Optimal treatment conditions for low-intensity pulsed ultrasound therapy for Alzheimer's disease: applications from mice to humans.

PURPOSE: We previously developed a novel therapy with low-intensity pulsed ultrasound (LIPUS) that ameliorates cognitive decline through upregulation of endothelial nitric oxide synthase (eNOS) in mouse models of Alzheimer's disease (AD). In a randomized, double-blind, placebo-controlled pilot trial, we demonstrated that whole-brain LIPUS therapy is safe and tends to suppress the cognitive decline in early AD patients. We herein report the findings of our basic experiments that we performed for the pilot trial in order to apply whole-brain LIPUS therapy to humans, as well. METHODS: First, we examined the relationship between bone density/thickness and ultrasound transmittance using human temporal bone. Next, based on the results of ultrasound transmittance, we further examined mRNA expression of VEGF, FGF2, and eNOS in response to variable ultrasound frequencies, duty cycles, and sound pressures. RESULTS: There was a significant correlation between bone thickness and transmittance (1.0 MHz, P < 0.001), while there was no significant correlation between bone density and transmittance (1.0 MHz, P = 0.421). At a frequency of 0.5 MHz, the optimum duty cycle was considered to be up to 20%. When the tissue amplitude was in the range of 0.05-0.5 MPa, VEGF, FGF2, and eNOS were significantly upregulated by LIPUS. Thus, the conditions necessary for LIPUS therapy for the human brain were identified as sound pressure just below the probe 1.3 MPa (tissue amplitude 0.15 MPa), duty cycle 5%, and frequency 0.5 MHz. CONCLUSION: We successfully identified the optimal treatment conditions for LIPUS therapy for patients with AD.

Influence of Power-Weighted Center of Echo Signal Within Window Function on Local Strain Rate Distribution in Left Ventricular Wall.

OBJECTIVE: The deviation of the power-weighted center of the echo signal from the geometric center within the velocity estimation window for calculating strain rate (SR) causes an estimation error. This study aimed to confirm whether an erroneous multilayer pattern in the SR distribution of the left ventricular wall could be corrected by considering the power-weighted center of the echo signal. METHODS: The SR distributions were measured locally in the transmural direction around the pre-ejection and early diastolic phases in healthy volunteers. The estimation error related to the power-weighted center of the echo signal was corrected using a previously proposed method, and the effectiveness of the correction was confirmed based on the accuracy of the estimated myocardial displacement. RESULTS: The SR distribution in early diastole was observed as multilayers of low- and high-amplitude negative SRs. However, this multilayer pattern disappeared after correction. In the pre-ejection phase, multilayers of positive and negative SRs were observed in the SR distributions with and without correction. This correction was sufficiently effective in accurately tracking the local peak of the echo signal. CONCLUSION: The multilayer pattern of low- and high-amplitude positive or negative SRs is caused by estimation errors related to the power-weighted center of the echo signal. The multilayer pattern of positive and negative SRs might not be caused by these errors and might relate to the actual change in myocardial thickness because the estimation errors do not convert the negative (positive) SR to positive (negative) in a homogeneous negative (positive) SR distribution.

A novel ultrasonic method for measuring minute sinusoidal displacement by network analyzer.

We developed a method for generating continuous sinusoidal displacements of an object to estimate viscoelastic parameters. However, the amplitude of the displacement caused by the ultrasonic excitation force under safe guidelines was small (a few micrometers), and it was difficult to stably measure the displacement. Therefore, to stably measure the amplitude of sinusoidal displacement as small as the order of micrometers, we proposed a novel method using a network analyzer. Ultrasonic waves were irradiated using an ultrasonic transducer on an object vibrating sinusoidally. The S parameter of the first reflected wave received from the surface of the object was measured using a network analyzer. The S parameter and the inverse Fourier transform were formulated theoretically, and the amplitude of the sinusoidal displacement of the object was estimated from the amplitude characteristics of the inverse Fourier-transformed signal. The proposed method was applied to measure sinusoidal displacements on the order of micrometers from 10 to 300 Hz on an object using a water tank experiment. The obtained sinusoidal displacement agreed well with the reference values measured using a laser displacement meter. The proposed method can accurately measure minute sinusoidal displacements that occur on an object.

