High intensity focused ultrasound (HIFU) is one of the promising minimally invasive therapeutic methods. Focused ultrasound creates coagulation area with minimally invasiveness. Reduction of thermal damage in non-target area during HIFU therapy as well as effective coagulation of the target area are important. In developing a safe and effective HIFU system and control algorithm for HIFU beam manipulation, experimental evaluation of 3D temperature distribution is required to verify the performance of the system under development. In previous studies, several methods for evaluating 3D temperature distribution were introduced, such as MRI thermometry. This method requires consideration of MRI compatibility of the HIFU system. Other methods including thermocouples and thermochromic liquid crystal (TLC) sheet have been used to visualize temperature distribution in experimental acoustic phantoms. However, there were several limitations in evaluating 3D temperature distribution during HIFU exposure. In this study, a 3D temperature distribution measurement system using micro-capsulated thermochromic liquid crystal (MTLC) was developed. We fabricated an optically transparent temperature sensitive phantom containing MTLC that emits a reflectance spectrum depending on temperature. The 3D temperature distribution was visualized using a light sheet method. The temperature distribution in the optical phantom during HIFU exposure was determined with errors as low as 0.6℃. Using this system, temperature distribution induced by HIFU exposure was visualized using different focusing methods to evaluate their performance.
The safety of physically disabled persons must be considered during the initial developmental stages of a rehabilitative or assistive device to prevent death or injury. Thus, in the research field of biomedical engineering, researchers specializing in physical assistive system development must take the initiative to ensure user safety. This paper proposes methods to plan and implement safety measures for physical assistive systems, using a wheelchair-compatible system with sit-to-stand (STS) movement support as an example. To promote daily independent active exercise and motor learning with progressively less assistance from specialists, the support system does not require the attachment of any device or sensor onto the user's body. To ensure safety, we first identified the possible dangers of injury or other potential hazards involved in STS movement support. Next, we developed safety measures to prevent all the identified dangers of injury and hazards. The steps taken to develop these safety measures were submitted for ethical review. Finally, we confirmed the effectiveness of the safety measures developed by conducting fundamental and realistic experiments. The safety measures described in this paper were developed for the STS support system, but the method used to identify the required safety measures can be implemented in the development of any physical assistive system. The proposed method will help engineers to improve the safety of rehabilitative and assistive devices.
The liver plays a central role in all metabolic processes in our body. Currently, orthotropic liver transplantation is the only proven effective treatment for serious acute liver failure (ALF) such as fulminant hepatic failure (FHF). However, liver transplantation has not yet become a standard treatment for FHF because of a shortage of donors. A bioartificial liver (BAL) device containing living hepatocytes might be a therapeutic alternative to liver transplantation. In this study, we developed a hollow fiber (HF)-type bioartificial liver (BAL) module and evaluated the performance of an embryonic stem cell (ESC)-immobilized bioartificial liver (ES-BAL) module in vitro and ex vivo. Mouse ESCs were immobilized in a BAL module and then cultured under perfusion conditions. Cells immobilized inside the HFs had a high proliferative activity and formed cylindrical organoids. For hepatic differentiation of cultured cells, differentiation-promoting agents were added to the culture medium. Ammonia removal and albumin secretion were detected after about 2 weeks of culture. Male Wistar rats were used in animal experiments. After induction of liver failure, rats were connected to an ES-BAL module containing differentiated ESCs or a control module (without cells) and underwent extracorporeal circulation for 1 h. We measured the changes in blood biochemical parameters. All rats in the control group died within 10 h, whereas two out of three rats treated with the ES-BAL module recovered after the operation. We evaluated the changes in blood biochemical parameters and the liver weight of surviving rats treated with the ES-BAL module. Most of the blood biochemical parameters had returned to normal ranges at 1 week after treatment. We also observed increased liver weight. In conclusion, we developed a novel HF-type BAL module containing differentiated ESCs with liver-specific functions. Our BAL module has the potential to support liver functions and induce liver regeneration in rats with liver failure.
