This dissertation evaluates the design of several defense schemes for wireless medical devices to address security issues. These schemes are designed to enable efficient and effective access control of wireless medical devices in both non-emergency and emergency situations. In recent years, the range of available wireless medical devices has increased and includes cardiac pacemakers, insulin pump, defibrillators, cochlear implants, neurostimulators, and various drug delivery systems. Unfortunately, most existing wireless medical devices lack sufficient security mechanisms to protect patients from malicious attacks. Thus, with the rise in use of medical implants, security becomes a critical issue as attacks on wireless medical devices may harm patients. Security on wireless medical devices is a relatively new field, which has not been thoroughly researched yet. The authors of a lot of articles have proposed token based, certification based and proximity based schemes to address the issue. However, most of the current solutions have many limitations and cannot be widely applied. Therefore, better solutions are needed. In order to address this issue, we design a novel and multiple-layer access control framework for wireless medical devices. In a low layer level, we utilize bi-channel technology and multi-factor authentication to defend against various attacks at wireless medical devices. Our system utilizes near field communication (NFC) to do device pairing and uses the medical device's wireless radio to perform remote programming. This approach defends against most attacks because our NFC pairing scheme guarantees that the successful communication range between the programmer and wireless medical devices is less than 6cm. When the patient is in a crowded area such as on public transportation, a different person's mobile devices and the patient's medical devices may be located less than 6cm apart; we use the patient's cell phone to detect such an environment. To avoid attacks in crowded areas, we design a scheme to detect such a situation using the patient's cell phone. User involvement is used on non-implantable medical devices (IMDs) and a patient access pattern based access control (PAPAC) scheme is used on IMDs. We also design a response time based scheme to defend against fake patient attacks. Our analyses and experiments show that the protection schemes are efficient and effective. In a high layer level, we design patient infusion pattern based access control (PIPAC) scheme for wireless medical devices. Specifically, insulin pumps are most widely applied wireless medical devices. The pump parameters and doses can be adjusted by anyone with an easily obtained USB device. The hacker can deliver a lethal dose without knowing the device's serial number in advance. To address this issue, we propose a PIPAC for wireless insulin pumps. This scheme employs a supervised learning approach to learn normal patient infusion patterns in terms of the dosage amount, rate, and time of infusion, which are automatically recorded in insulin pump logs. The generated regression models are used to dynamically configure a safe infusion range for abnormal infusion identification. Our proposed algorithms are evaluated with real insulin pump logs used by several patients for up to 6 months. The experimental results demonstrate that our scheme can reliably detect a single overdose attack with a success rate up to 98\% and defend against a chronic overdose attack with a very high success rate. For IMDs in non-emergency case, the PAPAC scheme we design utilizes the patient's IMD access pattern to address resource depletion (RD) attacks. It is a novel support vector machine (SVM) based scheme. This SVM based scheme is very effective at defending against RD attacks. Our experimental results show that the average detection rate is above 90\%. For IMDs in emergency cases, we design a novel biometrics based two-level secure access control scheme that utilizes a patient's biometrics to prevent unauthorized access to the IMD. The scheme consists of two levels: level-one employs a patient's some basic biometrics and is lightweight; level-two uses a patient's customized iris data to achieve effective authentication. The experimental results show that our IMD access control scheme is very effective and has small overhead in terms of battery, CPU and memory. Thus, it is suitable for IMDs. Both the false acceptance rate (FAR) and false rejection rate (FRR) are close to zero with a suitable threshold. Protecting wireless medical devices is a very challenging task due to their extremely limited resource constraints. It is necessary to balance the overhead of security schemes and security requirements. In this dissertation, we will first discuss security vulnerabilities in wireless medical device systems. Then we will present our framework using smart phones and other technologies, such as near field communication based access control. Further, we will describe the detailed design of this framework. Finally, extensive experiments show that our schemes can achieve good performance with small overhead. / Computer and Information Science

