Effect of Saturation Pulse Duration and Power on pH-weighted Amide Proton Transfer Imaging: A Phantom Study

Abstract

Purpose: Amide proton transfer (APT) imaging may detect changes in tissues’ pH based on the chemical exchange saturation transfer (CEST) phenomenon, and thus it may be useful for identifying the penumbra in ischemic stroke patients. We investigated the effect of saturation pulse duration and power on the APT effect in phantoms with different pH values.

Methods: Five samples were prepared from a 1:10 solution of egg-white albumin in phosphate-buffered saline at pH 6.53–7.65. The APT signal intensity (SI) was defined as asymmetry of the magnetization transfer ratio at 3.5 ppm. We measured the APT SIs in the egg-white albumin samples of different pH values with saturation pulse durations of 0.5, 1.0, 2.0, and 3.0 sec and saturation pulse powers of 0.5, 1.5, and 2.5 μT. The relative change in the APT SI in relation to the saturation duration and power at different pH values was defined as follows: (APT SI _{each saturation pulse} − APT SI _{shortest or weakest pulse})/APT SI_{shortest or weakest pulse}. The dependence of the APT SI on pH and the relative change in the APT SI were calculated as the slope of the linear regression.

Results: The lower the pH, the larger the relative change in the APT SI, due to the change in saturation pulse duration and power. The APT SI was highly correlated with the pH at all saturation pulse durations and powers.

Conclusion: The influence of saturation duration and power on the APT effect was greater at lower pH than higher pH. The combination of saturation pulse ≥ 1.0 s and power ≥ 1.5 μT was useful for the sensitive detection of changes in APT effects in the egg-white albumin samples with different pH values.

Introduction

Amide proton transfer (APT) imaging is a specific type of endogenous chemical exchange saturation transfer (CEST) imaging technique introduced by Zhou et al.^1,2 The CEST method is capable of detecting various low concentrations of compounds in vivo without contrast injection, and the APT imaging detects the exchange of amide protons (-NH), which resonate at + 3.5 ppm downfield from the bulk water protons of endogenous mobile proteins and peptides in tissue.² The contrast of APT imaging reflects both the concentration of between amide protons and the protons of bulk water and the proton exchange rate.

Concentration-weighted APT imaging is considered to reflect different tissue concentrations of amide protons contained in mobile proteins and peptides, which are significantly enhanced in active tumors.³ It has been demonstrated that concentration-weighted APT imaging is useful for the differentiation of radiation necrosis from active or recurrent glioma,³ for predicting the histopathological grades of diffuse gliomas in patients⁴ and for the evaluation of the therapeutic response to chemotherapy in glioblastoma multiforme.⁵ On the other hand, pH-weighted APT imaging may be useful for the identification of both the viable ischemic penumbra^6–9 and the ischemic core with irreversibly damaged tissue in the early stage.^10,11 The ischemic penumbra may exhibit increased oxygen extraction and anaerobic glycolysis in an effort to maintain transmembrane ion gradients.¹² Anaerobic metabolism leads to tissue acidosis, and thus acidosis may provide a sensitive indicator of impairment.¹³

The contrasts in APT imaging are dominated by the asymmetry analysis of the magnetization transfer at 3.5 ppm, which corresponds to the CEST effect in this technique. However, the APT signal intensity (SI) can be influenced not only by the concentration and pH of endogenous mobile proteins and peptides but also by scan parameters, such as the power and duration of the RF saturation pulse. In previous studies investigating scan parameters, Wada et al. demonstrated that APT imaging with long saturation pulses can sensitively reflect the concentrations of mobile proteins and peptides by using different concentrations of egg-white phantom.⁹ Togao et al. reported the utility of long RF saturation pulses, which were enabled by a parallel transmission-based technique to sensitively detect low concentrations of amide protons, in APT imaging of diffuse gliomas.¹⁴ However, these studies were concentration-weighted APT imaging studies, not pH-weighted APT imaging studies. Zhao et al. reported that a saturation pulse power of 2 μT can make the interpretation and differentiation of brain lesions (brain tumors and brain stroke) straightforward and should be useful for clinical applications of APT imaging in the brain.¹⁵ However, the combined influence of the duration and the power of the saturation pulses on pH-weighted APT imaging have not been investigated. Therefore, the purpose of our study was to investigate the effect of saturation pulse power and duration on the APT effect in pH-phantoms and to determine optimal settings for the saturation pulses in pH-weighted APT imaging.