Electrocardiograms (ECGs) captured by wearable ECG devices readily contain artifacts due to measurement faults. Since artifacts and R waves have quite similar frequency characteristics, R wave misdetection or R-R interval (RRI) miscalculation may result. Aiming at accurate analysis of heart rate variability (HRV), this paper proposes a new RRI outlier processing method consisting of three steps: evaluating RRI reliability, excluding RRI outlier, and complementing missing RRI. In the first step, the method evaluates the measurement status of all detected R waves and calculates RRI reliability based on the measurement status of a combination of the measurement status of two R waves. Since we target wearable ECG devices used in non-medical environment, the method evaluates R waves based on the threshold electric potential for left ventricular hypertrophy, and determines those exceeding the threshold as artifacts. The method accordingly sets lower reliability to RRIs containing R waves evaluated as artifacts. In the second step, the method excludes all RRIs with low reliability as outliers. These steps may be effective for HRV measures in the time domain, but are not sufficient for analyzing HRV measures in the frequency domain. Resampling the time series RRI data, which is essential for analyzing HRV in the frequency domain, may produce outliers if the target RRIs contain missing values. Our method accordingly complements missing RRIs before data resampling based on RRI characteristics. We postulate that consecutive changes in RRIs follow a simple formula consisting of three components: direct current, low frequency, and high frequency. Our method complements missing values according to the formula, which is calculated from RRIs time series regarded as having been properly measured. To confirm the effectiveness of the method before applying it to ECGs recorded by wearable devices, we evaluated all the steps using pseudo-ECGs generated artificially by adding noise and artifacts to open ECG data. Initial evaluation results showed that the proposed method outperformed conventional method regarding the precision of both time and frequency domain measures of HRV.
As the pathological characteristics of living tissue generally correlate with tissue stiffness, techniques such as ultrasound elastography play an important role in the medical field. One issue with elastography is the limited area of measurement. To determine the spatial distribution of elastic moduli, model-based estimation methods have been developed. However, these methods can only be applied to the observable area, and expanding the estimation area remains problematic. This paper introduces a method to estimate the spatial distribution of elastic moduli over the entire elastic body from locally observed deformation patterns. The reconstruction of elasticity from multiple deformation patterns is formulated as a minimization problem based on the sparseness of the gradient of tissue elasticity. Based on simulation experiments investigating the performance of the proposed method, we confirm that spatial variations in elasticity are effectively estimated and the area of elasticity reconstruction is extended.
Simultaneous measurements of mechanomyogram (MMG) and electromyogram (EMG) may be useful for accurate evaluation of skeletal muscle contraction. However, unlike the EMG, the MMG is rarely used in clinical tests. As the target muscle has to be fixed during conventional MMG measurements, it is not possible to measure MMG during dynamic exercises such as sports and rehabilitation. To solve these problems, the authors developed an MMG (displacement-MMG)/EMG hybrid transducer system that allows simultaneous MMG and EMG measurements. Furthermore, we also developed an analysis method that is able to evaluate muscle contraction using the power spectra of each signal. The measurement system and the analysis method were applied to recumbent bicycle pedaling, during which work rate was increased incrementally. The results showed that this transducer system provided MMG/EMG measurements stably during dynamic exercise. When the work rate of the bicycle pedaling increased; that is, when the dynamic muscle strength increased, the sum of the power spectra of the MMG/EMG also increased. The MMG/EMG hybrid transducer system and the analysis method were useful for evaluating muscle strength during dynamic exercise.
Gait characteristics vary among people and correspond to individual differences of external and internal traits. Recent studies on autism spectrum disorders (ASD) reported that gait characteristics are associated with a walker's autistic trait. Previous studies measured gait characteristics with walking alone and did not investigate gait characteristics in interactive situations. The goal of this study was to examine the correlation between ASD traits and gait characteristics in typically developed (TD) young adults. The subjects completed a Subthreshold Autism Trait Questionnaire (SATQ) for quantitative measurement of autistic traits. After completing the questionnaire, the subjects participated in walking experiments in pairs using an inertial measurement unit (IMU)-type three-dimensional motion capture system. Two participants walked toward each other and had to avoid collision with their counterparts. The norm in the X- and Y-axis directions of acceleration, and the norm in the X-, Y-, and Z-axis directions of the angular velocity of four body parts (waist, left/right foot, and head) were calculated. In this study, the standard deviation of each norm and the average pitch were used as evaluation indices of the magnitude of sway of the body. Each parameter was calculated in two areas: the Walking Area (from the 3rd to 6th steps) and the Passing Area (from the 7th to 10th steps). Multiple regression analysis was performed to determine the factors that explain the SATQ scores. All explanatory variables were standardized, and a multiple linear regression analysis was performed. The results revealed that when the distance between the two subjects at the time of passing each other was 20 cm, there was a strong correlation between the SATQ score and the standard deviation of the norm of the angular velocity of the waist from the 7th to 10th steps. This finding suggests that compared with TD individuals, individuals with severe ASD traits significantly increase the standard deviation of the norm of the angular velocity of the waist in order to avoid contact with their counterparts.