Optimizing irradiation conditions for low-intensity pulsed ultrasound to upregulate endothelial nitric oxide synthase.

PURPOSE: Here we aimed to develop a minimally invasive treatment for ischemic heart disease and demonstrate that low-intensity pulsed ultrasound (LIPUS) therapy improves myocardial ischemia by promoting myocardial angiogenesis in a porcine model of chronic myocardial ischemia. Studies to date determined the optimal treatment conditions within the range of settings available with existing ultrasound equipment and did not investigate a wider range of conditions. METHODS: We investigated a broad range of five parameters associated with ultrasound irradiation conditions that promote expression of endothelial nitric oxide synthase (eNOS), a key molecule that promotes angiogenesis in human coronary artery endothelial cells (HCAEC). RESULTS: Suboptimal irradiation conditions included 1-MHz ultrasound frequency, 500-kPa sound pressure, 20-min total irradiation time, 32-48-[Formula: see text] pulse duration, and 320-[Formula: see text] pulse repetition time. Furthermore, a proposed index, [Formula: see text], calculated as the product of power and the total number of irradiation cycles applied to cells using LIPUS, uniformly revealed the experimental eNOS expression associated with the various values of five parameters under different irradiation conditions. CONCLUSION: We determined the suboptimal ultrasound irradiation conditions for promoting eNOS expression in HCAEC.

Estimation error in speed of sound caused by rotation of measured cross-section from short-axis plane of blood vessels: a preliminary study.

PURPOSE: Estimating the speed of sound (SoS) in ultrasound propagation media is important for improving the quality of B-mode images and for quantitative tissue characterization. We have been studying a method for estimating the SoS by measuring the reception time distribution of waves scattered from a scatterer at the elements in a probe. Previously, the measurement cross section was assumed to be perpendicular to the long axis of the blood vessel. In this study, we experimentally investigated the relationship between rotation angle [Formula: see text] of the probe relative to the short-axis plane of the blood vessel and the estimated SoS, [Formula: see text]. METHODS: Water tank and phantom experiments were conducted to investigate the characteristics of [Formula: see text] and element signals when the probe was rotated. RESULTS: The received signal powers at the elements around both edges greatly decreased as [Formula: see text] increased. We introduced a parameter representing the decrease in power, [Formula: see text], in the received signal at the elements at both edges relative to the center element. [Formula: see text] was estimated to be larger as [Formula: see text] increased, especially for [Formula: see text]. [Formula: see text] also increased as [Formula: see text] increased. An approximately proportional relationship existed between the errors in [Formula: see text] and [Formula: see text]. CONCLUSION: Based on these results, we can distinguish between the presence and the absence of SoS misestimations using the difference in power among the elements in the received signal. In the absence of misestimation, we can obtain the true SoS, even if the target has a non-negligible size, by applying our previously proposed methods.

A randomized, double-blind, placebo-controlled pilot trial of low-intensity pulsed ultrasound therapy for refractory angina pectoris.