Materials and Methods

Phantoms

Five samples were prepared from egg-white albumin (Wako Pure Chemical Industries, Tokyo, Japan) at a concentration of 10% and agarose at a concentration of 3% diluted with phosphate buffer solution (Wako Pure Chemical Industries) at pH values of 6.53, 6.84, 6.91, 7.25, and 7.65. The samples were held stationary in a bath of warm polyvinyl alcohol (PVA) gel (37℃). The solution pH was measured with a standard pH electrode (Benchtop pH/Water Quality Analyzer LAQUA F-71; HORIBA, Tokyo, Japan). Physiologic total protein content is ~10% by weight in the human brain,^1,4–6 and the human brain is normally regulated at pH 7.0–7.2 through the cooperative action of a multitude of sensors and regulators within cellular compartments.^6,7

MR Imaging

Experiments were performed on a 3.0 T clinical MR scanner (Ingenia 3.0 T; Philips Medical Systems, Best, the Netherlands) using a 15-channel head coil for signal reception and two-channel parallel transmission via the body coil for RF transmission. The acquisition software was modified so that it alternated the operation of the two transmission channels during the RF saturation pulse, which enables long pseudo-continuous-wave RF saturation. Thus, each amplifier can be operated at full power while staying within the specified duty cycle of 50%, allowing for homogenized RF saturation pulses by adjustment of the individual power of the RF amplifiers during alternation, according to the results of the RF calibration scan (B1 calibration).¹⁶

APT images were acquired by the use of a single-shot turbo-spin-echo (TSE) sequence with the following parameters: TR = 12000 msec; TE = 6.7 msec; FOV = 250 × 250 mm²; slice thickness = 5 mm; number of slices = 1; matrix size = 332 × 332; spatial resolution = 0.75 × 0.75 mm²; number of acquisitions = 1; scan time = 6 min 36 sec; sinc-gaussian-shaped saturation pulse (50 msec for one element); saturation pulse power = 0.5, 1.5, and 2.5 µT; and saturation pulse duration = 0.5, 1.0, 2.0, and 3.0 sec. CEST images were acquired at 32 frequency offsets ± 16, ± 12, ± 8 and from −6 to + 6 ppm with a step of 0.5 ppm, as well as one far off-resonant frequency (−1560 ppm) for signal normalization.

We created a B0 map image for point-by-point off-resonance correction from a separately acquired gradient recalled echo sequence with the following parameters: dual echo (ΔTE = 1 msec), TR = 21.1 msec; TE = 11.0 msec; flip angle = 30°; FOV = 250 × 250 mm²; slice thickness = 5 mm; number of slices = 1; matrix size = 0.75 × 0.75 mm²; number of acquisitions = 8; and scan time = 1 min 52 sec.

The phantoms were scanned three times, and the total acquisition time for each one session was 41 min 28 sec.

Image Analysis

All image data were analyzed using the software program ImageJ (version 1.47v; National Institutes of Health, Bethesda, MD, USA). A dedicated plug-in was created for assessment of the z-spectrum and magnetization transfer ratio asymmetry (MTR_asym) equipped with a correction function for B0 inhomogeneity. The z-spectrum and MTR_asym are defined as follows:

$$\mathrm{Z}-\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{u}\mathrm{m}=\frac{S_{sat}\left(\mathrm{\Delta }ppm\right)}{S_0}$$ (1)

$$\mathrm{M}\mathrm{T}{\mathrm{R}}_{asym}\left(\text{Δ}ppm\right)=\frac{S_{sat}\left(-\mathrm{\Delta }ppm\right)-S_{sat}\left(+\mathrm{\Delta }ppm\right)}{S_0}$$ (2)

where $S_{sat}\left(\mathrm{\Delta }ppm\right)$ and $S_0$ are the signal intensities at a target frequency and −1560 ppm, respectively.

