Real-time mobile detection of drug use with wearable biosensors: a pilot study.

While reliable detection of illicit drug use is paramount to the field of addiction, current methods involving self-report and urine drug screens have substantial limitations that hinder their utility. Wearable biosensors may fill a void by providing valuable objective data regarding the timing and contexts of drug use. This is a preliminary observational study of four emergency department patients receiving parenteral opioids and one individual using cocaine in a natural environment. A portable biosensor was placed on the inner wrist of each subject, to continuously measure electrodermal activity (EDA), skin temperature, and acceleration. Data were continuously recorded for at least 5 min prior to drug administration, during administration, and for at least 30 min afterward. Overall trends in biophysiometric parameters were assessed. Injection of opioids and cocaine use were associated with rises in EDA. Cocaine injection was also associated with a decrease in skin temperature. Opioid tolerance appeared to be associated with a blunted physiologic response as measured by the biosensor. Laterality may be an important factor, as magnitude of response varied between dominant and nondominant wrists in a single patient with bilateral wrist measurements. Changes in EDA and skin temperature are temporally associated with intravenous administration of opioids and cocaine; the intensity of response, however, may vary depending on history and extent of prior use.

Autonomic changes with seizures correlate with postictal EEG suppression.

OBJECTIVE: Sudden unexpected death in epilepsy (SUDEP) poses a poorly understood but considerable risk to people with uncontrolled epilepsy. There is controversy regarding the significance of postictal generalized EEG suppression as a biomarker for SUDEP risk, and it remains unknown whether postictal EEG suppression has a neurologic correlate. Here, we examined the profile of autonomic alterations accompanying seizures with a wrist-worn biosensor and explored the relationship between autonomic dysregulation and postictal EEG suppression. METHODS: We used custom-built wrist-worn sensors to continuously record the sympathetically mediated electrodermal activity (EDA) of patients with refractory epilepsy admitted to the long-term video-EEG monitoring unit. Parasympathetic-modulated high-frequency (HF) power of heart rate variability was measured from concurrent EKG recordings. RESULTS: A total of 34 seizures comprising 22 complex partial and 12 tonic-clonic seizures from 11 patients were analyzed. The postictal period was characterized by a surge in EDA and heightened heart rate coinciding with persistent suppression of HF power. An increase in the EDA response amplitude correlated with an increase in the duration of EEG suppression (r = 0.81, p = 0.003). Decreased HF power correlated with an increase in the duration of EEG suppression (r = -0.87, p = 0.002). CONCLUSION: The magnitude of both sympathetic activation and parasympathetic suppression increases with duration of EEG suppression after tonic-clonic seizures. These results provide autonomic correlates of postictal EEG suppression and highlight a critical window of postictal autonomic dysregulation that may be relevant in the pathogenesis of SUDEP.

Cluster-based probability model and its application to image and texture processing.

We develop, analyze, and apply a specific form of mixture modeling for density estimation within the context of image and texture processing. The technique captures much of the higher order, nonlinear statistical relationships present among vector elements by combining aspects of kernel estimation and cluster analysis. Experimental results are presented in the following applications: image restoration, image and texture compression, and texture classification.

Video orbits of the projective group a simple approach to featureless estimation of parameters.

We present direct featureless methods for estimating the eight parameters of an "exact" projective (homographic) coordinate transformation to register pairs of images, together with the application of seamlessly combining a plurality of images of the same scene, resulting in a single image (or new image sequence) of greater resolution or spatial extent. The approach is "exact" for two cases of static scenes: (1) images taken from the same location of an arbitrary three-dimensional (3-D) scene, with a camera that is free to pan, tilt, rotate about its optical axis, and zoom, or (2) images of a flat scene taken from arbitrary locations. The featureless projective approach generalizes interframe camera motion estimation methods that have previously used a camera model (which lacks the degrees of freedom to "exactly" characterize such phenomena as camera pan and tilt) and/or which have relied upon finding points of correspondence between the image frames. The featureless projective approach, which operates directly on the image pixels, is shown to be superior in accuracy and the ability to enhance the resolution. The proposed methods work well on image data collected from both good-quality and poor-quality video under a wide variety of conditions (sunny, cloudy, day, night). These new fully automatic methods are also shown to be robust to deviations from the assumptions of static scene and no parallax.

On the efficiency of the orthogonal least squares training method for radial basis function networks.

The efficiency of the orthogonal least squares (OLS) method for training approximation networks is examined using the criterion of energy compaction. We show that the selection of basis vectors produced by the procedure is not the most compact when the approximation is performed using a nonorthogonal basis. Hence, the algorithm does not produce the smallest possible networks for a given approximation error. Specific examples are given using the Gaussian radial basis functions type of approximation networks.

M-lattice: from morphogenesis to image processing.

The paper is based on reaction-diffusion, a nonlinear mechanism first proposed by Turing in 1952 to account for morphogenesis, the formation of shape and pattern in nature. One of the key limitations of reaction-diffusion systems is that they are generally unbounded, making them awkward for digital image processing. In this paper we introduce the "M-lattice", a system that preserves the pattern-formation properties of reaction-diffusion and is bounded. On the theoretical front, we establish how the M-lattice is closely related to the analog Hopfield network and the cellular neural network, but has more flexibility in how its variables interact. Like many "neurally inspired" systems, the bounded M-lattice also enables computer or analog VLSI implementations to simulate a variety of partial and ordinary differential equations. On the practical front, we demonstrate two novel applications of reaction-diffusion formulated as the new M-lattice. These are adaptive filtering, applied to the restoration and enhancement of fingerprint images, and nonlinear programming, applied to image halftoning in both "faithful" and "special effects" styles.

Finding perceptually dominant orientations in natural textures.

An algorithm for detecting orientation in texture is developed and compared with results of humans detecting orientation in the same textures. The algorithm is based on the steerable filters of Freeman and Adelson (IEEE Trans. PAMI 13, 891-906, 1991), orientation-selective filters derived from derivatives of Gaussians. The filters are applied over multiple scales and their outputs non-linearly contrast-normalized. The data for humans were collected from forty subjects who were asked to identify 'the minimum number of dominant orientations' they perceived, and the 'strength' with which they perceived each orientation. Test data consisted of 111 grey-level images of natural textures taken from the Brodatz album, a standard collection used in computer vision and image processing. Results show that the computer and humans chose at least one of the same dominant orientations on 95 of the natural textures. Of these textures, 74 were also in 100% agreement on the location of all the dominant orientations chosen by both humans and computer. Disagreements are analyzed and possible causes are discussed. Some apparent limitations in the current filter shapes and sizes are illustrated, as well as some (surprisingly small) effects believed to be caused by semantic recognition and gestalt grouping.