As feature sizes and operating voltages decrease, single-event transients from particle strikes in logic circuits become more probable. Much literature is available on the effects of these events in memory, but with increasing clock speeds, combinational logic has also been shown to be at risk. In this work, several combinational circuits are selected and simulated, taking into account gate library and layout information, in order to characterize the effects of particle strikes which upset single nodes as well as multiple, physically adjacent nodes. It is shown that traditional reliability tests which simulate a single fault are not sufficient multiple faults stemming from a single strike occur and are more complex. However, multiple faults do not always translate to additional errors in the output logical reconvergence limits the effect of faults within a circuit. In order to properly understand reliability in circuit design, analysis of multiple faults should be taken into account.

Sensitive label-free optical biosensors based on grating-coupled porous silicon (PSi) waveguides are demonstrated for biosensing applications. This is the first time that the benefits of both PSi and diffraction gratings have been combined for the detection of small molecules. The large surface area and strong field confinement in the waveguide region where biomolecules are immobilized make it possible to detect biomolecule interactions with high sensitivity. Diffraction gratings provide a simple, convenient, and effective means of coupling light into the waveguide for DNA hybridization experiments. Through both experiments and theoretical calculations, it was shown that an all-PSi grating-coupled waveguide biosensor exhibits a more than 30-fold larger signal response for small molecule detection compared to planar waveguide sensors. Kinetics parameters of different sized molecules in nanoscale pores were also determined for the first time using PDMS flow cell integrated PSi grating-coupled waveguides.

As transistor density increases with each new CMOS technology node, the probability of a single ion causing a single-event transient in a circuit or inducing charge sharing among transistors increases. These transients can lead to single-event upsets that can cause a circuit or system to fail. Therefore, it is important to understand the characteristics of single-event transients at each new technology node and the resulting impacts on circuit designs. This dissertation uses both three-dimensional mixed-mode technology-computer-aided design simulations and experimental analysis at the 65 nm, 90 nm, and 130 nm technology nodes to fully characterize the mechanisms that affect single-event transients in sub-100 nm bulk CMOS technologies. Investigations show that the design parameter of n-well contact area influences the pulse width of single-event transients by controlling the degree of parasitic bipolar junction transistor amplification in pMOS transistors. Also the prevalence of charge sharing in sub-100 nm bulk CMO technologies has led to a new single-event mechanism called pulse quenching that can shorten or eliminate single-event transients. Furthermore, pulse quenching can lead to a new type of single-event transient called a double-pulse-single-event transient.

Levodopa remains the most effective medication for Parkinsons disease. Prolonged use of levodopa leads to a side effect called dyskinesia that is characterized by abnormal involuntary movements. The assessment of dyskinesia severity is essential to develop better therapies to treat it. Qualitative assessments using rating scales/questionnaires are rater-dependent and lack rating resolution. In this work, we propose a low-cost, patient-friendly, video-based technique that quantifies dyskinesia severity with a single score incorporating most of its attributes. This dissertation makes three main contributions to the field. The first contribution is the development of a severity score called SVS that is based on the covariance of points that were automatically tracked in video sequences using non-rigid image registration. The second contribution is the validation of SVS, which was done by comparing it to severity rankings by trained neurologists based on the clinical definition of dyskinesia. Results show a moderate but statistically significant correlation between automatic and manual rankings. The third contribution is a modified severity score that uses frequency dispersion parameters to differentiate between choreic and dystonic movement. Results show that adding these frequency dispersion parameters improves the correlation between manual and automatic scores.

Current deep sub-micron technologies are particularly susceptible to single events. The challenge derives from a conglomeration of effects that affect circuits radiation response. For instance, decreasing device size has led to decreased storage charge and increased operating speeds. Decreased storage charge contributes to single event sensitivity since the charge deposited by the ionized particle is now more on par with the storage node charge. The increased operating speeds increase circuit sensitivity since they are now comparable to the speed of a single event transient. Single events are expected to dominate other reliability concerns in deep sub-micrometer devices due to decreased transistor currents and nodal capacitances. Thus, it is vital to understand and quantify the impact of contributing mechanisms. In addition to the challenges presented by radiation, transistor and chip design also faces a challenge from the variations that are inherent to the manufacturing process. During chip fabrication, extreme measures are taken to ensure precision. However, variations in lithography, random dopant fluctuations, gate depletion, surface state charge, and line-edge roughness cause changes in individual transistor behavior and therefore, changes in the behavior-describing transistor parameters. The non-standard behavior then affects circuit performance and therefore impacts single event response. It is anticipated that process variations will substantially increase with shrinking device sizes. Consequently, the potential impact of process variations on SE circuit response is significant. In order to accurately predict the single-event response of any circuit, it is necessary to identify and quantify the impact of the specific process variations. This work correlates the shifts in radiation response to specific device and process-parameter variations by quantifying the impact of process variability on the range of single-event upset (SEU) critical charge. Monte-Carlo simulations are leveraged to assess the impact of process variations on SEU response. This dissertation analyzes the impact of process variations on the single-event upset response of sub-100nm CMOS memory storage circuits.

