Search

Recent functional neuroimaging and lesion studies demonstrate the involvement of the orbitofrontal/ventromedial prefrontal cortex as a key structure in decision making processes. This region seems to be particularly crucial when contingencies between options and consequences are unknown but have to be learned by the use of feedback following previous decisions (decision making under ambiguity). However, little is known about the neural correlates of decision making under risk conditions in which information about probabilities and potential outcomes is given. In the present study, we used functional magnetic resonance imaging to measure blood-oxygenation-level-dependent (BOLD) responses in 12 subjects during a decision making task. This task provided explicit information about probabilities and associated potential incentives. The responses were compared to BOLD signals in a control condition without information about incentives. In contrast to previous decision making studies, we completely removed the outcome phase following a decision to exclude the potential influence of feedback previously received on current decisions. The results indicate that the integration of information about probabilities and incentives leads to activations within the dorsolateral prefrontal cortex, the posterior parietal lobe, the anterior cingulate and the right lingual gyrus. We assume that this pattern of activation is due to the involvement of executive functions, conflict detection mechanisms and arithmetic operations during the deliberation phase of decisional processes that are based on explicit information.

Historical bases for the special role of the prefrontal cortex are outlined and the case of a 27-year-old woman with massive bilateral prefrontal damage of unknown etiology is then described. Frontal lobe degeneration was repeatedly examined with magnetic resonance imaging and fluoro-deoxy-D-glucose-positron emission tomography and was found to include both orbital and dorsolateral aspects of the frontal lobes. While the degeneration initially measured was limited to portions of the orbital, medial and dorsolateral parts of both frontal lobes, with right-sided predominance, a second brain scan 15 months later revealed massive shrinkage of both frontal lobes, together with additional involvement of the posterior association cortices. The patient had completed her high-school education and had superior verbal long-term memory, normal short-term memory, and normal priming, but manifested grossly deficient scores in various frontal lobe-sensitive tests. Though a number of neurological examinations were performed, no plausible cause for the damage was established.

Behavioral and neuroanatomical consequences of a thiamine-deficient diet, combined with the application of a thiamine-antagonist (pyrithiamine) were investigated in the cat. Eight cats (the experimental group) were subjected to a vitamin-B1-poor diet until they developed neurological symptoms (epileptic attacks, ataxia, gait disturbances), while 24 cats were fed normally and served as control group. Immediately following the appearance of neurological signs, a high dose of thiamine was given to the cats of the experimental group; they were then allowed to recover for ten days. Thereafter the performance in learning an alternation task in a T-maze was tested and compared with that of the control group. Behaviorally, the cats of the experimental group manifested drastically retarded acquisition rates in the learning task compared to the cats of the control group. Neuroanatomically, damage was found in the brains of each of the cats in the experimental group; this damage consisted mainly of enlarged ventricles, hemorrhages, neuronal loss and gliosis. Those regions most consistently affected were the periaqueductal gray, the inferior colliculi and the mamillary nuclei. The thalamic mediodorsal nucleus was affected to a minor degree in three cats only. Four cats manifested damage in the hippocampal formation. No damage was found in the cerebellum. Most of the damaged neuroanatomical loci resemble those found in patients with a Wernicke-Korsakoff syndrome.

Different areas of the frontal cortex of the cat were injected with small amounts of horseradish peroxidase. The region of labeled cells in the mediodorsal nucleus of the thalamus (MD) were related to the injected areas. Distinct relations between subdivisions of MD and of the prefrontal cortex were established: a rather large central sector of MD projects to the gyrus proreus and the anterior parts of the gyri sigmoideus, rectus, and frontalis. A narrow lateral band of anterior MD neurons projects predominantly to an area on both sides of the sulcus praesylvius, whereas a postero-lateral band sends fibers to a region on the ventral anterior sylvian gyrus. The area between the presylvian sulcus and the sylvian gyrus is apparently free of MD afferents, but not of other thalamic afferents. A fourth sector of MD, situated dorsomedially, projects to the middle parts of the gyri rectus and frontalis. And a fifth sector, located ventrally to the dorsomedial MD sector, projects to the ventral part of the gyrus rectus. The established subfields of MD and of the prefrontal cortex are discussed with respect to previous anatomical research in the cat.

Thirty-three cat brains with injections of horseradish peroxidase in various regions of the cerebral cortex were screened for afferent projections from the ventral tegmental area, the locus ceruleus, and the parabrachial nuclei. All three structures were found to project to rather divergent parts of the cortex, including regions in the posterior half of the hemisphere. These results, especially for the ventral tegmental area and, to a lesser degree, for the parabrachial neurons, disagree with most of the target loci of established cortical afferents in the rat. Though our results might be attributed to species differences in the cortical innervation of brain stem structures, we prefer explanations which emphasize different densities in the distribution of brain stem afferents to the cortex, and/or which suggest different cortical targets of catecholaminergic and noncatecholaminergic neurons.

The afferent projections to, and the interconnections between, four structures of the so-called limbic system were investigated in the cat. The retrograde horseradish peroxidase (HRP) technique was used to trace the origins of fibers projecting to each of these four loci. Particular emphasis was laid on tracing cortical afferents of these regions. Four injections were performed in the dorsal and two in the ventral subicular regions; six were centered within the mamillary nuclei, four within the anterior thalamic nuclei, and three within the cingulate gyrus. For each region, a number of projections were found which had apparently not been described before, at least not for the cat: For injections into the subicular regions, a hitherto unknown number of cortical afferents was detected, including labeled cells in the prefrontal and premotor fields and from large areas within the posterior parietal, temporal and occipital cortex (i.e., sensory and sensory integration cortex); numerous neurons were labeled in the anterior nuclear group of the thalamus. Injections of HRP into the mamillary nuclei revealed, aside from a strong projection from the subicular regions, frontocortical and cingulate projections to the mamillary nuclei; the mamillary nuclei also received subcortical projections from the septum, the diagonal band of Broca and from the periaqueductal gray. Following injections into the anterior thalamic nuclei, labeled cells were found in the prefrontal cortex, and to a lesser extent in lateral parts of the cortical hemisphere; subcortically, the mamillary nuclei received connections from hypothalamic areas, the periaqueductal gray, the diagonal band of Broca and the claustrum. Cingulate injections labeled cells in temporal and parietal cortical areas, in the subicular region, and also in the periaqueductal gray. Our findings reveal that each of the four injected areas receives a large number of afferents from divergent regions of the brain; of these, a considerable number is shared by each of the four injection loci. Furthermore, the present results reveal that the subiculum, the mamillary bodies, and the anterior thalamus are more strongly interconnected than previously assumed.