To reduce unnecessary contributions from the conventional MT effect and the direct saturation of bulk water, we performed an asymmetry analysis of MTR with respect to the water frequency. In APT imaging, the asymmetry of MTR at 3.5 ppm (MTR_asym [3.5 ppm]) downfield from the bulk water signal is calculated as follows:

$$\mathrm{M}\mathrm{T}{\mathrm{R}}_{asym}\left(3.5ppm\right)=\frac{S_{sat}\left(-3.5ppm\right)-S_{sat}\left(+3.5ppm\right)}{S_0}$$ (3)

In this study, we defined the APT SI as MTR_asym (3.5 ppm) × 100 (%). We measured the APT SIs in egg-white albumin samples with different pH values at the saturation pulse durations of 0.5, 1.0, 2.0, and 3.0 sec (saturation pulse power = 2.5 µT) and power of 0.5, 1.5, and 2.5 µT (duration = 3.0 sec). The phantoms were scanned three times with repositioning and rewarming to about 37℃, and the average values and standard deviations of the APT SIs were calculated. Circular ROIs (approx. 1.5 cm², 18 pixels) were placed in the center of each phantom, and the ROI location was carefully adjusted to avoid including regions with large B0 shifts (> 15 Hz) by referring to the B0 map. The z-spectrum and MTR_asym were calculated pixel-by-pixel, and then the mean values within the ROI were obtained. The APT SIs were calculated pixel-by-pixel; the mean values derived within the ROI were then determined, and the mean values were further averaged using the data of three scans.

The relative change in the APT SI in relation to the saturation pulse duration and power at different pH values were defined as follows:

$$\begin{array}{rl}& Relativechange=\\ & \frac{APTS{I}_{eachsaturationpulse}-APTS{I}_{shortestorweakestpulse})}{APTS{I}_{shortestorweakestpulse}}\end{array}$$ (4)

Statistical analysis

All statistical data analyses were performed with JMP 11 (SAS Institute, Cary, NC, USA). The dependence of the APT SI to pH and the relative change in the APT SI were evaluated as the slope of the linear regression. We compared the APT SIs among the different settings of saturation pulses by one-way analysis of variance (ANOVA) and Tukey’s honestly significant difference (HSD) post hoc test. P values < 0.05 were considered to indicate significant differences.

Results

Figure 1 shows the comparison between the saturation pulse duration and the APT SI at each pH. The APT SI increased with increasing saturation pulse duration at each pH. The APT SIs were significantly higher at the saturation duration of 1.0 sec than that of 0.5 sec at pH 6.53, 6.84, and 7.65. The APT SI was significantly higher at the saturation duration of 2.0 sec than that of 1.0 sec only at pH 6.53. At all pH values, there were no significant differences between the APT SIs at saturation pulses of 2.0 sec and those at saturation pulses of 3.0 sec.

Fig. 1.

Comparison between the saturation pulse duration and the APT SI at each pH with the saturation pulse power 2.5 µT. APT, amide proton transfer; SI, signal intensity.

Figure 2 shows the z-spectrum and MTR_asym of the different saturation pulse durations at each pH. The z-spectrums became broader with increasing saturation pulse duration at all pH values, and the MTR around 3.5 ppm became larger.

Fig. 2.

Z-spectrum (a–e) and MTR_asym (f–j) for four saturation pulse durations measured at five different pH values with a saturation pulse power of 2.5 µT: (a and f) 6.53, (b and g) 6.84, (c and h) 6.91, (d and i) 7.25, (e and j) 7.65. MTR_asym, magnetization transfer ratio asymmetry.

Figure 3 shows the comparison between the saturation pulse power and the APT SI at each pH. The APT SIs at saturation powers of 1.5 µT or 2.5 μT were significantly higher than the APT SIs at a saturation power of 0.5 µT at all pHs. The APT SIs were not significantly different between the saturation pulse powers of 1.5 µT and 2.5 μT at all pHs.

Fig. 3.

Comparison between the saturation pulse power and the APT SI at each pH with the saturation pulse duration 3.0 sec. APT, amide proton transfer; SI, signal intensity.

Figure 4 shows the z-spectrum and MTR_asym of the different saturation pulse powers at each pH. The z-spectrums became broader with increasing saturation pulse power at all pH values, and the MTR around 3.5 ppm became larger.

Fig. 4.

Z-spectrum (a–e) and MTR_asym (f–j) for three saturation pulse powers measured at five different pH values with a saturation pulse duration of 3.0 sec: (a and f) 6.53, (b and g) 6.84, (c and h) 6.91, (d and i) 7.25, (e and j) 7.65. MTR_asym, magnetization transfer ratio asymmetry.

