Over the past decade, there has been an increasing interest in the real-time monitoring of ambulatory vital signs such as heart rate (HR) and arterial blood oxygen saturation (SpO2) using wearable medical sensors during field operations. These measurements can convey valuable information regarding the state of health and allow first responders and front-line medics to better monitor and prioritize medical intervention of military combatants, firefighters, miners and mountaineers in case of medical emergencies. However, the primary challenge encountered when using these sensors in a non-clinical environment has been the presence of persistent motion artifacts (MA) embedded in the acquired physiological signal. These artifacts are caused by the random displacement of the sensor from the skin and lead to erroneous output readings. Several signal processing techniques, such as time and frequency domain segmentation, signal reconstruction techniques and adaptive noise cancellation (ANC), have been previously developed in an offline environment to address MA in photoplethysmography (PPG) with varying degrees of success. However, the performance of these algorithms in a spasmodic noise environment usually associated with basic day to day ambulatory activities has still not been fully investigated. Therefore, the focus of this research has been to develop novel MA algorithms to combat the effects of these artifacts. The specific aim of this thesis was to design two novel motion artifact (MA) algorithms using a combination of higher order statistical tools namely Kurtosis (K) for classifying 10 s PPG data segments, as either ‘clean’ or ‘corrupt’ and then extracting the aforementioned vital parameters. To overcome the effects of MA, the first algorithm (termed ‘MNA’) processes these ‘corrupt’ PPG data segments by identifying abnormal amplitudes changes. The second algorithm (termed ‘MNAC’), filters these ‘corrupt’ data segments using a 16th order normalized least mean square (NLMS) ANC filter and then extracts HR and SpO2.

Over the past decade, there has been an increasing interest in the real-time monitoring of ambulatory vital signs such as heart rate (HR) and arterial blood oxygen saturation (SpO2) using wearable medical sensors during field operations. These measurements can convey valuable information regarding the state of health and allow first responders and front-line medics to better monitor and prioritize medical intervention of military combatants, firefighters, miners and mountaineers in case of medical emergencies. However, the primary challenge encountered when using these sensors in a non-clinical environment has been the presence of persistent motion artifacts (MA) embedded in the acquired physiological signal. These artifacts are caused by the random displacement of the sensor from the skin and lead to erroneous output readings. Several signal processing techniques, such as time and frequency domain segmentation, signal reconstruction techniques and adaptive noise cancellation (ANC), have been previously developed in an offline environment to address MA in photoplethysmography (PPG) with varying degrees of success. However, the performance of these algorithms in a spasmodic noise environment usually associated with basic day to day ambulatory activities has still not been fully investigated. Therefore, the focus of this research has been to develop novel MA algorithms to combat the effects of these artifacts. The specific aim of this thesis was to design two novel motion artifact (MA) algorithms using a combination of higher order statistical tools namely Kurtosis (K) for classifying 10 s PPG data segments, as either ‘clean’ or ‘corrupt’ and then extracting the aforementioned vital parameters. To overcome the effects of MA, the first algorithm (termed ‘MNA’) processes these ‘corrupt’ PPG data segments by identifying abnormal amplitudes changes. The second algorithm (termed ‘MNAC’), filters these ‘corrupt’ data segments using a 16th order normalized least mean square (NLMS) ANC filter and then extracts HR and SpO2.

Despite having the numerous evolved heart rate measuring devices and progress in their development over the years, there always remain the challenges of modern signal processing implementation by a comparatively small size wearable device. This thesis paper presents a wearable reflectance photoplethysmography (PPG) sensor system for measuring the heart rate of a user both in steady and moving states. The size and, power consumption of the device are considered while developing, to ensure an easy deployment of the unit at the measuring site and the ability to power the entire unit with a battery .The selection of both the electronic circuits and signal processing techniques is based on their sensitivity to PPG signals, robustness against noise inducing artifacts and miniaturization of the entire measuring unit. The entire signal chain operates in the discrete-time, which allows the entire signal processing to be implemented in firmware on an embedded microprocessor. The PPG sensor system is implemented on a single PCB that consumes around 7.5mW of power. Benchmarking tests with standard heart rate measuring devices reveal that the developed measurement unit (combination of the PPG sensor system, and inertial measurement unit (IMU) developed in-house at Acreo Swedish ICT, and a battery) is comparable to the devices in detecting heart rate even in motion artifacts environment. This thesis work is carried out in Acreo Swedish ICT, Gothenburg, Sweden in collaboration with MidSweden University, Sundsvall, Department of Electronics Design. This report can be used as ground work for future development of wearable heart rate measuring units at Acreo Swedish ICT.

