Advances in brain SPECT : methodological and human investigations

Abstract

The aim of this work was to optimise and validate a new approach for examination of die regional cerebral blood flow (rCBF) using Single Photon Emission Computed Tomography (SPECT) for application in clinical research and patient care. It also involved the implementation of a software system for evaluation and analysis of rCBF. The task was accomplished by validating and using the SPECT-methodology and forming a group of healthy individuals serving as a reference material. Examinations were made with a three-headed gamma camera operating with ultra-high resolution collimators. The radiopharmaceutical used was 99Tcm-D,L-Hexamethylpropylenearnine oxime (99Tcm-HMPAO).

To provide a means for semi-quantitative evaluations of rC13F, a Computerised Brain Atlas (CBA) was adopted. This is a software tool that can be applied to any brain imaging modality. By this, an individual brain SPECT examination is brought into registration in the same 3D space with a reference image created by averaging scans of several normal controls. Once the standardised volume is defined, the CBA brings the registered images into a form in which each assessed image unit corresponds to a part of an anatomo-functional region and processes functional comparisons between various homologous regions. The CBA allows comparisons between different studies of the same individual in order to study the influence of an intervening factor as well as subtractions between a single individual and a reference group in order to highlight the changes in rCBF distribution. As 99Tcm-HMPAO does not allow an accurate absolute assessment of the global blood flow, a normalisation procedure is performed by the atlas, giving rise to relative figures.

The validation of the technique was performed by comparison to Positron Emission Tomography (PET), the "golden standard" in nuclear medicine. In the same individuals, a SPECT-study and a PET-study using [15O]-butanol, the latter also representing the brain blood flow, were made at separate occasions but under similar conditions. SPECT and PET demonstrated a rather good agreement, despite an expected higher spatial resolution of PET resulting in higher relative flow data in central structures. The reproducibility and the repeatability of rCBF SPECT examination were also evaluated in order to investigate the variability of this technique. Analogous SPECT-studies were repeated in the same individual after a few hours and again after three months. The variation in normalised flow between individuals was less than 3% for all brain regions, thus reflecting a high reproducibility.

The approach was utilised in a baromedicine research project where the influence of acute hypobaric hypoxia on the rC13F was investigated in normal volunteers. Motor cortex and basal ganglia were significantly influenced by hypoxia and the increased rCB17 was imaged by subtracting the data sets at hypoxia and normoxia. In further study, in addition to the known rCB17 pattern in Alzheimer Disease and Frontal Lobe Dementia, two regions with unexpected low CBF were identified: the anterior cingulate cortex in Frontal Lobe Dementia and the nc caudatus at both conditions. This confirmed the possibility to clinically implement the method as an adjunct to the visual evaluation of the SPECT image. The proposed approach has revised the SPECT routine technique, clinically validated a brain standardisation software and underscored the necessity of a reliable and numerically large control group of healthy individuals in order to increase the diagnostic accuracy. The combination of these factors adds value in the clinical routine to the analysis of SPECT data and to subsequent clinical diagnosis.