ROC and reproducibility analysis of designed algorithm for potential diagnosis of Parkinson's disease in ultrasound images

Authors:
Jiří Blahuta;Tomáš Soukup;Petr Čermák;Michal Večerek;Milan Jakel;David Novák
Affiliations:
Department of Informatics, Silesian University in Opava, Opava, Czech Republic;Department of Informatics, Silesian University in Opava, Opava, Czech Republic;Department of Informatics, Silesian University in Opava, Opava, Czech Republic;Department of Informatics, Silesian University in Opava, Opava, Czech Republic;Department of Informatics, Silesian University in Opava, Opava, Czech Republic;Department of Informatics, Silesian University in Opava, Opava, Czech Republic
Venue:
MMES'11/DEEE'11/COMATIA'11 Proceedings of the 2nd international conference on Mathematical Models for Engineering Science, and proceedings of the 2nd international conference on Development, Energy, Environment, Economics, and proceedings of the 2nd international conference on Communication and Management in Technological Innovation and Academic Globalization
Year:
2011

Citing 3
Cited 0

Using Multivariate Statistics (5th Edition)

Using Multivariate Statistics (5th Edition)
The image recognition of brain-stem ultrasound images with using a neural network based on PCA

ACELAE'11 Proceedings of the 10th WSEAS international conference on communications, electrical & computer engineering, and 9th WSEAS international conference on Applied electromagnetics, wireless and optical communications
Image processing of medical diagnostic neurosonographical images in MATLAB

Proceedings of the 15th WSEAS international conference on Computers

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper introduces our experimental software which has been developed in MATLAB for potential detection of pathology to detection of Parkinson's disease. The algorithm is based on thresholding of intensity images, using the elliptical ROI followed by computing of area in this ROI for each intensity. The second step is statistical analysis of these data such as ROC curve, variability and Cohen's kappa coefficient. We used DICOM or converted images into JPEG of TCS (transcranial ultrasound images). This statistics is critical for appraisal of repeatability and reproducibility of designed algorithm.