NMF-Based analysis of SPECT brain images for the diagnosis of alzheimer's disease

  • Authors:

  • Pablo Padilla;Juan-Manuel Górriz;Javier Ramírez;Elmar Lang;Rosa Chaves;Fermin Segovia;Ignacio Álvarez;Diego Salas-González;Miriam López

  • Affiliations:

  • University of Granada, Granada, Spain;University of Granada, Granada, Spain;University of Granada, Granada, Spain;CIML Group, Biophysics, University of Regensburg, Regensburg, Germany;University of Granada, Granada, Spain;University of Granada, Granada, Spain;University of Granada, Granada, Spain;University of Granada, Granada, Spain;University of Granada, Granada, Spain

  • Venue:
  • HAIS'10 Proceedings of the 5th international conference on Hybrid Artificial Intelligence Systems - Volume Part I
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper offers a computer-aided diagnosis (CAD) technique for early diagnosis of Alzheimer's disease (AD) by means of single photon emission computed tomography (SPECT) image classification The SPECT database for different patients is analyzed by applying the Fisher discriminant ratio (FDR) and non-negative matrix factorization (NMF) for the selection and extraction of the most significative features of each patient SPECT data, in order to reduce the large dimensionality of the input data and the problem of the curse of dimensionality, extracting score features The NMF-transformed set of data, with reduced number of features, is classified by means of support vector machines (SVM) classification The proposed NMF+SVM method yields up to 94% classification accuracy, thus becoming an accurate method for SPECT image classification For the sake of completeness, comparison between conventional PCA+SVM method and the proposed method is also provided.