BACKGROUND: Despite the advances in the treatment of cardiovascular diseases, effective treatment remains to be established to improve the quality of life and prognosis of patients with chronic coronary syndromes. This study was aimed to evaluate the effectiveness and safety of the low-intensity pulsed ultrasound (LIPUS) therapy, which we have developed as a novel non-invasive angiogenic therapy through upregulation of endothelial nitric oxide synthase (eNOS). METHODS AND FINDINGS: We conducted a randomized, double-blind, placebo-controlled (RCT) pilot trial of the LIPUS therapy for patients with refractory angina pectoris. The patients who received optimal medical therapy without indication of PCI or CABG due to the lack of graftability or complexity of coronary lesions were enrolled. They were randomly divided into the LIPUS treatment group (N = 31) and the placebo group (N = 25) in a 1:1 fashion. The LIPUS therapy was performed in a transthoracic manner for 20 min for 3 sections each (mitral, papillary muscle, and apex levels) under the conditions that we identified; frequency 1.875 MHz, intensity 0.25 MPa, and 32 cycles. The primary endpoint was weekly use of nitroglycerin. Secondary endpoints included stress myocardial perfusion imaging and others. The average weekly nitroglycerin use (times/week) was decreased from 5.50 to 2.44 in the LIPUS group and from 5.94 to 2.83 in the placebo group. The changes in the average weekly nitroglycerin use were comparable; -3.06 (95% CI: -4.481 to -1.648) in the LIPUS group (P<0.01) and -3.10 (95% CI: -4.848 to -1.356) in the placebo group (P<0.01). No adverse effects were noted. CONCLUSIONS: In the present study, the LIPUS therapy did not further ameliorate chest pain as compared with optimal medications alone in patients with refractory angina pectoris. The present findings need to be confirmed in another trial with a large number of patients. (Registration ID: UMIN000012369).

Beneficial Effects of Low-Intensity Pulsed Ultrasound Therapy on Right Ventricular Dysfunction in Animal Models.

Right ventricular failure (RVF) is a leading cause of death in patients with pulmonary hypertension; however, effective treatment remains to be developed. We have developed low-intensity pulsed ultrasound therapy for cardiovascular diseases. In this study, we demonstrated that the expression of endothelial nitric oxide synthase (eNOS) in RVF patients was downregulated and that eNOS expression and its downstream pathway were ameliorated through eNOS activation in 2 animal models of RVF. These results indicate that eNOS is an important therapeutic target of RVF, for which low-intensity pulsed ultrasound therapy is a promising therapy for patients with RVF.

Speed-of-sound estimation in ultrasound propagation medium by considering size of target scatterer.

PURPOSE: Accurate speed-of-sound (SoS) estimation in an ultrasound propagation medium improves imaging quality and contributes to better diagnosis of diseases. In conventional time-delay-based SoS estimation approaches studied by several groups, a received wave is assumed to be scattered from an ideal point scatterer. In these approaches, the SoS is overestimated when the target scatterer has a non-negligible size. In this paper, we propose the SoS estimation method that considers target size. METHODS: In the proposed method, the error ratio of the estimated SoS using the conventional time-delay-based approach is determined from measurable parameters using the geometric relationship between the received elements and target. Subsequently, the SoS erroneously estimated using conventional estimation, assuming the ideal point scatterer as a target, is corrected by the determined estimation error ratio. To validate the proposed method, the SoS in water was estimated for several wire sizes. RESULTS: The SoS in the water was overestimated using the conventional SoS estimation method, with a maximum positive error of 38 m/s. The proposed method corrected the SoS estimates, and the errors were suppressed to within 6 m/s, irrespective of the wire diameter. CONCLUSION: The present results demonstrate that the proposed method can estimate the SoS by considering the target size without using information on the true SoS, true target depth, and true target size, which is applicable to in vivo measurements.

Lateral M-Mode: Ultrasound Visualization of Displacement Along Longitudinal Direction at Intima-Media Complex.

Quantification of the dynamics of the carotid artery wall is useful in evaluating arteriosclerosis and atherosclerosis. As the carotid artery wall moves not only in the radial direction but also in the longitudinal direction, longitudinal movement should be considered in the analysis of the dynamic properties of the carotid artery wall. In this study, we propose a "lateral M-mode" method for visualizing the longitudinal movement of the intima-media complex (IMC). For the lateral M-mode, we set the target line in the longitudinal direction along the IMC and visualize the signals on the target line frame-by-frame by correcting the position of the target line along the radial displacement estimated by the phased tracking method. Differentiating the envelope signals between consecutive ultrasound beams was effective in visualizing the lateral movement of the IMC. The precision of the longitudinal displacement of the IMC estimated using the conventional block-matching method was validated by comparing it with the lateral M-mode. Because the conventional M-mode sequence plays an important role in evaluation of the dynamics of various tissues, the proposed "lateral M-mode" contributes to a detailed understanding of vascular dynamics and the development of diagnostic methods for vascular diseases.

