Quantitative brain SPECT in Alzheimer's disease and normal aging

Area of Science:

• Neuroimaging
• Radiology
• Neurology

Background:

• Alzheimer's disease (AD) diagnosis relies on clinical assessment and neuroimaging.
• Brain single-photon emission computed tomography (SPECT) offers potential for AD diagnosis.
• Objective quantitative methods are needed to enhance SPECT's diagnostic utility in AD.

Purpose of the Study:

• To develop and evaluate an objective, quantitative image analysis protocol for differentiating AD patients from healthy elderly controls using HMPAO-SPECT.
• To assess the diagnostic accuracy of the developed quantitative SPECT method.

Main Methods:

• HMPAO-SPECT datasets from 29 probable AD patients and 78 controls were analyzed.
• Automated procedures quantified activity levels in 120 standardized cortical volumes.
• Quadratic discriminant analysis classified subjects using global and regional normalized activity levels.

Main Results:

• The quantitative SPECT method achieved a classification performance (area under ROC curve) of 0.923 +/- 0.036.
• At a threshold likelihood ratio of 1, sensitivity was 91% and specificity was 86% for distinguishing AD patients.
• The analysis identified significant differences in regional brain activity between AD patients and controls.

Conclusions:

• Quantitative SPECT imaging provides an accurate and objective method for differentiating Alzheimer's disease patients from healthy elderly individuals.
• This approach enhances the diagnostic utility of SPECT in clinical settings for AD.
• The findings support the use of quantitative SPECT analysis as a valuable tool in AD diagnostics.