Using results of eye movement signal analysis in the neural network recognition of otoneurological patients

We extracted a collection of eye movement signals employed for almost two decades in clinical otoneurological tests at a balance laboratory. During those years we designed and programmed signal analysis methods to analyse their features in detail and to compute medically important attributes. In the present study, using such attributes and their results computed we classified test cases into groups of healthy subjects and patients with multilayer perceptron neural networks. Classification succeeded in total accuracies from 60% to 90% depending on the type of eye movements, which were saccades, nystagmus, sinusoidal movements and vestibulo-ocular reflex stimulated in two different ways; these are the chief eye movement tests applied in otoneurology.     

How to construct small probabilistic roadmaps with a good coverage?

We investigate how we can construct small probabilistic roadmaps in a reasonable time while still keeping a good coverage and connectivity. We propose a new neighborhood-based method that can reduce the size of the roadmaps by filtering out unnecessary nodes. We then experimentally test it against a basic probabilistic roadmap planner and a visibility-based planner. We use both a uniform sampling and a bridge test sampling in our tests. The results show that the neighborhood-based method can reduce the number of nodes considerably. The neighborhood-based method is simple to implement, it works well with a uniform sampling, and it does not need any additional parameters when compared with the basic planner.     

Use of adipose stem cells and polylactide discs for tissue engineering of the temporomandibular joint disc

There is currently no suitable replacement for damaged temporomandibular joint (TMJ) discs after discectomy. In the present study, we fabricated bilayer biodegradable polylactide (PLA) discs comprising a non-woven mat of poly(L/D)lactide (P(L/D)LA) 96/4 and a P(L/DL)LA 70/30 membrane plate. The PLA disc was examined in combination with adipose stem cells (ASCs) for tissue engineering of the fibrocartilaginous TMJ disc in vitro. ASCs were cultured in parallel in control and chondrogenic medium for a maximum of six weeks. Relative expression of the genes, aggrecan, type I collagen and type II collagen present in the TMJ disc extracellular matrix increased in the ASC-seeded PLA discs in the chondrogenic medium. The hypertrophic marker, type X collagen, was moderately induced. Alcian blue staining showed accumulation of sulphated glycosaminoglycans. ASC differentiation in the PLA discs was close to that observed in pellet cultures. Comparison of the mRNA levels revealed that the degree of ASC differentiation was lower than that in TMJ disc-derived cells and tissue. The pellet format supported the phenotype of the TMJ disc-derived cells under chondrogenic conditions and also enhanced their hyalinization potential, which is considered part of the TMJ disc degeneration process. Accordingly, the combination of ASCs and PLA discs has potential for the development of a tissue-engineered TMJ disc replacement.     

Neural network modelling of word production in Finnish: coding semantic and non-semantic features

The objective of our research is to computationally model word production and its disorders by means of artificial neural networks. In the current study we develop and analyze an algorithm that generates a distributed semantic coding from a given semantic tree-structure classification of words. With the algorithm it is possible to generate semantic representations that are compact and easy to modify. This renders the coding method suitable for our multilayer perceptron-based neural network model of word production. The model is shown to be able to account for a variety of performance patterns observed in four Finnish aphasia patients suffering from word-finding difficulties.     

Kernel selection in multi-class support vector machines and its consequence to the number of ties in majority voting method

Support vector machines are a relatively new classification method which has nowadays established a firm foothold in the area of machine learning. It has been applied to numerous targets of applications. Automated taxa identification of benthic macroinvertebrates has got generally very little attention and especially using a support vector machine in it. In this paper we investigate how the changing of a kernel function in an SVM classifier effects classification results. A novel question is how the changing of a kernel function effects the number of ties in a majority voting method when we are dealing with a multi-class case. We repeated the classification tests with two different feature sets. Using SVM, we present accurate classification results proposing that SVM suits well to the automated taxa identification of benthic macroinvertebrates. We also present that the selection of a kernel has a great effect on the number of ties.     

Segmentation of digital signals based on estimated compressionratio

The authors have studied the problem of approximating a digital signal with a suitable continuous broken line. They use the approximative broken line for further analysis of the signal as detection of peaks, waves, and other structural features. They can also save considerable amount of storage space with an approximation that does not lose too much significant information about the original signal. The authors' work is based on examining different distance metrics and different segmentation methods with respect to the remaining residual error in the resulting approximation. The aim of the work has been to develop a method that can perform segmentation with an acceptable amount of residual error without a need to define a large set of parameters that control the segmentation process. The authors' contribution is to examine the effect of the estimated compression ratio of the resulting approximation and finding an estimate of this compression ratio. They first define a target in the form of a compression ratio of the resulting approximation and then by applying their method, try to find a suitable threshold parameter to achieve this target. The authors have tested their method with electrocardiogram (EGG) signals and the compression ratio of the approximation has been found to be a suitable target to control the segmentation process     