Figure 5 shows the relative change in the APT SI to the saturation pulse duration (saturation power = 2.5 μT) and the power (saturation pulse duration = 3.0 sec) at different pH values. At the lower pH values, the larger relative change in APT SI were observed due to the change in saturation pulse duration and power, except for the pH values of 6.84 and 6.91 at the same saturation pulse duration and power (Table 1).

Fig. 5.

The relative change in the APT SI to the saturation pulse duration (a) and power (b) at different pH values. APT, amide proton transfer; SI, signal intensity.

Table 1.

Coefficient of determination and slope of linear regression between the relative change in APT SI and saturation pulse duration and power at each pH

pH	Change in saturation pulse
	Duration (sec)		Power $\left(\mu T\right)$
	R2	Slope	R2	Slope
6.53	0.938	0.62	0.995	77.08
6.84	0.930	0.53	0.984	8.46
6.91	0.921	0.56	0.981	15.32
7.25	0.886	0.48	0.960	5.88
7.65	0.816	0.35	0.943	4.92

APT, amide proton transfer; SI, signal intensity.

Figure 6 shows the relationship between the pH and the APT SI at each saturation pulse duration and power. The slopes of the linear regression were 3.84, 5.61, 5.48, and 4.76 at saturation durations 0.5, 1.0, 2.0, and 3.0 sec, respectively. The slopes of the linear regression were 1.41, 5.57, and 4.76 at saturation powers of 0.5, 1.5, and 2.5 µT, respectively. The APT SI was highly correlated with the pH at all saturation pulse durations and powers (R² = 0.929–0.968).

Fig. 6.

The relationship between the pH and the APT SI at each saturation pulse duration and saturation pulse power: (a) each saturation pulse duration with a saturation pulse power of 2.5 µT and (b) each saturation pulse power with a saturation pulse duration of 3.0s. APT, amide proton transfer; SI, signal intensity.

The z-spectrum and MTR_asym for each pH at a saturation pulse duration of 3.0 sec and a power of 2.5 µT are shown in Fig. 7. The z-spectrum became broader with increasing pH, and further decreased with increasing pH at around + 3.5 ppm. The MTR_asym increased with increasing pH at around 3.5 ppm. These results exhibited the same trend as the results for the other saturation pulse parameters.

Fig. 7.

The dependence of z-spectrum (a) and MTR_asym (b) with pH at a saturation pulse duration of 3.0s and saturation pulse power of 2.5 µT. MTR_asym, magnetization transfer ratio asymmetry.

Figure 8 shows APT images of egg-white phantoms with five different pH values, obtained with the four saturation pulse durations and three powers. Each APT image in the figure has the same window width and adjusted window level. The APT SI decreased with each of the saturation pulse duration, power, and lower pH. Note that the contrast was seemed to be similar for all parameters except 0.5 µT when the window width was fixed and the window level was adjusted.

Fig. 8.

pH-weighted images obtained under different combinations of saturation pulse duration and power: (a) 0.5s and 2.5 µT, (b) 1.0s and 2.5 µT, (c) 2.0s and 2.5 µT, (d) 3.0s and 2.5 µT, (e) 3.0s and 0.5 µT, (f) 3.0s and 1.5 µT, (g) 3.0s and 2.5 µT, and (h) The T1-weighted image of phantoms. Each APT image is shown with the same window width and adjusted window level. APT, amide proton transfer.

Discussion

We have demonstrated the effect of the saturation pulse duration and power on the pH-weighted APT imaging on a 3 T clinical MR scanner. The APT SIs were smaller at lower pH due to a decreased exchange rate. The APT SIs were positively correlated with the pH value at all saturation pulse durations and powers. This indicates that the APT-weighted images can detect decreased signal intensity in regions with decreased tissue pH in vivo. However, the CEST phenomenon is dependent not only on the concentration and pH of the compound but also on the experimental parameters. Thus, the APT SIs were changed by the saturation pulse duration and power in present study.