Evaluation of two semi-supervised learning methods and their combination for automatic classification of bone marrow cells.

Differential bone marrow (BM) cell counting is an important test for the diagnosis of various hematological diseases. However, it is difficult to accurately classify BM cells due to non-uniformity and the lack of reproducibility of differential counting. Therefore, automatic classification systems have been developed in which deep learning is used. These systems requires large and accurately labeled datasets for training. To overcome this, we used semi-supervised learning (SSL), in which learning proceeds while labeling. We used three methods: self-training (ST), active learning (AL), and a combination of these methods, and attempted to automatically classify 16 types of BM cell images. ST involves data verification, as in AL, before adding them to the training dataset (confirmed self-training: CST). After 25 rounds of CST, AL, and CST + AL, the initial number of training data increased from 425 to 40,518; 3682; and 47,843, respectively. Accuracies for the test data of 50 images for each cell type were 0.944, 0.941, and 0.976, respectively. Data added with CST or AL showed some imbalances between classes, while CST + AL exhibited fewer imbalances. We suggest that CST + AL, when combined with two SSL methods, is efficient in increasing training data for the development of automatic BM cells classification systems.

Application of low-complexity generalized coherence factor to in vivo data.

PURPOSE: Beamforming using the generalized coherence factor (GCF) reduces sidelobe artifacts and provides an excellent contrast-to-noise ratio. We previously proposed GCFreal, a method to calculate GCF without generating analytic signals, and GCFbin, a method to calculate GCF by binarizing the received signals. In this study, we applied these methods to in vivo data and showed the effect of the computational complexity reduction on contrast performance. METHODS: Channel RF data were acquired from the human liver and gallbladder. We set up several observation points in each data set and investigated the mechanism that causes the differences in contrast performance among the methods based on the signals and their power spectra in the channel direction. RESULTS: For GCF and GCFreal, the obtained values were almost the same. However, there were large differences in GCFbin from GCF when the signals from the focus point or from outside the focus point were received on different channels. This is because the amplitudes of the signals with high coherence and those with low coherence were changed by binarizing the signals. CONCLUSION: While GCFbin can significantly reduce the computational complexity, there are differences in the values of GCFbin and GCF due to binarizing of the received signals. However, this difference resulted in GCFbin being superior to GCF in terms of artifact reduction. This is owing to the elimination of amplitude information in GCFbin, which makes it a new efficient coherence factor with different characteristics from GCF.

Evaluation of local changes in radio-frequency signal waveform and brightness caused by vessel dilatation for ascertaining reliability of elasticity estimation inside heterogeneous plaque: a preliminary study.

PURPOSE: To diagnose plaque characteristics, we previously developed an ultrasonic method to estimate the local elastic modulus from the ratio of the pulse pressure to the strain of the arterial wall due to dilatation in systole by transcutaneously measuring the minute thinning in thickness during one cardiac cycle. For plaques, however, some target regions became thicker as the vessel dilates, resulting in false elasticity. Therefore, a method to identify a reliable target for the elastic modulus estimation is indispensable. As a candidate for an identification index of plaques that become thicker during one cardiac cycle, the correlation of the radio-frequency (RF) signals remains high and it is not sufficient to obtain the elasticity. In this study, we thoroughly observed the target with a high correlation but positive strain in the plaque and characterized it by the property of the surrounding area. METHODS: For the plaque formed in the right carotid sinus of a patient with hyperlipidemia and the wall of the right common carotid artery of a young healthy male, (1) the correlation value as the similarity between the RF signals, (2) change in brightness obtained from the log-compressed envelope signals, and (3) strain obtained between the time of the R-wave and that of the maximum vessel dilatation were observed to characterize the region in the plaque. RESULTS: In the plaque, it was found that the region with high correlation and positive strain and its surrounding area could be classified into one of the three typical patterns. CONCLUSION: As a preliminary study, this study provides a clue to assert the reliability of elasticity estimates for a region with high correlation and positive strain in the plaque based on measurable properties.

