and grey matter (GM) ROIs (Figs.2,3a-b). The ihMTR spectrum was similar in shape but somewhat changed by a concentrated preparation; ihMTR at higher frequency was amplified more by the concentrated irradiation than at lower |Δ|. Examining individual components of the calculated ihMTR, the MTR was greatly reduced for concentrated single frequency irradiation and substantially less so for the dual frequency concentrated case (Fig.4). Based on the MTR difference, the WM/GM ratio increased from 1.3±0.1 to 1.5±0.1 going from single frequency to dual frequency data, and for lower B_1,RMS with a ratio of 1.8±0.4 at B_1,RMS=15mG (Figs.3c-d).The ability to greatly increase the ihMT signal within safe power levels at 3T enhances the diagnostic and scientific potential of ihMT. Recent work on development of a model for the ihMT signal highlights its sensitivity to the dipolar relaxation time, T_1D parameter [3]. Since dipolar order effects and T_1D only affect single frequency irradiation, the greater reduction of concentrated single frequency irradiation suggests additional T_1D sensitivity. The MTR difference between concentrated and distributed single frequency irradiation might also be considered as an alternative ihMT-like contrast when power constraints are severe, as at 7T. Indeed an example volunteer image of the ihMTR (5% duty cycle; B_1,RMS=30mG) and MTR difference from data acquired at the lowest B_1,RMS shows comparable contrast (Figs.3b,d).Use of shorter duration, high power saturation pulses during preparation at equal B_1,RMS (i.e. lower duty cycles) increased the ihMTR achievable and provided a means to reduce the average SAR. Further investigation is required to see how this effect relates to MT/ihMT related model parameters, such as the free and bound pool exchange rate, and the dipolar relaxation time, T_1D.