The selection of the appropriate experimental parameters is critical in APT imaging based on the CEST effect. The APT SIs were higher at longer durations and stronger powers for the saturation pulses, but a few significant differences were observed among the saturation pulses of ≥ 1.0 sec or saturation powers of ≥ 1.5 μT. Basically, the CEST effect should be larger when a stronger and longer saturation pulse is irradiated because it builds up and maintains the saturation of exchangeable protons, and then it enables a further accumulation of exchanged protons in bulk water.^9,17 Although the APT SI was highly correlated with the pH at all saturation pulse durations and powers, the slopes of the linear regression between APT SI and pH were almost the same in the saturation pulses ≥ 1.0 sec or at the saturation powers ≥ 1.5 μT. The APT ratio (APTR) is magnetization transfer-based and depends on many parameters:

$$APTR=APT{R}_{\mathrm{\infty }}\left(1-e^{-R_{1w}{t}_{sat}}\right)$$ (5)

Here, $R_{1w}$ is the spin-lattice rate of water and $t_{sat}$ is the saturation pulse duration$.APT{R}_{\mathrm{\infty }}$ is the APTR value in the allowed maximum transfer status when t_sat $\to \mathrm{\infty }.$^1,9 The exponential relationship might be applicable to the relative change in the APT SI in this work, but the scaling of the image display is linear. Therefore, linear analysis was applied to align the visual information in this work. In addition, the relative change in APT SI was larger at lower pH than at higher pH. Since the concentration of albumin was the same in all phantoms in this study, it is assumed that the proton exchange was saturated at higher pH due to the fast exchange rate. On the other hand, the increase in APT SI was greater due to the increase in the proton exchange time during the longer saturation pulse duration at lower pH with the slow exchange rate. This indicates that longer saturation pulse duration and stronger power are not necessary for the pH-weighted APT imaging.

The z-spectrum became broader when we used longer saturation pulse duration and stronger saturation pulse power in all pH phantoms because the MT effect and direct water saturation were frequently occurred; moreover, the MTR_asym at around 3.5 ppm became larger increased with saturation pulse duration and power due to enhancement of the CEST effect. The resonance frequency of exchangeable protons in mobile proteins and peptides is not limited to 3.5 ppm but exhibits various frequencies ranging from 1 to 5 ppm, resulting in a wide frequency range of MTR_asym value. It was also possible that the peak shifted due to changes in the exchange rate of protons. A decrease in MTR_aym at 3.5 ppm with a decrease in pH might have resulted in the higher peak at around 2 ppm. A similar trend was found in previous in vivo studies,^2,13,15 although it is difficult to make a direct comparison with the present in vitro results. As the saturation pulse duration decreases, the z-spectrum becomes narrower and more susceptible to B0 inhomogeneity. In images obtained at these short saturation durations, the effect of B0 inhomogeneity that could not be corrected by the B0 map seems to appear as a peak around 0.5 ppm. At the saturation pulse duration of 0.5 sec, it was likely that the free water signal was not sufficiently suppressed and deviated from zero due to insufficiency of the saturation effect. However, this is not a problem because the APT signal intensity is defined as the difference in signal intensities measured at − 3.5 ppm and + 3.5 ppm. Glycosaminoglycan (GAG) CEST (gagCEST) imaging, which detects GAGs in the cartilage of the knee joint and the nucleus pulposus of the intervertebral disc using CEST methods, is highly sensitive to the direct water saturation effect, since the resonance frequency of GAG hydroxyl protons (around + 1.0 ppm from bulk water) is critically close to that of bulk water protons.¹⁷ In gagCEST imaging, the SI values were compared at the saturation pulses of 0.5, 1.0, and 2.0 sec and saturation powers of 0.6, 0.8, and 1.6 μT, and the highest SIs were obtained at 1.0 sec and 0.8 μT. In addition, the MTR_asym at around 1.0 ppm was larger at the saturation pulse and power of 1.0 sec and 0.8 μT, respectively. The reason for not obtaining the largest SI was that there was direct water suppression at the longest and strongest saturation pulses. On the other hand, the resonance frequencies of amide protons (around + 3.5 ppm) are less affected by the direct water saturation effect and do not get smaller MTR_asym than that by gagCEST imaging because they are not so close as that of GAG hydroxyl protons.