Appropriate Window Function and Window Length in Multifrequency Velocity Estimator for Rapid Motion and Locality of Layered Myocardium.

The heart wall has a multilayered structure and moves rapidly during ejection and rapid filling periods. Local strain rate (SR) measurements of each myocardial layer can contribute to accurate and sensitive evaluations of myocardial function. However, ultrasound-based velocity estimators using a single-frequency phase difference cannot realize these measurements owing to insufficient maximum detectable velocity, which is limited by a quadrature frequency. We previously proposed a velocity estimator using multifrequency phase differences to improve the maximum detectable velocity. However, the improvement is affected by a spatial discrete Fourier transform (DFT) window length that represents the locality of the velocity estimation. In this article, we theoretically describe that shortening the window increases the interference between different frequency components and decreases the maximum detectable velocity. The tradeoff between the maximum detectable velocity and the window length was confirmed through simulations and a water-tank experiment. Under the tradeoff, the Hanning window, which was used in previous studies, is not always appropriate for the local measurement of the velocity, which sometimes exceeds 100 mm [Formula: see text] depending on the subject, direction of the ultrasound beam to the heart wall, and cardiac periods. In the in vivo measurement with the short window, the Tukey window with a large flat part that has a high-frequency resolution and ameliorates the discontinuity at both edges of the windowed signal was appropriate to measure the maximum velocity. This study offers the potential for local measurements of each myocardial layer using the multifrequency velocity estimator with the appropriate window function and window length.

Strain Rate Distribution in Layered Myocardium Measured Using Local Velocity Estimator with Multifrequency Phase Differences.

Measurement of the myocardial strain rate (SR), with high spatial resolution, is useful in evaluation of the transmurality of myocardial infarction. As the SR distribution is calculated using velocities observed at multiple positions in the heart wall, it is necessary to estimate the local velocity to measure SR distribution. In the present study, our previously proposed local velocity estimator, with multifrequency phase differences, was used to measure SR distribution in the heart wall. The SR distribution measured with the proposed local velocity estimator revealed alternate layers of contraction and relaxation, which were not measured with the conventional velocity estimator with spatial averaging. The reproducibility of the SR distributions was confirmed in three consecutive heartbeats with three subjects. High-spatial-resolution SR measurement with the proposed local velocity estimator will allow myocardial layer-specific analysis in the transmural direction.

Low-complexity generalized coherence factor estimated from binarized signals in ultrasound beamforming.

PURPOSE: In coherence-based beamforming (CBB) using a generalized coherence factor (GCF), unnecessary signals caused by sidelobes are reduced, and an excellent contrast-to-noise ratio (CNR) is achieved in ultrasound imaging. However, the GCF computation is complex compared to the standard delay-and-sum (DAS) beamforming. In the present study, we propose a method that significantly reduces the number of GCF computations. METHODS: In the previously proposed GCFreal, generation of the analytic signal for each element in the conventional GCF could be omitted. Furthermore, in GCF estimated from binarized signals (GCFB) proposed in the present study, the GCF value is calculated after the received signal of each element is binarized to reduce the computational complexity of the GCF. RESULTS: The values of GCFB and GCFreal estimated from simulation and experimental data were compared. We also evaluated the image quality of B-mode images weighted by GCFB and GCFreal. Compared with GCFreal, GCFB was superior in reducing unnecessary signals but tended to reduce the brightness of the diffused scattering media. The CNR improvement was comparable for both methods. CONCLUSION: Generalized coherence factor estimated from binarized signals exhibits excellent CNR improvement compared to DAS. CNR improvements yielded by GCFB and GCFreal may depend on the observation target; however, under the conditions of the present study, comparable performances were obtained. Because GCFB can significantly reduce the computational complexity, it is potentially applicable in clinical diagnostic equipment.