Motion artifacts are often a significant component of the measured signal in functional near-infrared spectroscopy (fNIRS) experiments. A variety of methods have been proposed to address this issue, including principal component analyses (PCA), Kalman filtering, correlation-based signal improvement (CBSI), wavelet filtering, spline interpolation, and autoregressive algorithms. The efficacy of these techniques has been compared using simulated data; however, our understanding of how these techniques fare when dealing with task-based cognitive data is limited. Recently, Brigadoi et al. (2014) quantitatively compared 6 motion correction techniques in a sample of adult data measured during a simple cognitive task. Wavelet filtering showed the most promise as an optimal technique for motion correction. Because fNIRS is often used with infants and young children, it is critical to evaluate the effectiveness of motion correction techniques directly with data from these age groups. Here we examined which techniques are most effective with data from young children. The efficacy of each technique was compared quantitatively using objective metrics related to the physiological properties of the hemodynamic response using two different sets of parameters to ensure maximum retention of included trials. Results showed that targeted PCA (tPCA) and CBSI retained a higher number of trials. These techniques also performed well in direct head-to-head comparisons with the other approaches using both quantitative metrics and a qualitative assessment. The CBSI technique corrected many of the artifacts present in our data; however, this technique was highly influenced by the parameters used to detect motion. The tPCA technique, by contrast, was robust across changes in parameters while also performing well across all comparison metrics. We conclude, therefore, that tPCA is an effective technique for the correction of motion artifacts in fNIRS data from young children.

Pulse oximetry is used to measure heart rate (HR) and arterial oxygen saturation (SpO2) from photoplethysmographic (PPG) waveforms. PPG waveforms are highly sensitive to motion artifact (MA), limiting the implementation of pulse oximetry in mobile physiological monitoring using wearable devices. Previous studies have shown that multichannel pulse oximetry can successfully acquire diverse signal information during simple, repetitive motion, thus leading to differences in motion tolerance across channels. In this study, we introduce a multichannel forehead-mounted pulse oximeter and investigate the performance of this novel sensor under a variety of intense motion artifacts. We have developed a multichannel template-matching algorithm that chooses the channel with the least amount of motion artifact to calculate HR and SpO2 every 2 seconds. We show that for a wide variety of random motion, channels respond differently to motion, and the multichannel estimate outperforms single channel estimates in terms of motion tolerance, signal quality, and HR and SpO2 error. Based on 31 data sets of PPG waveforms corrupted by random motion, the mean relative HR error was decreased by an average of 5.6 bpm when the multichannel-switching algorithm was compared to the worst performing channel. The percentage of HR measurements with absolute errors ≤ 5 bpm during motion increased by an average of 27.8 % when the multichannel-switching algorithm was compared to the worst performing channel. Similarly, the mean relative SpO2 error was decreased by an average of 4.3 % during motion when the multichannel-switching algorithm was compared to each individual channel. The percentage of SpO2 measurements with absolute error ≤ 3 % during motion increased by an average of 40.7 % when the multichannel-switching algorithm was compared to the worst performing channel. Implementation of this multichannel algorithm in a wearable device will decrease dropouts in HR and SpO2 measurements during motion. Additionally, the differences in motion frequency introduced across channels observed in this study shows precedence for future multichannel-based algorithms that make pulse oximetry measurements more robust during a greater variety of intense motion.

