Wyniki wyszukiwania - BazTech

1

Human activity detection based on the iBeacon technology

Lewandowski M., Orczyk T., Płaczek B.

Journal of Medical Informatics & Technologies

|

2016

|

Vol. 25

38--45

EN

Paper presents a new method of patient activity monitoring, by using modern ADL (Activities of Daily Living) techniques. Proposed method utilizes energy efficient Bluetooth iBeacon BLE (Bluetooth Low Energy) modules, developed by Apple. Main advantage of this technology is the ability to detect neighboring devices, which belong to the same device family. Proposed method is based on observing changes of received signal strength indicator (RSSI) in the time domain. The RSSI analysis is performed in order to asses a human activity. Such observation may be particularly useful for monitoring consciousness of elder people, where reaction time of emergency rescuers and appropriate rescue operations may save the human lives.

2

Feature projection k-NN classifier model for imbalanced and incomplete medical data

Porwik P., Orczyk T., Lewandowski M., Cholewa M.

Biocybernetics and Biomedical Engineering

|

2016

|

Vol. 36, no. 4

644--656

EN

Many datasets, especially various historical medical data are incomplete. Various qualities of data can significantly hamper medical diagnosis and are bottlenecks of medical support systems. Nowadays, such systems are often used in medical diagnosis. Even great number of data can be unsuitable when data is imbalanced, missing or corrupted. In some cases these troubles can be overcome by machine learning algorithms designed for predictive modeling. Proposed approach was tested on real medical data and some benchmarks dataset form UCI repository. The liver fibrosis disease from a medical point of view is difficult to treatment and has a significant social and economic impact. Stages of liver fibrosis are diagnosed by clinical observation and evaluations, coupled with a so-called METAVIR rating scale. However, these methods may be insufficient, especially in the recognition of phase of the disease. This paper describes a newly developed algorithm to non-invasive fibrosis stage recognition using machine learning methods – a classification model based on feature projection k-NN classifier. This solution allows extracting data characteristics from the historical data which may be incomplete and may contain imbalance (unequal) sets of patients. Proposed novel solution is based on peripheral blood analysis without using any specialized biomarkers, and can be successfully included to medical diagnosis support systems and might be a powerful tool for effective estimation of liver fibrosis stages.

3

Instance based kNN modification for classification of medical data

Orczyk T., Porwik P., Lewandowski M., Cholewa M.

Journal of Medical Informatics & Technologies

|

2015

|

Vol. 24

99--106

EN

Paper describes a novel modification to a well known kNN algorithm, which enables using it for medical data, which often is a class-imbalanced data with randomly missing values. Paper presents the modified algorithm details, experiment setup, results obtained on a cross validated classification of a benchmark database with randomly removed values (missing data) and records (class imbalance), and their comparison with results of the state of the art classification algorithms.

4

Image analysis of immunohistochemical stains for detection of parathyroid disease

Płaczek B., Orczyk T., Bułdak R., Michalski M., Segiet O., Polaniak R.

Journal of Medical Informatics & Technologies

|

2015

|

Vol. 24

141--146

EN

In this paper a method is introduced which enables automatic detection of parathyroid hyperplasia and parathyroid adenoma on the basis of immunohistochemical angiogenesis markers expression in micrographs. The proposed method uses digital image processing techniques and classification algorithms to detect diseased tissue. The disease detection is performed by classification of normalized color intensity histograms. Accuracy of this method was evaluated by using micrographs of parathyroid tissue sections obtained from patients that have undertaken surgery due to primary hyperparathyroidism. Use of different color models, various classifiers, and immunohistochemical markers was considered during the experiments. The experimental results show that the introduced method enables accurate detection of parathyroid disease. The most promising results were obtained for k-nearest neighbor and neural network classifiers.

5

Method of signature recognition with the use of the complex features

Doroz R., Pałys M., Orczyk T., Safaverdi H.

Journal of Medical Informatics & Technologies

|

2014

|

Vol. 23

155--162

EN

In this paper a new method of handwritten signatures verification has been proposed. This method, for each signature, creates complex features which are describing this signature. These features are based on dependencies analysis between dynamic features registered by tablets. These complex features are then used to create vectors describing the signature. Elements of these vectors are calculated using measures proposed in this work. The similarity between signatures is assessed by determining the similarity of vectors in the compared signatures. Research, whose results will be presented in the further part of this work, have shown a high efficiency of verification using proposed method.