Evaluation of the impeller radial stability is important from the bioengineering point of view in the development of mechanical circulatory support devices (MCSDs) for safer use as bridge for several months or destination therapy for years. In this study, radial stability of a magnetically levitated impeller in a centrifugal blood pump with an axially magnetic suspension system was evaluated by investigating the effects of the eccentric impeller position on passive stability, aiming to propose a pump design guide for the development and safer clinical use of MCSDs. First, impeller displacements in the prototype pump were measured using a mock loop together with laser displacement sensors. Then, the radial hydraulic forces exerted on an eccentric impeller were calculated using computational fluid dynamics (CFD) analysis for four volute-casing geometries. In addition, hemocompatibility was assessed using CFD calculations of scalar shear stress exerted on blood. Measurement of impeller displacement showed that the displacement varied from 0.56 to 0.27 mm at a rotational speed of 1800 rpm as the flow rate increased from 0 to 6.5 L/min. In the CFD calculation, the radial hydraulic force increased linearly from 0.2 to 1.7 N as the impeller displacement increased from 0 to 0.5 mm for all the double volute geometries, under conditions of a rotational speed of 1800 rpm and flow rates of 3, 5 and 7 L/min. These results indicate that the impeller stability in the prototype pump is acceptable at the operation conditions of ventricular assist devices, because the magnetic bearing stiffness of radial component was 4.1 N/mm. In the pressure recovery analysis of eccentric impellers, a double volute was not effective because of the unbalanced pressure field generated by the unbalanced pressure recovery. Thus, the increase in radial hydraulic force associated with an eccentric impeller could not be avoided by changing the conventional double volute design. The CFD analysis of geometrical variation indicated that widening of the radial clearance is an effective approach to improve the radial stability as well as hemocompatibility, although the radial clearance should be designed based on trade-offs among impeller stability, hemocompatibility and pump performance.
In this work, we developed a measurement system that uses LEDs to estimate multiple components such as urea and creatinine in spot urine samples using near-infrared spectroscopy, considering future transition to LED light sources. In this study, we chose LEDs with 10 standard wavelengths (1400–2300 nm, in 100 nm increments). A multiple regression analysis using all combinations of 10 wavelengths was performed. We prepared glucose-added urine samples (GAU, urine samples from 10 healthy adults, each mixed with glucose). Wavelength selection was performed by comparing the minimum standard error of prediction (SEP, calculated from actual concentration and predicted concentration) for each wavelength combination. We obtained high accuracy for estimating urinary urea and creatinine levels (SEP: 42.4 mg/dl and 7.34 mg/dl, respectively) using four wavelengths for urea including two wavelengths showing negative absorbance, and five wavelengths for creatinine. Furthermore, an extremely high correlation coefficient (γ > 0.99) was obtained for both components. We calculated urea concentration, creatinine concentration, and urea-to-creatinine ratio using this optical, reagentless method. The low SEP and high γ show that our method is suitable for practical determination of urea-to-creatinine ratio. Thus, this method of analyzing urine samples using NIR spectroscopy can be used to assess protein intake in CKD patients.
Extracellular volume fraction mapping (ECV Map) can provide quantitative measurements of myocardial tissue with amyloid deposition and myocardial edema. ECV measurements have been shown to correlate well with myocardial fibrosis. Pixel-wise ECV Maps are calculated from acquired precontrast and postcontrast T1 Maps calibrated by blood hematocrit. The maps are acquired with ECG triggering and breath holding. However, ECV measurement is not accurate when heart motion occurs because of inconsistent and inadequate breath holding during image acquisition. We present an application of motion-correction algorithm for ECV Maps in cardiac MRI. Our proposed method is based on aligning the position of the heart between precontrast and postcontrast T1 Maps before calculating the ECV Map. The problem with this registration is spatial displacement of the myocardium because of different diaphragm positions. We have developed an automatic approach to detect the displacement before and after contrast injection, and the ECV Map is measured with correction of the myocardium position considering the displacement. We confirmed that our proposed method improves the accuracy of the ECV Map regardless of the size of displacement.
Wearable thermometers are popular devices for measuring body temperature during fever, as well as for monitoring basal temperature in women. They are easy to handle, inexpensive, accurate and provide continuous recordings. Most wearable thermometers are connected to a smart phone or tablet to display data. Many types of wearable thermometer are available, such as touch, patch and invisible (radiometric) types. In this review, we describe and discuss currently available wearable thermometers.