Radiation therapy (RT) is commonly used to treat patients with lung cancer. One of the limitations of RT is that irradiation of the surrounding healthy lung tissues during RT may cause damage to the lungs. Radiation-induced pulmonary toxicity may be mitigated by minimizing doses to high-function lung tissues, which we refer to as functional avoidance RT. Lung function can be computed by image registration of treatment planning four-dimensional computed tomography (4DCT), which we refer to as CT ventilation imaging. However, the accuracy of functional avoidance RT is limited by lung function imaging accuracy and artifacts in 4DCT. The goal of this dissertation is to improve the accuracy of functional avoidance RT by overcoming those two limitations. A common method for estimating lung ventilation uses image registration to align the peak exhale and inhale 3DCT images. This approach called the 2-phase local expansion ratio is limited because it assumes no out-of-phase lung ventilation and may underestimate local lung ventilation. Out-of-phase ventilation occurs when regions of the lung reach their maximum (minimum) local volume in a phase other than the peak of inhalation (end of exhalation). This dissertation presents a new method called the N-phase local expansion ratio for detecting and characterizing locations of the lung that experience out-of-phase ventilation. The N-phase LER measure uses all 4DCT phases instead of two peak phases to estimate lung ventilation. Results show that out-of-phase breathing was common in the lungs and that the spatial distribution of out-of-phase ventilation varied from subject to subject. On average, 49% of the out-of-phase regions were mislabeled as low-function by the 2-phase LER. 4DCT and Xenon-enhanced CT (Xe-CT) of four sheep were used to evaluate the accuracy of 2-phase LER and N-phase LER. Results show that the N-phase LER measure was more correlated with the Xe-CT than the 2-phase LER measure. These results suggest that it may be better to use all 4DCT phases instead of the two peak phases to estimate lung function. The accuracy of functional avoidance RT may also be improved by reducing the impact of artifacts in 4DCT. In this dissertation, we propose a a geodesic density regression (GDR) algorithm to correct artifacts in one breathing phase by using artifact-free data in corresponding regions of the other breathing phases. Local tissue density change associated with CT intensity change during respiration is accommodated in the GDR algorithm. Binary artifact masks are used to exclude regions of artifacts from the regression, i.e., the GDR algorithm only uses artifact-free data. The GDR algorithm estimates an artifact-free CT template image and its time flow through a respiratory cycle. Evaluation of the GDR algorithm was performed using both 2D CT time-series images with simulated known motion artifacts and treatment planning 4DCT with real motion artifacts. The 2D results show that there is no significant difference (p-value = 0.95) between GDR regression of artifact data using artifact masks and regression of artifact-free data. In contrast, significant errors (p-value = 0.005) were present in the estimated Jacobian images when artifact masks were not used. We also demonstrated the effectiveness of the GDR algorithm for removing real duplication, misalignment, and interpolation artifacts in 4DCT. Overall this dissertation proposes methods that have the potential to improve functional avoidance RT by accommodating out-of-phase ventilation, and removing motion artifacts in 4DCT using geodesic image regression.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第22376号 / 医博第4617号 / 新制||医||1043(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 松田 道行, 教授 林 康紀, 教授 江藤 浩之 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Modern Automatic Blood Pressure Measurement Techniques are based on measuring the cuff pressure and on sensing the pulsatile amplitude variations. These measurements are very sensitive to motion of the patient or the surroundings where the patient is. The slightest unexpected movements could offset the readings of the automatic Blood Pressure meter by a large amount or render the readings totally meaningless. Every effort must be taken to avoid subjecting the body of the patient or the patient's surroundings to motion for obtaining a reliable reading. But there are situations in which we need Blood Pressure Measurements with the patient or his surroundings in motion; for instance in an ambulance while a patient is being transported to a hospital. In this thesis, we present a technique to reduce the effect of motion artifact from Blood Pressure measurements. We digitize the blood pressure waveform and use Digital Signal Processing Techniques to process the corrupted waveform. We use the differences in frequency spectra of the Blood Pressure signal and motion artifact noise to remove the motion artifact noise. The motion artifact noise spectrum is not very well defined, since it may consist of many different frequency components depending on the kind of motion. The Blood Pressure signal is more or less a periodic signal. That translates to periodicity in the frequency domain. Hence, we designed a digital filter that could take advantage of the periodic nature of the Blood Pressure Signal waveform. The filter is shaped like a comb with periodic peaks around the signal frequency components. Further processing of the filtered signal: baseline restoration and level shifting help us to further reduce the noise corruption.

No description available.

Laser speckle contrast imaging (LSCI) is a non-invasive method for assessment of microcirculatory blood flow. The technique is based on analysis of speckle patterns to build 2D maps of perfusion with high spatial and temporal resolution. A drawback of the method is that it is highly sensitive to motion artifacts since the perfusion estimates are based on quantification of the motion blurring in the images. The camera is today limited to a bulky stand for good measurements, but even as it is fixed, it does not ensure that the patient is completely still. In many clinical settings, it would be advantageous to have a more flexible camera and to be able to detect if an image is influenced by external motion. Multi-exposure laser speckle contrast imaging (MELSCI) is an extension to LSCI that utilizes the contrast from multiple exposure times. The gain in information has paved way for more accurate perfusion estimates. The technique has been limited due to its computational complexity, but recently a real time system has been developed. The goals of this thesis was twofold, firstly find a quantifiable measure of motion artifacts to be able to evaluate and compare LSCI and MELSCI. Secondly, propose an algorithm that detects movements in LSCI recordings. Motion artifacts in LSCI and MELSCI were investigated by developing a setup where repeatable movements could be made. Measurements of a hand influenced by motions of different speeds and directions were acquired and the relative difference between motion and static states were calculated and compared for the two systems. The relative difference of the MELSCI measurements were lower for all speeds above 0.57 mm/s, indicating more robustness to motion artifacts. A detection algorithm using image registration to calculate the instantaneous speed in each frame of the recording was developed. The method successfully detects movements perpendicular to the camera and shows that the intensity images of an LSCI recording can be used to give a direct indication of when movement has occurred.