<p dir="ltr">Microfluidic techniques have been widely adopted in biomedical research due to the pre- cise control of fluids, small volume requirement, low cost and etc, and have boosted the development of biomolecular interaction analysis, point-of-care diagnostics, and biosensors.</p><p dir="ltr">Protein-protein interaction plays a key role in biological, biomedical and pharmaceutical research. The technical development of biosensors, new drugs and vaccines, and disease diagnostics heavily rely on the characterization of protein-protein interaction kinetics. The current gold standard assays for measuring protein-protein interaction are surface plasmon resonance (SPR), and bio-layer interferometry (BLI). These commercial devices are accurate but expensive, however.</p><p dir="ltr">Here, I have developed new microfluidic techniques and models in protein-protein in- teraction kinetics measurement, rotational diffusion coefficient modeling, electrochemical impedance spectroscopy-based biosensors, and two-phase porous media flow models. Firstly, I applied particle diffusometry (PD) in the streptavidin-biotin binding kinetics measurement, utilizing a Y-junction microchannel. Secondly, to reduce solution volumes used in an analysis experiment, I designed a low-volume chip and coupled it with PD to measure the binding kinetics of human immunodeficiency virus p24 antibody-antigen interactions. Thirdly, con- sidering the Brownian motion of the non-symmetric particles, I developed a new model to efficiently compute particles’ rotational diffusion coefficients. Fourthly, to make economic biosensors to detect multiple biomarkers, I created a new chip, enabling hundreds of tests in a single droplet (∼ 50 μL) on one chip. Finally, to understand the liquid flow in porous media, such as nitrocellulose in lateral flow assays, I built a new two-phase porous media flow model based on the Navier-Stokes equation and compared it with experiments. These techniques and models underwent rigorous experimental and computational validation, demonstrating their effectiveness and performance.</p>

<p dir="ltr">Robotic surgery has transformed the landscape of minimally invasive procedures, offering unmatched precision and quicker patient recovery times. Despite all these advancements, training surgeons to use these sophisticated surgical systems effectively remains a daunting challenge, primarily due to high costs, limited accessibility, increased learning curve, and inconsistent training quality. Existing training modalities are limited by the high costs of original training robots, logistical challenges, lack of emphasis on hand movements, the necessity of expert presence, and limited scalability and effectiveness. This thesis introduces TrainVR, a low-cost based training system designed to overcome these hurdles and enhance the skillset of surgical trainees. TrainVR integrates affordable Virtual Reality (VR) technology with enhanced fidelity, creating an engaging and realistic training environment. TrainVR is designed to simulate realistic surgical environments and procedures, focusing on the development of motor, cognitive, and spatial skills for tasks required for robotic surgery through computer vision algorithms, gamified environments, performance analytics, and supporting both asynchronous and remote expert-led training scenarios. This system features customizable training modules, enabling trainees to practice a wide array of surgical procedures in a safe, virtual setting. The device also focuses on the importance of user’s hand, clutch, and ergonomics during surgical training which is crucial based on feedback from surgeons. The development of TrainVR involved crafting detailed 3D models of surgical instruments and anatomical structures, by integrating hardware, software and designing a user-friendly interface. We conducted testing with different game environments which compare the performance of the users and provide insights to improve the learning. The thesis concludes by experimenting and proposing new configurations to improve the fidelity and hand tracking which should closely match with the experience provided by the present training simulators at a substantially lower cost. TrainVR’s scalable design and compatibility with standard VR hardware make it accessible to a wide range of institutions, including those with limited resources. By offering a cost-effective, immersive, and adaptive training solution, TrainVR aims to enhance surgical education and ultimately improve patient care outcomes.</p>

The aim of this thesis is to derive an unsupervised method for detecting anomalies in time series. Autoencoder-based approaches are widely used for the task of detecting anomalies where a model learns to reconstruct the pattern of the given data. The main idea is that the model will be good at reconstructing data that does not contain anomalous behavior. If the model fails to reconstruct an observation it will be marked as anomalous. In this thesis, the derived method is applied to data from active medical devices manufactured by B. Braun. The given data consist of 6,000 length-varying time series, where the average length is greater than 14,000. Hence, the given sample size is small compared to their lengths. Subsequences of the same pattern where anomalies are expected to appear can be extracted from the time series taking expert knowledge about the data into account. Considering the subsequences for the model training, the problem can betranslated into a problem with a large dataset of short time series. It is shown that a common autoencoder is able to reconstruct anomalies well and is therefore not useful to solve the task. It is demonstrated that a variational autoencoder works better as there are large differences between the given anomalous observations and their reconstructions. Furthermore, several thresholds for these differences are compared. The relative number of detected anomalies in the two given datasets are 3.12% and 5.03%.