After experiencing natural disasters, accidents, and shocking incidents, some children experience post-traumatic stress disorder (PTSD). A respiration control method that relaxes the body and mind may efficiently improve PTSD symptoms. Therefore, we developed a stuffed toy using two airbags to measure the respiration wave and guide a child's respiration using the up-and-down movements of the toy's abdomen to help them relax. The respiration wave was measured by the sensing device, and the child's respiration could be led by the moving device. We then performed an evaluation experiment. Participants in the experiment consisted of nine healthy girls aged 8–10 years. The results showed that the respiration guidance resulted in fewer variations in the children's mean heart rate than that with just hugging (p < 0.05). This result suggests that the respiration-guiding trial relaxes children compared with just hugging the stuffed toy without motion. The effectiveness of this stuffed toy in improving PTSD symptoms in children should be further evaluated.
Surface electromyographic (EMG) signals are known to be strongly influenced by anatomical, physiological and detection system parameters. Among the detection system parameters, we are interested in the effect of muscle fiber inclination on the electrode arrangement. The purpose of this study was to determine the best and the worst orientation of the electrodes arranged in nine detection systems relative to the muscle fiber direction and also to classify the investigated systems according to their degree of isotropy. The study was based on simulated surface EMG (sEMG) signals generated in a cylindrical multilayer volume conductor. The orientation of electrodes with respect to the fiber direction was defined by the fiber inclination angle (FIA). For each detection system, the mean power (MP) of the simulated signals was computed at different FIAs and used as a basis for evaluating the effect of muscle fiber inclination. We showed that for the FIA range of 0–180°, approximately isotropic systems had three positions to record sEMG signals under good conditions (MP was maximum). However, longitudinal and transversal highly anisotropic systems had two and one positions, respectively, at which sEMG signals were detected under good conditions. We showed also that the degree of isotropy of the nine detection systems investigated was less affected by the increase in muscle and fat thicknesses. However, with an increase in inter-electrode distance (IED), the degree of isotropy of approximately isotropic systems decreased while the degree of isotropy of highly anisotropic systems increased.
AI-based medical and healthcare devices and systems have unique characteristics including 1) plasticity causing changes in system performance through learning, and need of creating new concepts about the timing of learning and assignment of responsibilities for risk management; 2) unpredictability of system behavior in response to unknown inputs due to the black box characteristics precluding deductive output prediction; and 3) need of assuring the characteristics of datasets to be used for learning and evaluation. The Subcommittee on Artificial Intelligence and its Applications in Medical Field of the Science Board, the Pharmaceuticals and Medical Devices Agency (PMDA), Tokyo, Japan, examined “new elements specific to AI” not included in conventional technologies, thereby clarifying the characteristics and risks of AI-based technologies. This paper summarizes the characteristics and clinical positioning of AI medical systems and their applications from the viewpoint of regulatory science, and presents the issues related to the characteristics and reliability of data sets in machine learning.
We examined muscle stiffness at various pedaling rates under conditions of constant power output. Eight healthy young participants pedaled a cycle ergometer at a power output of 47 W. The combinations of pedaling rate and workload were 40 revolutions per min (rpm) and 11.7 N, 60 rpm and 7.8 N, and 80 rpm and 5.9 N, respectively. One electrical stimulus per two pedal rotations was applied to the vastus lateralis muscle at a crank angle of 30° in the down phase. Mechanomyograms (MMGs) were measured using a capacitor microphone, and the evoked MMG was extracted. An evoked MMG system was identified, and the coefficients of the denominator of the transfer function were used to estimate stiffness and viscous coefficient of the muscle. Muscle stiffness was 236–705 Nm−1, and was proportional to the pedaling rate when power output was held constant, while the viscous coefficient did not change from approximately 15 Nm−1s. In conclusion, our findings demonstrate that stiffness of the vastus lateralis muscle increases with increasing pedaling rate under conditions of constant power output, while the viscous coefficient does not change.
