Commonalities and Discrepancies in the Relationships between Behavioural Outcome and the Results of Neuroimaging in Brain-Damaged Patients

Abstract

Variables which are of influence in establishing clear predictions of neuropsychological alterations from neuroradiological data (and vice versa) are documented and discussed. It is concluded that personality factors and the kind and locus of brain lesions are the most crucial determinants. The locus of the brain damage may have cumulative effects either when it is situated in a strategic place (usually within the white matter, affecting interneuronal communication) or when various types of lesions appear superimposed (combination of focal and diffuse lesions). Consequently, the consideration of the patient's personality background and of as many neuropsychological facts as possible may considerably increase the validity of outcome predictions. When static or dynamic neuroimaging fails to show abnormalities in spite of obvious psychological alterations, an intensive neuropsychological documentation may even replace neuroradiology. Keywords: Brain damage -Magnetic resonance imaging -Computer tomography -Neuropsychological assessment INTRODUCTIONThe availability of cranial computer tomography (CT) has revolutionized the diagnostic power for patients with brain damage, and the subsequent appearance of magnetic resonance imaging (MRI), positron emission tomography (PET) and functional magnetic resonance imaging has further increased and refined the power of relating brain tissue changes to behavioural alterations. Nevertheless, there are numerous case reports in which neuroimaging and neuropsychological data indicate a different prognosis, thereby leading to an improper rehabilitation treatment of the patient.Part of the discrepancy may be explained by the false assumption of linear relationships between lesion size and behavioural outcome (e.g. Irle, 1987Irle, , 1990Grafman et al., 1986), but other parts are most likely due to insensitive measurement on the imaging or the psychological level. Alesch et al. (1991), for instance, found that CT is quite insensitive to the identification of lymphomas although gliomas are readily detected. Similarly, although subdural empyemas and most sterile effusions and chronic subdural haematomas appear similar in CT scans, they are (~ 1996 Rapid Science Publishers much more readily distinguishable on the basis of signal intensity differences in MRI (Takamura et al., 1995);and Kertesz et al. (1987) concluded from their analyses that MRI surpasses CT scanning in the early detection of cerebral infarcts. CT on the other hand should be the method of choice to rule out intracerebral bleeding when a patient is screened initially. For detecting anatomical correlates of abnormal behavioural functioning, such as in psychogenic amnesia, conventional static imaging techniques have been useless up to now, although dynamic methods, such as PET imaging, may reveal cerebral blood flow changes indicative of abnormal processing (Amsterdam and Mozley, 1992;Markus et al., 1992).Congruence and incongruence of neuropsychology and neuroradiology are the topics of this commentary. We will first illustr...