<p>  </p> <p>Understanding the mechanical properties of peripheral nerves is essential for chronically implanted device design. The work in this thesis aimed to understand the relationship between local deformation responses to global strain changes in peripheral nerves. A custom-built mechanical testing rig and sample holder enabled an improved cyclical uniaxial tensile testing environment on rabbit sciatic nerves (N=5). A speckle was placed on the surface of the nerve and recorded with a microscope camera to track local deformations. The development of a semi-automated digital image processing algorithm systematically measured local speckle dimension and nerve diameter changes. Combined with the measured force response, local and global strain values constructed a stress-strain relationship and corresponding elastic modulus. Preliminary exploration of models such as Fung and 2-Term Mooney-Rivlin confirmed the hyperelastic nature of the nerve. The results of strain analysis show that, on average, local strain levels were approximately five times smaller than globally measured strains; however, the relationship was dependent on global strain magnitude. Elastic modulus values corresponding to ~9% global strains were 2.070 ± 1.020 MPa globally and 10.15 ± 4 MPa locally. Elastic modulus values corresponding to ~6% global strains were 0.173 ± 0.091 MPa globally and 1.030 ± 0.532 MPa locally.   </p>

Master of Science / Department of Computing and Information Sciences / John Hatcliff / Medical devices are safety-critical systems whose failure may put human life in danger. They are becoming more advanced and thus more complex. This leads to bigger and more complicated code-bases that are hard to maintain and verify. Model-driven development provides high-level and abstract description of the system in the form of models that omit details, which are not relevant during the design phase. This allows for certain types of verification and hazard analysis to be performed on the models. These models can then be translated into code. However, errors that do not exist in the models may be introduced during the implementation phase. Automated translation from verified models to code may prevent to some extent. This thesis proposes approach for model-driven development and verification of medical devices. Models are created in AADL (Architecture Analysis & Design Language), a language for software and hardware architecture modeling. AADL models are translated to SPARK Ada, contract-based programming language, which is suitable for software verification. Generated code base is further extended by developers to implement internals of specific devices. Created programs can be verified using SPARK tools. A PCA (Patient Controlled Analgesia) pump medical device is used to illustrate the primary artifacts and process steps. The foundation for this work is "Integrated Clinical Environment Patient-Controlled Analgesia Infusion Pump System Requirements" document and AADL Models created by Brian Larson. In addition to proposed model-driven development approach, a PCA pump prototype was created using the BeagleBoard-xM device as a platform. Some components of PCA pump prototype were verified by SPARK tools and Bakar Kiasan.