We have proposed and developed a vibratory microinjection system (VMS) to facilitate gene transfer and to increase the efficiency in production of transgenic animals. The VMS injects transgene into a cell by vibrating an injection micropipette longitudinally at a frequency in the order of kilohertz. The vibrator in the second version of VMS consisted of one cylindrical stacked-type piezoelectric actuator and a housing, and provided any vibration frequency lower than 18 kHz. First of all, we found that the amplitudes at the level of the vibrator were perfectly proportional to the voltages applied to the vibrator. Then, we investigated both possible negative and positive effects of VMS. (1) Vibration may break transgene, but we could not find any fragmentation of a 3.5-kilobase-pair (kb) transgene even after an extremely long period of vibration at an extremely high injection pressure. (2) In a one-hour in vitro experiment, the vibratory microinjection ejected more than double the volume of transgene solution that the conventional (non-vibratory) microinjection did, indicating that the vibration increased the ejection speed by more than 100%. (3) Compared with the non-vibratory microinjection, we investigated the effects of VMS on pronuclear microinjection, in which a small amount of transgene is injected into one of two pronuclei in a zygote, and examined two vibration conditions: 5 kHz - 15 Vp-p and 10 kHz - 8 Vp-p. The vibratory microinjection pierced zygotes with significantly smaller cellular deformation, decreased the lethal events of a micropipette pulling out pronuclear components, and resulted in significantly better in vitro development to the blastocyst stage, although considerably higher death rates were also observed. The coexistence such as better embryonic development and higher death rate seemed contradictory, but the contradiction was probably derived from wide variations in amplitude among micropipettes. More specifically, we applied a fixed vibration condition to different micropipettes because we had confirmed that the amplitudes at the level of vibrator were stable. The micropipettes, however, could have been considerably different in vibration characteristic although they were commercially available products in accordance with industrial standard. Therefore, some micropipettes would have been effective, others ineffective or even injurious. In conclusion, the second version of VMS heralded a new approach to pronuclear microinjection because it showed several merits except for the higher death rates.
An extracorporeal membrane oxygenator (ECMO) is used for the management of severe heart or respiratory failure. In the conventional ECMO system, centrifugal blood pumps are commonly used in recent years. However, a relatively high rotational speed is required to perfuse against the high hydrodynamic resistance of the system circuit. High rotational speed causes high shear stress. To provide a blood pump that can be operated at a lower rotational speed than conventional centrifugal blood pumps, we propose a novel high-pressure type rotary blood pump named the toroidal convolution pump (TCP). In the TCP, fluid that acquires centrifugal force from rotation of the impeller is returned to the impeller through the flow path and again acquires centrifugal force. With repetition of this cycle, fluid is accelerated to generate high pressure. We investigated the performance and basic property of the TCP using an experimental model and computational fluid dynamic (CFD) analysis. The TCP generated pump output of 5 L/min against a pressure head of 350 mmHg at a rotational speed of 2450 rpm. This rotational speed is much lower than that of conventional centrifugal blood pumps, which is usually higher than 3000 rpm. The efficiency of the TCP including the motor was approximately 4.2% at that setting. CFD analysis showed symmetrical pressure distribution about the central pivot bearing. The pressure difference between inlet and outlet ports was approximately 40% higher than that of the experimental model. We found no excessive negative pressure that would cause hemolysis. Although we identified areas of high sheer stress, hemolysis is estimated to be low because of the short exposure time to the high sheer stress. We found no stagnant area that would cause thrombus formation.
In this paper, we propose an algorithm to construct a spatiotemporal statistical shape model (SSM) of the brain surface of a human embryo. Our model is based on level sets and the model is constructed independently for each Carnegie stage (CS), in which feature space is constructed by spatially-weighted principal component analysis (PCA). The statistics of the shape variation were calculated under the assumption of q-Gaussian distribution to reduce the risk of overfitting for datasets of small sizes. To define the statistics of the shape distribution of the intermediate CS, we consider 18 interpolation methods, which are the combinations of three average interpolations (linear, B-spline and information geometry) and six covariance interpolation (rotation, affine-invariant, log-Euclidean, information geometry, Wasserstein geometry and tensor B-spline) techniques, which are exhaustively compared in the experiments. We also propose a method to allow interpolation between distributions with a mixed number of dimensions, i.e., the rank of the covariance matrix. The SSM was constructed and evaluated using 60 sets of brain labels acquired from human embryos from the Kyoto collection. We found that the best interpolation method was a combination of the linear average interpolation method with an information geometry–based covariance interpolation method. We also found that use of the q-Gaussian distribution and selection of the number of dimensions are effective methods for improving the performance of the spatiotemporal SSM of the human embryo.
Because radiofrequency (RF) coils used in magnetic resonance imaging (MRI) are specific for given regions of interest in the human head (HH), it is necessary to adjust the capacitance when setting the resonance frequency of MRI. Simulation-based development can reduce the number of components and manufacturing processes, but to obtain the same conditions as in the actual HH when RF coils are loaded, it is necessary to execute simulation of the head model under loading conditions. The HH model and spherical model are commonly reported for modeling the HH. In simulation, the HH model requires long calculation time because of its complicated shape. On the other hand, the calculated resonance characteristics of a spherical model are significantly different from those of the HH model, because the spherical model is oversimplified. We designed a new HH model by modifying the 3D Shepp-Logan Phantom model and evaluated its performance using a high frequency electromagnetic field simulator. Using our new head model in simulation, the approximate resonant frequency of the RF coil was calculated at 1/6 the calculation time of that when using the HH model, and the actual loading condition of the HH was reproduced.