1.7 Volt, DC to 15 GHz Differential Amplifier with Constant Group Delay in InP-HBT Technology

10 dB gain–, 15 GHz-Bandwidth amplifier has been designed andfabricated in InP-HBT technology. Operation of the amplifier was achieved at1.7 V at total current consumption of 40 mA. The amplifier designed to havea Bessel-transfer-function has almost constant group delay up to 30 GHz.Pulse- and eye-diagram- measurements have been performed to verify largesignal operation. At 3.3 V, 24 dB gain and 10 GHz bandwidth was achieved.The 1dB compression point at 1.7 V supply voltage is measured at –10dBm output power.     

National research contributions: A case study on Finnish biomedical research

The long-term influence and contribution of research can be evaluated relatively reliably by bibliometric citation analysis. Previously, productivity of nations has been estimated by using either the number of published articles or journal impact factors and/or citation data. These studies show certain trends, but detailed analysis is not possible due to the assumption that all articles in a journal were equally cited. Here we describe the first comprehensive, longterm, nationwide analysis of scientific performance. We studied the lifetime research output of 748 Finnish principal investigators in biomedicine during the years 1966–2000, analysed national trends, and made a comparison with international research production. Our results indicate that analyses of the scientific contribution of persons, disciplines, or nations should be based on actual publication and citation counts rather than on derived information like impact factors. 51% of the principal investigators have published altogether 75% of the articles; however, the whole scientific community has contributed to the growth of biomedical research in Finland since the Second World War.     

Vestibulo-ocular reflex function as measured with the head autorotation test

The vestibulo-ocular reflex (VOR) of 125 healthy subjects was examined over the frequency range of 0.5-5 Hz with the head autorotation test (HART). During the HART the subjects fixated at a steady target while moving their heads horizontally from side to side with increasing frequencies according to auditory signals. The gain was determined as the ratio between the amplitude of the eye and head movements in five frequency bands between 0.5 and 5 Hz. The phase difference between the eye and head movements was determined in both degrees and milliseconds. The ability to reach high-frequency bands was evaluated. The mean gain was close to unity up 5 Hz, when it decreased to 0.91. The mean phase difference showed a lead of approximately 5 degrees at frequencies below 2 Hz, and at frequencies above 2 Hz there was no phase difference within the resolution of the test. The frequency band of 5 Hz was reached by 78% of the subjects, and that of 4 Hz was reached by 94% of the subjects. In summary, the HART is a new approach with which to study VOR function and determine accurately the VOR for healthy subjects. The normal upper frequency limit is 4 Hz in the HART.     

Nonsteroidal anti-inflammatory drug-associated upper gastrointestinal lesions in rheumatoid arthritis patients. Relationships to gastric histology, Helicobacter pylori infection, and other risk factors for peptic ulcer

BACKGROUND: Nonsteroidal anti-inflammatory drugs (NSAIDs) are risk factors for peptic ulcer in rheumatoid arthritis (RA) patients, but the contribution of reactive gastritis, concomitant Helicobacter pylori infection, or RA activity to NSAID ulcer pathogenesis is unknown. METHODS: Ninety-six RA patients taking NSAIDs and dyspeptic sex- and age-matched control patients without NSAID use or an RA diagnosis were enrolled in the study. RESULTS: Gastric ulcer (GU) was detected in 29 (30%) RA patients and 3 control patients (P < 0.001). Sixteen RA patients and no control patient had an H. pylori-negative GU. The GUs of the RA patients were mainly located in the prepyloric region (28%) and antrum (62%). Nine of the 29 RA patients (31%) with GU had more than 1 ulcer. Erosive gastropathy was detected in 34 (71% H. pylori-negative) RA patients and in 13 (62% H. pylori-negative) control subjects (P < 0.001). Chronic gastritis was observed in 65 RA patients (48% H. pylori-negative) and in 58 control subjects (43% H. pylori-negative) (NS). whereas reactive gastritis was found in only 2 RA patients and in none of the controls. Corticosteroid use was the only independent risk factor for GU: odds ratio was 6.8 (95% confidence interval, 1.3-36.0). The prevalences of duodenal ulcer or esophagitis were not increased in RA patients. CONCLUSIONS: RA patients using NSAIDs continuously are at a greatly increased risk of developing both H. pylori-negative and -positive GUs, and corticosteroid use is an independent risk factor for ulcer development. Most RA patients have chronic gastritis, whereas reactive gastritis is rarely associated with continuous NSAID use in RA patients.