Prélevés de manière non invasive, les gaz respiratoires sont constitués de nombreux composés organiques volatils (VOCs) dont la quantité dépend de l’état de santé du sujet. L’analyse quantitative de l’air expiré présente alors un fort intérêt médical, que ce soit pour le diagnostic ou le suivi de traitement. Dans le cadre de ma thèse, nous proposons d’étudier un dispositif d’analyse des gaz respiratoires, et notamment de ces VOCs. Cette thèse multidisciplinaire aborde différents aspects, tels que le choix des capteurs, du matériel et des modes d’acquisition, l’acquisition des données à l’aide d’un banc gaz, et ensuite le traitement des signaux obtenus de manière à quantifier un mélange de gaz. Nous étudions la réponse d’un capteur à oxyde métallique (MOX) à des mélanges de deux gaz (acétone et éthanol) dilués dans de l’air synthétique (oxygène et azote). Ensuite, nous utilisons des méthodes de séparation de sources de manière à distinguer les deux gaz, et déterminer leur concentration. Pour donner des résultats satisfaisants, ces méthodes nécessitent d’utiliser plusieurs capteurs dont on connait la forme mathématique du modèle décrivant l’interaction du mélange avec le capteur, et qui présentent une diversité suffisante dans les mesures d’étalonnage pour estimer les coefficients de ce modèle. Dans cette thèse, nous montrons que les capteurs MOX peuvent être décrits par un modèle de mélange linéaire quadratique, et qu’un mode d’acquisition fonctionnant en double température permet de générer deux capteurs virtuels à partir d’un unique capteur physique. Pour quantifier précisément les composants du mélange à partir des mesures sur ces capteurs (virtuels), nous avons conçu des méthodes de séparation de sources, supervisées et non supervisées appliquées à ce modèle non-linéaire : l’analyse en composantes indépendantes, des méthodes de moindres carrés (algorithme de Levenberg-Marquardt), et une méthode bayésienne ont été étudiées. Les résultats expérimentaux montrent que ces méthodes permettent d’estimer les concentrations de VOCs contenus dans un mélange de gaz, de façon précise, en ne nécessitant que très peu de points de calibration. / Non-invasively taken, exhaled breath contains many volatile organic compounds (VOCs) whose amount depends on the health of the subject. Quantitative analysis of exhaled air is of great medical interest, whether for diagnosis or for a treatment follow-up. As part of my thesis, we propose to study a device to analyze exhaled breath, including these VOCs. This multidisciplinary thesis addresses various aspects, such as the choice of sensors, materials and acquisition modes, the acquisition of data using a gas bench, and then the processing of the signals obtained to quantify a gas mixture. We study the response of a metal oxide sensor (MOX) to mixtures of two gases (acetone and ethanol) diluted in synthetic air (oxygen and nitrogen). Then, we use source separation methods in order to distinguish the two gases, and to determine their concentration. To give satisfactory results, these methods require first to use several sensors for which we know the mathematical model describing the interaction of the mixture with the sensor, and which present a sufficient diversity in the calibration measurements to estimate the model coefficients. In this thesis, we show that MOX sensors can be described by a linear-quadratic mixing model, and that a dual temperature acquisition mode can generate two virtual sensors from a single physical sensor. To quantify the components of the mixture from measurements on these (virtual) sensors, we have develop supervised and unsupervised source separation methods, applied to this nonlinear model: independent component analysis, least squares methods (Levenberg Marquardt algorithm), and a Bayesian method were studied. The experimental results show that these methods make it possible to estimate the VOC concentrations of a gas mixture, accurately, while requiring only a few calibration points.

A economia da saúde, composta pela demanda social de bens e serviços, pela formação de profissionais, por indústrias de alta tecnologia e por produção e consumo de insumos e medicamentos, pode ser entendida por meio dos conceitos de circuito espacial produtivo e círculos de cooperação no espaço, assim como pelo de complexo industrial da saúde. Essa economia desenvolveu-se na relação entre serviços públicos e privados, tendo o Estado como mediador, para corrigir as assimetrias das estratégias de mercado. Estudando suas principais periodizações e o processo brasileiro de industrialização, esta pesquisa sistematiza informações que explicitam a vertente espacial da produção de equipamentos para radiologia e diagnóstico por imagem, que se caracteriza pela profunda interação entre ciência e técnica sob a égide do sistema capitalista de produção e que logram tornar-se globais, e discute a tipologia e a topologia dos equipamentos avaliados, refletindo sobre sua distribuição desigual no território, o que acaba reforçando as contradições socioeconômicas já existentes. Assim, esse período técnico-científico-informacional assiste à permanente introdução de novos objetos técnicos, conformando uma tecnosfera e uma psicosfera que orientam a racionalidade, a irracionalidade e a contrarracionalidade no território e nas práticas médicas contemporâneas. / Health economics, formed by the social demand for goods and services, specific professional education, high technology industries, the production and consumption of medicines and intermediary goods, can be understood through the concepts of productive spatial circuits, circles of cooperation in space and the health industry complex. It has been developed through the relations between public and private services, having the State as mediator to rebalance the asymmetric market strategies. Stressing its most important periods and to the brazilian industrialization process, this research intents to undertake: systematically address data regarding the spatial perspective on the production of radiology and diagnostic imaging devices for medical uses, which is deeply involved with the interaction between science and technique under the auspices of the capitalist system of production, thanks to whom it became global; the characteristics and the topology of the devices studied, reflecting the uneven distribution of these devices in the use the brazilian territory, which end up to reinforce the existing socioeconomic contradictions. This period, technical, scientific and informational, thus, witness the constant addition of technical objects, creating a tecnosphere and a psychosphere that conducts the rationality, counter-rationality in brazilian territory and the contemporary medical practice.

