Chemical exchange saturation transfer (CEST) MRI can detect low-concentration compounds with exchangeable protons via saturation transfer to water^1,2. One of the CEST contrasts, amide proton transfer (APT), was developed to detect the exchange of amide protons, which resonate at 3.5 ppm downfield from the bulk water protons of endogenous mobile proteins and peptides in tissue. It has been successful applied to many pathological studies such as stroke, tumor grading in clinical patient^3,4. Experimental measurement of the APT effect is complex, and depends on not only amide proton concentration and exchange rate, saturation scheme, but also possibly pulse sequence readout. A variety of imaging sequences have been used for detecting APT effect, including echo planar imaging (EPI), fast spin-echo (FSE), and gradient echo⁵. To date, many studies have been conducted on the design and optimization of saturation scheme. However, little literature is available on the effects of varied readout sequences. In this study, three imaging sequences, spin-echo EPI (seEPI), single-shot FSE (ssFSE), and a recent developed 3D FSE spiral (3Dspiral), were employed for comparing APT effect and imaging quality on tumor patients.MRI acquisition was performed on a 3T clinical scanner (Discovery MR750, GE Healthcare, Milwaukee, USA) using a 32-channel brain coil as the signal detection and whole body coil for RF transmission. APT imaging were acquired with three different pulse sequences (seEPI, ssFSE, 3Dspiral) with identical TR (2.5 s) and saturation scheme. The parameters for saturation scheme were as follows: saturation pulse= 400 ms × 4 fermi pulses, B₁=1.5 μT, 37 saturation frequencies with S_sat(w) (0, ±25, ±50, ±75, ±100, ±150, ±200, ±250, ±300, ±350, ±375, ±400, ±425, ±450, ±475, ±500, ±525, ±550, ±600 Hz), and 2 S₀ (without saturation pulse). The imaging parameters were as follow: FOV= 22 cm; matrix size=128×128; slice thickness=5 mm; and TE=25.3 ms, 30.24 ms, and 10.1 ms for seEPI, ssFSE, and 3Dspiral, respectively. In the data analysis, the APT effect was quantified using the magnetization transfer ratio asymmetry (MTRasym) at 3.5 ppm with respect to the water resonance as the formula of (S_sat(-3.5ppm)-S_sat(+3.5ppm))/S₀ with pixel-by-pixel B₀ correction.Gd-enhancing T1-weighted image (T1WI) and APT images generated by three pulse sequences in brain tumor patient with cavernous hemangioma were shown in Fig. 1. Enhanced signal on post-contrast T1WI is observed (Fig. 1a), implying the tumor location is in suprasellar cistern with extension to the basal ganglion and thalamic region. Hyperintense signal was noted on APT imaging (Fig. 1b-d) within Gd-enhancing tumor area, suggesting increased content of protein and peptides. APT images generated by 2D seEPI (Fig. 1b) and ssFSE (Fig. 1c) appear to be similar, while APT image showed the blurring in 3Dspiral (Fig. 1d). For acquiring APT data on whole brain with reasonable acquisition time (6 min with 39 saturation frequencies in this case), the spiral arm was reduced for 3Dspiral acquisition and consequently, low resolution APT image was found at 3Dspiral. Z-spectrum and MTR_asym within tumor area and gray matter were shown in Fig. 2. Z-spectrum was found to be slightly broader at seEPI (Fig.2). The possible reason is that seEPI could involve longer readout and means the chemical exchange duration is prolonged. More exchangeable saturated protons, including MT pool and amide proton, could exchange with bulk water during the whole EPI readout, thus slightly broadening the Z-spectrum. APT quantitative values among the sequences within tumor area (Fig. 2a) are slightly varied (4.6%, 5.2%, and 3.7% for ssFSE, seEPI, and 3Dspiral, respectively) but more consistent at gray matter region (Fig. 2b).We have demonstrated that varied readout sequences with the same saturation scheme showed the similar APT and Z-spectrum. Careful consideration may be required when choosing the appropriate CEST readout sequences for your own applications⁵, though it appears like among various sequences readout. For example, although FSE-based CEST sequence produces exquisite images with less sensitive to field inhomogeneity and may be suitable to neck region for detecting nasopharyngeal carcinoma in comparison with EPI, the additional refocusing pulses can create concerns of energy deposit on clinical scanners. When considering doing whole brain study with voxel-based analysis, 3Dspiral would be good candidate since limited slice number are acquired at 2D sequence due to possible saturation inference on neighboring slices.