Low-intensity pulsed ultrasound therapy promotes recovery from stroke by enhancing angio-neurogenesis in mice in vivo.

Since the treatment window of thrombolytic therapy for stroke is limited, new therapy remains to be developed. We have recently developed low-intensity pulsed ultrasound (LIPUS) therapy to improve cognitive dysfunction in mouse models of vascular dementia and Alzheimer's disease. Here, we further aimed to examine whether our LIPUS therapy improves neurological recovery from ischemic stroke, and if so, to elucidate the mechanisms involved. In a mouse model of middle cerebral artery occlusion (MCAO), we applied LIPUS (32 cycles, 193 mW/cm2) to the whole brain 3 times in the first week (days 1, 3, and 5) after MCAO. We evaluated neurological functions using behavioral tests and performed histological analyses. Furthermore, to elucidate how LIPUS works within the injured brain, we also tested the effects of LIPUS in endothelial nitric oxide synthase (eNOS)-deficient (eNOS-/-) mice. In wild-type mice, the LIPUS therapy markedly improved neurological functions in the tightrope and rotarod tests at 28 days after MCAO. Histological analyses showed that the LIPUS therapy significantly increased the numbers of CD31-positive blood vessels in the perifocal lesion and doublecortin (DCX)-positive neurons in the ischemic striatum, indicating the angio-neurogenesis effects of the therapy. Importantly, these beneficial effects of the LIPUS therapy were totally absent in eNOS-/- mice. No adverse effects of the LIPUS therapy were noted. These results indicate that the LIPUS therapy improves neurological functions after stroke through enhanced neuro-angiogenesis in mice in vivo in an eNOS-dependent manner, suggesting that it could a novel and non-invasive therapeutic option for stroke.

Multifrequency Phased Tracking Method for Estimating Velocity in Heart Wall.

Local high-accuracy velocity estimation is important for the ultrasound-based evaluation of regional myocardial function. The ultrasound phase difference at the center frequency of the transmitted signal has been conventionally used for velocity estimation. In the conventional method, spatial averaging is necessary owing to the frequency-dependent attenuation and interference of backscattered waves. Here, we propose a method for suppressing these effects using multifrequency phase differences. The resulting improvement in velocity estimation in the heart wall was validated by in vivo experiments. In the conventional method, the velocity waveform exhibits spike-like changes. The velocity waveform estimated using the proposed method did not exhibit such changes. Because the proposed method estimates myocardium velocity without spatial averaging, it can be used for measuring heart wall dynamics involving thickness changes.

Evaluation method of the degree of red blood cell aggregation considering ultrasonic propagation attenuation by analyzing ultrasonic backscattering properties.

PURPOSE: Red blood cell (RBC) aggregation is one of the main factors that determines blood viscosity and an important indicator for evaluating blood properties. As a noninvasive and quantitative method for diagnosing blood properties, our research group estimated the size of RBC aggregates by fitting the scattered power spectrum from the blood vessel lumen with the theoretical scattering characteristics to evaluate the degree of RBC aggregation. However, it was assumed that the propagation attenuation of ultrasound in the vascular lumen was the same regardless of whether RBCs were aggregated or not, which caused systematic errors in the estimated size. METHODS: To improve the size estimation accuracy, we calculated and corrected the attenuation of the blood vessel lumen during RBC aggregation and non-aggregation. The attenuation in the blood vessel lumen was calculated with the spectra acquired from two different depths. RESULTS: In the basic experiments using microparticles, the estimation accuracy decreased as the concentration increased in the case of the conventional method, but the estimated size tended to approach the true size irrespective of the concentration, removing the propagation attenuation component with the proposed method. In the in vivo experiment on the human hand dorsal vein, the size was estimated to be larger during RBC aggregation and smaller during non-aggregation using the proposed method. CONCLUSION: These results suggest that the proposed method can provide precise size estimation by considering the propagation attenuation component regardless of differences in blood conditions such as RBC concentration and degree of aggregation.