<p>A droplet subjected to external aerodynamic disturbances disintegrates into smaller droplets and is known as secondary atomization. Droplet breakup has been studied for low Ohnesorge (<b><i>Oh < </i></b>0.1) numbers and good agreement has been seen amongst researchers. However, when it comes to cases with high the <b><i>Oh</i></b> number, i.e. atomization where the influence of viscosity is significant, very little data is available in the literature and poor agreement is seen amongst researchers. </p> <p> </p> <p>This thesis presents a complete analysis of the modes of deformation and breakup exhibited by a droplet subjected to continuous air flow. New modes of breakup have been introduced and an intermediate case with no droplet fragmentation has been discovered. Further, results are presented for droplet size-velocity distributions. In addition, Digital in-line holography (DIH) was utilized to quantify the size-velocity pdfs using a hybrid algorithm. Finally, particle image velocimetry (PIV) was employed to characterize the air flow in the unique cases where drops exhibited no breakup and cases with multiple bag formation. </p> <p> </p> <p>A droplet subjected to external aerodynamic disturbances disintegrates into smaller droplets and is known as secondary atomization. Secondary breakup finds relevance is almost every industry that utilizes sprays for their application. </p> <p> </p>

<div>Magnetic resonance imaging (MRI) has become an invaluable tool in health care. Despite its popularity, there is still an ever-increasing need for faster scans and better image quality. Multi-coil MRI, which uses multiple transmit and/or receive coils, holds the potential to address many of these MRI challenges. Multi-coil MRI systems can utilize parallel transmission (pTx) technology using multi-dimensional radio-frequency (RF) pulses for parallel excitation. The pTx platform is shown to be superior in high-field MRI. Therefore, this dissertation is focused on the RF pulse design and optimization on an MRI system with multiple transceiver coils.</div><div> </div><div>This dissertation addresses three major research topics. First, we investigate the optimization of pTx RF pulses when considering both transmitters and receivers of the MRI system. We term this framework multiple-input multiple-output (MIMO) MRI. The RF pulse design method is modeled by minimizing the excitation error while simultaneously maximizing the signal-to-noise ratio (SNR) of the reconstructed MR image. It further allows a key trade-off between the SNR and the excitation accuracy. Additionally, multiple acceleration factors, different numbers of used receive coils, maximum excitation error tolerance, and different excitation patterns are simulated and analyzed within this model. For a given excitation pattern, our method is shown to improve the SNR by 18-130% under certain acceleration schemes, as compared to conventional parallel transmission methods, while simultaneously controlling the excitation error in a desired scope.</div><div> </div><div>Second, we propose a pTx RF pulse design method that controls the peak local specific absorption rates (SARs) using a compressed set of SAR matrices. RF power, peak local SARs, excitation accuracy, and SNR are simultaneously controlled in the designed pTx RF pulses. An alternative compression method using k-means clustering algorithm is proposed for an upper-bounded estimation of peak local SARs. The performance of the pTx design method is simulated using a human head model and an eight-channel transceiver coil array. The proposed method reduces the 10-g peak local SAR by 44.6-54.2%, as compared to the unconstrained pTx approach, when it has a pre-defined lower bound of SNR and an upper bound of excitation error tolerance. The k-means clustering-based SAR compression model shows its efficiency as it generates a narrower and more accurate overestimation bound than the conventional SAR compression model.</div><div> </div><div>Finally, we propose two machine learning based pTx RF pulse design methods and test them for the ultra-fast pTx RF pulse prediction. The two methods proposed are the kernelized ridge regression (KRR) based pTx RF pulse design and the feedforward neural network (FNN) based pTx RF pulse design. These two methods learn the training pTx RF pulses from the extracted key features of their corresponding B1+ fields. These methods are compared with other supervised learning methods (nearest-neighbor methods, etc.). All learned pTx RF pulses should be reasonably SAR-efficient because training pTx RF pulses are SAR-efficient. Longer computation time and pre-scan time are the drawbacks of the current pTx approach, and we address this issue by instantaneously predicting pTx RF pulses using well-trained machine learning models.</div>