Biomedical subjects

Evangelos Milios

Publications and source records attributed to Evangelos Milios.

5 recordsLinked to original sources

PubMed2005

Acoustic positioning using multiple microphone arrays.

Passive acoustic techniques are presented to solve the localization problem of a sound source in three-dimensional space using off-the-shelf hardware. Multiple microphone arrays are employed, which operate independently, in estimating the direction of arrival of sound, or, equivalently, a direction vector from the array's geometric center towards the source. Direction vectors and array centers are communicated to a central processor, where the source is localized by finding the intersection of the direction lines defined by the direction vectors and the associated array centers. The performance of the method in the air is demonstrated experimentally and compared with a state-of-the-art method that requires centralized digitization of the signals from the microphones of all the arrays.

Acoustics↗

PubMed2004

Knowledge management in pediatric pain: mapping on-line expert discussions to medical literature.

Clinical decision-making can be vastly improved with the availability of the right medical knowledge at the right time. This concept paper presents a knowledge management re-search program to (a) identify, capture and organize the tacit knowledge inherent within on-line problem-solving discussions between pediatric pain practitioners; (b) establish linkages between topic-specific pediatric pain discussions and corresponding published medical literature on children's pain available at PubMed--i.e. linking tacit expert knowledge to explicit medical literature; and (c) make these knowledge re-sources available to pediatric pain practitioners via the WWW for timely access to various modalities of clinical knowledge.

Artificial Intelligence↗

PubMed2003

Sonification of range information for 3-D space perception.

We present a device that allows three-dimensional (3-D) space perception by sonification of range information obtained via a point laser range sensor. The laser range sensor is worn by a blindfolded user, who scans space by pointing the laser beam in different directions. The resulting stream of range measurements is then converted to an auditory signal whose frequency or amplitude varies with the range. Our device differs from existing navigation aids for the visually impaired. Such devices use sonar ranging whose primary purpose is to detect obstacles for navigation, a task to which sonar is well suited due to its wide beam width. In contrast, the purpose of our device is to allow users to perceive the details of 3-D space that surrounds them, a task to which sonar is ill suited, due to artifacts generated by multiple reflections and due to its limited range. Preliminary trials demonstrate that the user is able to easily and accurately detect corners and depth discontinuities and to perceive the size of the surrounding space.

Acoustics↗

PubMed2003

Active stereo sound localization.

Estimating the direction of arrival of sound in three-dimensional space is typically performed by generalized time-delay processing on a set of signals from a fixed array of omnidirectional microphones. This requires specialized multichannel A/D hardware, and careful arrangement of the microphones into an array. This work is motivated by the desire to instead only use standard two-channel audio A/D hardware and portable equipment. To estimate direction of arrival of persistent sound, the position of the microphones is made variable by mounting them on one or more computer-controlled pan-and-tilt units. In this paper, we describe the signal processing and control algorithm of a device with two omnidirectional microphones on a fixed baseline and two rotational degrees of freedom. Experimental results with real data are reported with both impulsive and speech sounds in an untreated, normally reverberant indoor environment.

Algorithms↗

PubMed2002

Filtering for medical news items using a machine learning approach.

We address the problem of filtering medical news articles for targeted audiences. The approach is based on terms and one of the difficulties is extracting a feature set appropriate for the domain. This paper addresses the medical news-filtering problem using a machine learning approach. We describe the application of two supervised machine learning techniques, Decision Trees and Naïve Bayes, to automatically construct classifiers on the basis of a training set, in which news articles have been pre-classified by a medical expert and four other human readers. The goal is to classify the news articles into three groups: non-medical, medical intended for experts, and medical intended for other readers. While the general accuracy of the machine learning approach is around 78%, the accuracy of distinguishing non-medical articles from medical ones is shown to be 92%.

Artificial Intelligence↗