Exploration of Game Consoles as a legitimate computing platform for in-the-field biomedical data acquisition and management

Biomedical research increasingly requires for testings be conducted outside the lab, in the field such as the participant’s home or work environment. This type of research requires semi-autonomous computer systems that collect such data and send it back to the lab for processing and dissemination. A key aspect of this type of research is the selection of the required software and hardware components. These systems need to be reliable, allow considerable customizability and be readily accessible but also able to be locked down. In this paper we report a set of requirements for the hardware and software for such a system. We then utilise these requirements to evaluate the use of game consoles as a hardware platform in comparison to other hardware choices.

Mandibular cortical width measurement based on dental panoramic radiographs with computer-aided system

The paper presents a method of the determining a mandibular cortical width on dental panoramic radiographs. Cortical width of lower border of mandible may potentially be associated with recognition of osteoporosis in postmenopausal women. An algorithm to perform a semiautomatic cortical width measurement in a given region of interest was developed. The algorithm is based on separate extraction of lower and upper boundaries of cortical bone. Results of boundaries extraction performed on 34 panoramic radiographs of healthy and osteoporotic individuals are presented, together with automatic measurements of particular distances. They were compared with results of hand-made measurements done by two maxillofacial radiologists. Presented algorithm may potentially be useful for screening patients with osteoporosis.

Optimally-balanced Hash Tree Generation in Ad Hoc Networks

Ideally a hash tree is a perfect binary tree with leaves equal to power of two. Each leaf node in this type of tree can represent a mobile node in an ad hoc network. Each leaf in the tree contains hash value of mobile node’s identification (ID) and public key (PK). Such a tree can be used for authenticating PK in ad hoc networks. Most of the previous works based on hash tree assumed perfect hash tree structures, which can be used efficiently only in networks with a specific number of mobile nodes. Practically the number of mobile nodes may not be always equal to a power of two and the conventional algorithms may result in an inefficient tree structure. In this paper the issue of generating a hash tree is addressed by proposing an algorithm to generate an optimally-balanced structure for a complete hash tree. It is demonstrated through both the mathematical analysis and simulation that such a tree is optimally-balanced and can efficiently be used for public key authentication in ad hoc networks.

An Efficient Parallel IP Lookup Technique for IPv6 Routers Using Multiple Hashing with Ternary marker storage

Internet address lookup is a challenging problem because of the increasing routing table sizes, increased traffic, higher speed links, and the migration to 128 bit IPv6 addresses. Routing lookup involves computation of best matching prefix for which existing solutions scale poorly when traffic in the router increases or when employed for IPV6 address lookup. Our paper describes a novel approach which employs multiple hashing on reduced number of hash tables on which ternary search on levels is applied in parallel. This scheme handles large number of prefixes generated by controlled prefix expansion by reducing collision and distributing load fairly in the hash buckets thus providing faster worst case and average case lookups. The approach we describe is fast, simple, scalable, parallelizable, and flexible.

Novel Model for the Computation of Linguistic Hedges in Database Queries

Most query languages are designed to retrieve information from databases containing precise and certain data using precisely specified commands. Due to the advancements in various kinds of data repositories in the recent years, there is a steep increase in complex queries. Most of the complex Queries are uncertain and vague. The existing Structured Query Language exhibits its inefficiency in handling these complex Queries. This paper proposes a model to handle the complexities by using fuzzy set theory. In this model, the Fuzzy Query with linguistic hedges is converted into Crisp Query, by deploying an application layer over the Structured Query Language.

System for Detection of Vital Signals with an Embedded System

Rapid advancement in the field of Embedded Systems and Wireless communications has permitted development of Revolutionary Medical Monitoring Systems and thus improving the lifestyle of patients. The system captures and analyzes the ECG signals in real time through a low cost embedded development board. The system can detect cardiac abnormalities with high precision. One of the objectives at the time of building the proposed system has been to optimize the resources, memory size and communication costs.

Copyright Transfer Form

Authors of accepted papers must fill this copyright transfer form and email/fax to the edtior of the AJICT before publication can take place.

Authors Guidelines

Authors wishing to submit papers to the AJICT should use this guideline and format for their papers.

Single Channel Wireless EEG: Proposed Application in Train Drivers

Electroencephalography (EEG) can be used as an indicator of fatigue. Several studies have shown that slow wave brain activities, delta (0-4 Hz) and theta (4- 8 Hz), increase as an individual becomes fatigued, while the fast brain activities, alpha (8-13 Hz) and beta (13-35 Hz), decrease. However, an EEG is a complex piece of equipment that is generally used in laboratory based studies. In order to develop a fatigue countermeasure device for train drivers using EEG, there is a need for a simple and wireless EEG monitor. This paper explains the development of a single channel wireless EEG device.

On the Combination of Logistic Regression and Local Probability Estimates

In this paper we give a survey of the combination of classifiers. We briefly describe basic principles of machine learning and the problem of classifier construction and review several approaches to generate different classifiers as well as established methods to combine different classifiers. Then, we introduce our novel approach to assess the appropriateness of different classifiers based on their characteristics for each test point individually