Alcohol-dependent individuals usually favor instant gratification of alcohol use and ignore its long-term negative consequences, reflecting impaired decision-making. According to the somatic marker hypothesis, decision-making abilities are subtended by an extended brain network. As chronic alcohol consumption is known to be associated with brain shrinkage in this network, the present study investigated relationships between brain shrinkage and decision-making impairments in alcohol-dependent individuals early in abstinence using voxel-based morphometry. Thirty patients performed the Iowa Gambling Task and underwent a magnetic resonance imaging investigation (1.5T). Decision-making performances and brain data were compared with those of age-matched healthy controls. In the alcoholic group, a multiple regression analysis was conducted with two predictors (gray matter [GM] volume and decision-making measure) and two covariates (number of withdrawals and duration of alcoholism). Compared with controls, alcoholics had impaired decision-making and widespread reduced gray matter volume, especially in regions involved in decision-making. The regression analysis revealed links between high GM volume in the ventromedial prefrontal cortex, dorsal anterior cingulate cortex and right hippocampal formation, and high decision-making scores (P < 0.001, uncorrected). Decision-making deficits in alcoholism may result from impairment of both emotional and cognitive networks.

Using magnetic resonance imaging, this study compared hippocampal volume between 145 very preterm children and 34 children born full-term at 7 years of age. The relationship between hippocampal volume and memory and learning impairments at 7 years was also investigated. Manual hippocampal segmentation and subsequent three-dimensional volumetric analysis revealed reduced hippocampal volumes in very preterm children compared with term peers. However, this relationship did not remain after correcting for whole brain volume and neonatal brain abnormality. Contrary to expectations, hippocampal volume in the very preterm cohort was not related to memory and learning outcomes. Further research investigating the effects of very preterm birth on more extensive networks in the brain that support memory and learning in middle childhood is needed. (JINS, 2013, 19, 1–11)

Congenital hypothyroidism is a pediatric endocrine disorder caused by insufficient endogenous thyroid hormone production. Children with congenital hypothyroidism have difficulties with episodic memory and abnormalities in hippocampal structure, suggesting deficient hippocampal functioning. To assess hippocampal activation in adolescents with congenital hypothyroidism (N = 14; age range, 11.5–14.7 years) compared with controls (N = 15; age range, 11.2–15.5 years), a functional magnetic resonance imaging visuospatial memory task was used. In this task, participants had to decide if object pairings were novel or were previously studied or if object pairs were in the same location as they were at study or had switched locations. Despite no group differences in task performance, adolescents with congenital hypothyroidism showed both increased magnitude of hippocampal activation relative to controls and bilateral hippocampal activation when only the left was observed in controls. Furthermore, the increased activation in the congenital hypothyroidism group was correlated with the severity of the hypothyroidism experienced early in life. These results suggest that perinatal deprivation of thyroid hormone has longstanding effects on hippocampal function and may account for memory problems experienced by adolescents with congenital hypothyroidism. (JINS, 2012, 18, 49–56)

To evaluate whether the lines occasionally detected on clinical magnetic resonance (MR) images are genuine hippocampal layers, a formalin fixed hippocampal specimen was scanned using T2 weighted sequences at 7 Tesla (voxel dimensions 0.064×0.064×1 mm) and at 1.5 Tesla (voxel dimensions: 0.156×0.156×1 mm) and compared with the results of histological examination. In addition, a healthy volunteer was scanned with a T2 weighted sequence at 1.5 Tesla (voxel dimensions: 0.469×0.469×2 mm). On 7 Tesla images hippocampal layers and the granule cell layer of the dentate were visible. On 1.5 Tesla images of the specimen, the hippocampal layers were again identified, but the granule cell layer of the dentate was not detectable. On 1.5 Tesla images of the hippocampus in vivo, 3 layers could be distinguished in the hippocampus on some slices. These mainly represented the alveus, pyramidal cell layer and stratum radiatum. A dark line consisting of a few pixels possibly represented the dentate gyrus. Our results show that the lines occasionally detected on clinical MR images are likely to be real hippocampal layers. However, the resolution currently used in clinical imaging (typically 0.469×0.469×2 mm or lower) is not sufficient for the detection of all hippocampal layers. For the reliable detection of all hippocampal layers on MR images an increase by a factor of approximately 20 would be necessary.