Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Jul;111(1):255-68.
doi: 10.1016/j.cmpb.2013.03.015. Epub 2013 May 6.

Functional activity maps based on significance measures and Independent Component Analysis

F J Martínez-Murcia  1 , J M Górriz, J Ramírez, C G Puntonet, I A Illán; Alzheimer's Disease Neuroimaging Initiative
Collaborators, Affiliations

Functional activity maps based on significance measures and Independent Component Analysis

F J Martínez-Murcia et al. Comput Methods Programs Biomed. 2013 Jul.

Abstract

The use of functional imaging has been proven very helpful for the process of diagnosis of neurodegenerative diseases, such as Alzheimer's Disease (AD). In many cases, the analysis of these images is performed by manual reorientation and visual interpretation. Therefore, new statistical techniques to perform a more quantitative analysis are needed. In this work, a new statistical approximation to the analysis of functional images, based on significance measures and Independent Component Analysis (ICA) is presented. After the images preprocessing, voxels that allow better separation of the two classes are extracted, using significance measures such as the Mann-Whitney-Wilcoxon U-Test (MWW) and Relative Entropy (RE). After this feature selection step, the voxels vector is modelled by means of ICA, extracting a few independent components which will be used as an input to the classifier. Naive Bayes and Support Vector Machine (SVM) classifiers are used in this work. The proposed system has been applied to two different databases. A 96-subjects Single Photon Emission Computed Tomography (SPECT) database from the "Virgen de las Nieves" Hospital in Granada, Spain, and a 196-subjects Positron Emission Tomography (PET) database from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Values of accuracy up to 96.9% and 91.3% for SPECT and PET databases are achieved by the proposed system, which has yielded many benefits over methods proposed on recent works.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest

There is no conflict of interest.

Figures

Fig. 1 –
Fig. 1 –
Several selected slices of the significance map obtained for each brain coordinate by applying the Relative Entropy criterium to the ADNI database.
Fig. 2 –
Fig. 2 –
Some selected slices depicting a spatial representation of the first four components obtained by ICA in the ADNI database.
Fig. 3 –
Fig. 3 –
Accuracy, sensitivity and specificity values for a Quadratic Bayes classifier in function of the number of selected voxels and the number of features extracted, using the SPECT database. The highest value obtained by the baseline method is presented for comparison purposes in Fig. 5(b).
Fig. 4 –
Fig. 4 –
Accuracy, sensitivity and specificity values for a SVM linear classifier, in function of the number of selected voxels and the number of features extracted, using the ADNI database. The highest value obtained by the baseline method is presented for comparison purposes in Fig. 5(b).
Fig. 5 –
Fig. 5 –
Average accuracy for each database and each evaluation method in funcdon of the number of selected voxels (a) and the number of independent components extracted (b), computed over the number of independent components and the number of selected voxels, respectively.
Fig. 6 –
Fig. 6 –
Value of the three first ICA projections of the images on the database, and decision surfaces designed by a SVM classifier with linear kernel ((a) for VDLN and (b) for ADNI) and a Naive Bayes Classifier with quadratic discriminant function ((c) for VDLN and (d) for ADNI). The number of voxels selected was N = 16, 000 and the number of Independent Components was K = 4
Fig. 7 –
Fig. 7 –
ROC curves obtained by the proposed system for SPECT (a) and ADNI (b) databases, the base systems VAF and SVAF and the PCA and FA based methods.
See this image and copyright information in PMC

References

    1. Friston K, Ashburner J, Kiebel S, Nichols T, Penny W, Statistical Parametric Mapping: The Analysis of Functional Brain Images, Academic Press, Amsterdam, 2007.
    1. Salas-Gonzalez D, Gorriz JM, Ramirez J, Illan IA, Lopez M, Segovia F, Chaves R, Padilla P, Puntonet CC, Feature selection using factor analysis for Alzheimer’s diagnosis using F-FDG pet images, Medical Physics 37 (11) (2010) 6084–6095. - PMC - PubMed
    1. Dhawan AP, A review on biomedical image processing and future trends, Computer Methods and Programs in Biomedicine 31 (3/4) (1990) 141–183, 10.1016/0169-2607(90)90001-P, URL http://www.sciencedirect.com/science/article/pii/016926079090001P - DOI - PubMed
    1. Saxena P, Pavel DC, Quintana JC, Horwitz B, An automatic threshold-based scaling method for enhancing the usefulness of Tc-HMPAO SPECT in the diagnosis of Alzheimer’s disease, in: Medical Image Computing and Computer-Assisted Intervention – MICCAI, Vol. 1496 of Lecture Notes in Computer Science, Springer, Heidelberg, Germany, 1998, pp. 623–630.
    1. Stoeckel J, Ayache N, Malandain G, Koulibaly PM, Ebmeier KP, Darcourt J, Automatic classification of SPECT images of Alzheimer’s disease patients and control subjects, in: Medical Image Computing and Computer-Assisted Intervention – MICCAI, Vol. 3217 of Lecture Notes in Computer Science, Springer, Heidelberg, Germany, 2004, pp. 654–662.

Publication types

MeSH terms