6

Medical diagnosis support system based on the ensemble of single-parameter classifiers

Orczyk T., Porwik P., Bernaś M.

Journal of Medical Informatics & Technologies

|

2014

|

Vol. 23

173--179

EN

This paper presents a medical diagnosis support system based on an ensemble of single parameter k–NN classifiers [1]. System was verified on a database containing real blood test results of diagnosed patients with a liver fibrosis. This dataset contains problems typical to a real medical data – especially missing values. Paper also describes the process of selecting a subset of parameters used for further evaluation (feature selection/elimination algorithm). Complete database contains many parameters, but not all are important for diagnosis, thus eliminating them is an important step. A comparison of proposed method of classification and feature selection with methods known from literature has also been presented.

7

Influence of missing data imputation method on the classification accuracy of the medical data

Orczyk T., Porwik P.

Journal of Medical Informatics & Technologies

|

2013

|

Vol. 22

111--116

EN

Aim of this study is to show the dangers of filling missing data - particularly medical data. Because there are many dedicated medical expert systems and medical decision support systems, a special attention must be paid on the construction of classifiers. Medical data are almost never complete, and completion of the missing data requires a special care. The safest approach of dealing with missing data would be removing records with missing parameters and/or removing parameters that are missing in the records. Unfortunately reducing data set that is already very small is not always an option. Dangers coming out from data imputation are shown in the article, which presents the influence of selected missing data filling algorithms on the classification accuracy.

8

Classification techniques for non-invasive recognition of liver fibrosis stage

Krawczyk B., Woźniak M., Orczyk T., Porwik P., Musialik J., Błońska-Fajfrowska B.

Journal of Medical Informatics & Technologies

|

2012

|

Vol. 20

121--127

EN

Contemporary medicine should provide high quality diagnostic services while at the same time remaining as comfortable as possible for a patient. Therefore novel non-invasive disease recognition methods are becoming one of the key issues in the health services domain. Analysis of data from such examinations opens an interdisciplinary bridge between the medical research and artificial intelligence. The paper presents application of machine learning techniques to biomedical data coming from indirect examination method of the liver fibrosis stage. Presented approach is based on a common set of non-invasive blood test results. The performance of four different compound machine learning algorithms, namely Bagging, Boosting, Random Forest and Random Subspaces, is examined and grid search method is used to find the best setting of their parameters. Extensive experimental investigations, carried out on a dataset collected by authors, show that automatic methods achieve a satisfactory level of the fibrosis level recognition and may be used as a real-time medical decision support system for this task.

9

Simple and non-invasive liver fibrosis stage prediction method

Orczyk T., Pałys M., Porwik P., Musialik J., Błońska-Fajfrowska B.

Journal of Medical Informatics & Technologies

|

2011

|

Vol. 17

227--231

EN

In this paper a simple and non-expensive indirect fibrosis stage prediction method is described. Presented method is non-invasive and is based on the results of the generic blood tests. The method is based on a statistical analysis of wide range of blood tests results supported with the experience of hepatologists.

10

The new rule based colour classifier in the problem of human skin colour detection

Głowacki D., Porwik P., Orczyk T.

Journal of Medical Informatics & Technologies

|

2010

|

Vol. 14

39--48

EN

This article presents new rules, which can be used to construct a classifier for image areas segmentation. Segmentation is made on upon the colours, which are commonly associated with human skin colour. The new rules of this classifier have been developed on the basis of the analysis and modifications of two other classifiers, which has been described in the literature. Nowadays, such classifiers are commonly used in practice: in photographic equipment, photo-editing software, biological images analysis or in-room person detecting systems.

11

An attempt to improve Eigenface algorithm efficiency for colour images

Orczyk T., Porwik P.

Journal of Medical Informatics & Technologies

|

2010

|

Vol. 16

201--207

EN

This article presents an attempt to improve Eigenface algorithm efficiency by using image pre–filtering in order to eliminate background areas of the picture and illumination influence. The background is treated as noise, so when noise is present then efficiency of the algorithm decreases. In order to eliminating this inconvenience, analysed image is pre–filtered by means of the colour classifier. The classifier eliminates pixels which have different colour than an average human skin colour on a digital photo. This causes that the Eigenface algorithm is less sensitive to background noise. The illumination influence was minimized by using hue information instead of traditionally used luminance. The main advantage of the proposed approach is possibility of using in environments where diverse image background texture and scene illumination appears. The eigenfaces technique can be applied in handwriting analysis, voice recognition, hand gestures interpretation and medical imaging.