Single Events Upsets (SEU) have long been a concern of the space and aviation fields. An SEU occurs when the logic state of a data-storage element in an integrated circuit changes because of the charge generated by a single ionizing particle. In the past, the minimum amount of deposited charge required from an event to cause an upset, was relatively large compared to present technologies. This kept the focus of the industry on heavy ions that had relatively high kinetic energy. As devices continue to scale down, the value for critical charge will also decrease. This will require attention to particles that could previously be ignored. Even terrestrial circuits will be exposed to particles that now have the ability to upset a cell. Current transients resulting from a single-event strike can be generated from 3D technology computer aided design (TCAD) device simulations to aid in determining an accurate SEU rate. These device level transient simulations can be time intensive, requiring significant overhead to calibrate a device response. This work describes a physics-based approach to estimating prompt current transient pulse shapes efficiently and accurately using a closed-form equation. This method of pulse shape calibration is accomplished without running time-consuming finite element transient simulations and only requires minimal device geometry information. Creating these pulses efficiently with this method enables designers to estimate transient responses quickly, allowing for seamless integration into higher level simulations.

Single event latchup (SEL) is a serious reliability concern for CMOS integrated circuits (ICs), and can be especially problematic in the space radiation environment. It can occur because of a parasitic pnpn circuit inherent in CMOS, which, if activated, introduces a low-impedance path across the power supply lines. This leads to a large current that can cause thermal damage in the IC. This dissertation describes an experimental study that focuses on SEL hardening strategies, triggering mechanisms, and testing considerations. The frontside single-photon absorption laser test method is used extensively. The backside two-photon absorption laser, broadbeam heavy ion, proton, and neutron test methods are also used. The majority of the work is done using custom test structures fabricated in a 180 nm bulk CMOS technology. Data are also presented on a commercial power device controller and 130 nm technology SRAMs. We evaluate the effectiveness of various SEL hardening strategies, including thick-film silicon-on-insulator (SOI), triple well, and guard rings. Although SOI technology is widely reported to be immune to SEL, conventional pnpn latchup can occur and has been observed in non-dielectrically isolated SOI processes. The introduction of triple well is shown to be an effective zero-area-penalty hardening strategy, although it does not result in the SEL immunity that was achieved when guard rings were introduced. After triggering latchup with the pulsed laser in a given pnpn region, latchup was observed to spread and infect many adjacent pnpn regions. The physical mechanisms of this spreading are discussed, along with the implications for device characterization. Lasers are used in other experiments to map the shapes and positions of SEL sensitive regions, and show that the position of maximum sensitivity is not centered on a pnpn region, but between two neighboring pnpn regions, due to synergistic triggering. Finally, the SEL sensitivity maps demonstrate that laser light reflected from metal lines toward the silicon can contribute to the single event effect response in some cases, for both backside- and frontside-incident laser tests.

In this thesis, single event transient (SET) effects in sub-70 nm bulk and SOI FinFETs are investigated through topside and backside laser and heavy ion irradiations. Pulsed laser induced current transients in bulk FinFETs show distinct signatures for charge collection from drift and diffusion, demonstrating the contribution of charge generated in the substrate to the charge collection process. This result was validated through heavy ion testing on advanced bulk FinFETs with two different junction contact schemes (dumbbell and saddle). The drain region dominates the charge collection response of bulk FinFETs, with the maximum charge collected in devices with dumbbell contacts. Recorded current transients in the drain and source terminals of bulk and SOI FinFETs indicated that shunt effect plays a key role in the charge collection process of these highly scaled structure. Top-side laser and heavy ion results on bulk and SOI FinFETs demonstrate a significantly higher tolerance of SOI devices to SEEs, thanks to the buried oxide (BOX) layer that reduces their collection volume to the fins.

Cognitive robotics is one of the branches of robotics that is concerned with the design and implementation of robots that will accomplish cognitive tasks such as perceiving the dynamic world around them and making acceptable decisions in real-time by imitating humans. A flexible and natural way to make robots learn new skills is to implement them with the ability to learn by imitation accompanied by effective affordance learning, which would be a feasible method to automate the tedious manual programming of robotic tasks. Imitation learning offers promising directions for gaining insight into faster affordance learning with which robots would have a good likelihood of learning complex behaviors from a small set of experiences much like human beings, and ultimately develop autonomous perceptual-motor control mechanisms. Thus, the goal of the dissertation is to develop a flexible mechanism for the robot so that it can learn high level motor tasks by experiencing action outcomes via its own sensors and forming an action-perception coupling similar to what happens in human beings. In order to learn basic behaviors, the robot goes through certain experimental stages involving self-exploration, affordance learning, and imitation learning which are parallel to the developmental pattern found in babies developing new motor skills. In this dissertation, using biologically inspired cognitive mechanisms such as affordance relations and imitation learning on a humanoid robot, we propose a new developmental approach for cognitive robots to learn novel motor behaviors in virtual and real environments.

Cancers in the head and neck region account for approximately 3 percent of all cancers in the United States, as it is reported by the American Cancer Society. Depending on the location and stage of the cancers, surgery, chemotherapy and radiation therapy are considered the major treatment options. Since the past decade intensity-modulated radiation therapy (IMRT) has become the state of the art in head and neck radiation therapy, segmentation of head and neck structures to be treated, including the level II, III, and IV lymph nodes, as well as structures to be spared, including the thyroid glands and the parotid glands, from the diagnostic computed tomography (CT) images is of great importance for treatment planning. In order to reduce the time required to manually segment these structures, a set of innovative approaches are proposed to automate the process: An active shape model (ASM) is constructed to segment the level II, III, and IV lymph node regions; a multiple-atlas-based approach is implemented to segment the thyroid gland; a constrained ASM with landmark uncertainty is used to segment the bilateral parotid glands. Both qualitative and quantitative results have shown that the automatically generated segmentations are of high precision, and the proposed approaches have the potential to reduce the delineation efforts required for clinical IMRT treatment planning.