In previous clinical studies, Zhao et al. have demonstrated that an RF irradiation power of 2 µT allows for effective interpretation and differentiation of the tumor and stroke types of brain lesions and should be useful for clinical applications of APT imaging in the brain.¹⁵ Togao et al. found that the APT-weighted contrast was enhanced with the use of a longer saturation pulse at 2.0 μT in diffuse gliomas.¹⁴ Wada et al. demonstrated that APT imaging with long saturation pulses can sensitively reflect the concentrations of mobile proteins and peptides by using an egg-white albumin phantom with a saturation pulse of 2.0 μT.⁹ The utility of a long and strong saturation pulse should be detected using high APT signal contrast for the concentration of albumin. The reason for setting the power of saturation pulse to 2.0 µT is that the MTR_asym at 3.5 ppm of normal white matter (NAWM) was around zero (0.18% ± 0.23%). The MTR_asym at 3.5 ppm of both the brain tumor and stroke region increased with increasing saturation pulse power from 1.0, 2.0, to 3.0 µT.¹³ On the other hand, the difference of MTR_asym at 3.5 ppm was only significant at 1.0 µT in stroke patients (almost zero), and no significant difference was observed between the saturation powers of 2.0 and 3.0 µT.¹³ In addition, the APT imaging with a saturation pulse power set at 1.5 µT in patients with cerebral infarction, the 50th percentile of the APT SI in the infarction area (− 0.53), was significantly lower than in the NAWM (− 0.24).¹¹ These findings indicate that the contrast between the infarct area and NAWM can be obtained even if the MTR_asym at 3.5 ppm of NAWM is not zero. The use of a saturation pulse power set at 2.0 µT is necessary for grading gliomas in APT imaging because of the reported cutoff value,⁴ but the contrast is important for pH-weighted APT images in patients with cerebral infarction because no cutoff value has been reported. In the present study, no significant difference was observed in the correlation between pH and APT SI with saturation pulses ≥ 1.0 sec or saturation pulse powers ≥ 1.5 µT. Moreover, the contrast of APT imaging was rendered comparable in all scan parameters by adjusting the window level and window width, except in the case of the weakest saturation pulse power of 0.5 µT. Thus, it may not be necessary to use saturation pulses of longer than 1.0 sec or stronger than 1.5 μT to detect pH changes. Moreover, the decrease in saturation pulse duration enables us to reduce the TR while maintaining the same recovery time, thereby shortening the examination time. The use of shorter saturation pulse duration is effective for patients with cerebral infarction due to the reduction in motion effects. Recently, Zhang et al. proposed a quasi-steady-state (QUASS)-CEST algorithm.¹⁸ Then, by performing both simulations and in vivo experiments in patients with brain tumors, they estimated the spin-lock relaxation rate (R1ρ) from the measured Z-spectrum and calculated the steady-state CEST effect. In this way, they were able to demonstrate that the proposed algorithm minimizes the effect of the relaxation delay and saturation duration time on CEST measurements.^18,19 In the future, it would be desirable to develop a method such as QUASS-CEST that could be performed without dependence on saturation pulse parameters.

Our study has the following limitations. We evaluated only seven pairs of duration and power of saturation pulses; this was because of restrictions in the image acquisition time required to keep the phantom at body temperature. As noted above, the APT SI can be affected by factors, such as the tissue water content, pH, temperature, the T1 of water, and the background MT effect.² In this study, the temperature of the phantom before the start of each scan session was 40.7 ± 0.2℃, whereas that after the scan session was 33.6 ± 1.6℃. Any further decrease in the temperature of the phantom would be a significant deviation from the human body’s temperature, and the scan sequence protocol was thus restricted. The combination of the imaging condition for saturation pulse in this experiment was selected from the minimum saturation pulse duration of 0.5 sec to a maximum of 3 sec, and minimum saturation power of 0.5 μT to a maximum of 2.5 μT that are often used in clinical CEST imaging. Realistically, it was not possible to try every combination of saturation duration and power. Although the combination of imaging conditions was limited in this work, the results of the present study may make it possible to predict the results for the saturation settings that were not tested in this study.

In addition, we were only able to investigate the effect of the saturation pulse on the pH-weighted APT imaging by using different pH phantoms due to ethical considerations attendant on the use of clinical cases, as thrombolytic therapy must not be delayed. In the future work, the development of a shorter APT imaging sequence would be helpful for use in the clinical diagnosis of acute brain infarction.²⁰

Conclusion

The APT SI was increased with increasing saturation pulse power and longer saturation pulse duration. The APT SIs were positively correlated with the pH at all saturation pulse durations and powers. The saturation pulse duration and power combination of ≥ 1.0 s and ≥ 1.5 μT were useful for the sensitive detection of changes in APT effects in egg-white albumin samples with different pH values.

Funding Statement

This work was supported by a KAKENHI grant (no. GAG9K17144).