T1-weighted 3D Magnetic Resonance Imaging Research Articles

Green exercise versus indoor urban exercise: Related frontal brain thickness and cognitive performance

Background and aimsThis study compared, for the first time, cognitive function, and the cortical thickness of frontal areas in healthy adults who exercise regularly in outdoor green spaces ("green exercise group," GEG; n=22, mean age=39.91 years, females=11) with those who exercise in indoor urban settings ("urban exercise group," UEG; n=25, mean age=40 years, females=11). We further explored the associations between frontal brain thickness and cognitive measures in each group separately. MethodForty-seven healthy adults participated. All underwent a neuropsychological evaluation for short-term memory, working memory, and attention. High-resolution T1-weighted 3D magnetic resonance imaging (MRI) scans were acquired for each participant. The thickness of frontal brain regions was extracted and exported to JASP software for statistical analyses. ResultsThe GEG performed better on working memory and short-term memory tasks compared to the UEG. Additionally, the GEG exhibited increased thickness in the right anterior cingulate cortex (rACC) and the right superior frontal sulcus (rSFS). These areas were associated with short-term memory within the GEG. Fisher's test and Zou's intervals revealed that only the correlation between the rSFS and short-term memory was significantly different between the UEG and GEG. ConclusionThese findings suggest that exercising in outdoor natural areas might be related to better cognitive function and greater brain thickness in frontal areas. This study will contribute to the growing body of evidence linking brain, cognition and exercising in natural areas. Further research is needed to confirm these findings and explore the underlying mechanisms.

3-1-3 Weight averaging technique-based performance evaluation of deep neural networks for Alzheimer’s disease detection using structural MRI

Alzheimer’s disease (AD) is a progressive neurological disorder. It is identified by the gradual shrinkage of the brain and the loss of brain cells. This leads to cognitive decline and impaired social functioning, making it a major contributor to dementia. While there are no treatments to reverse AD’s progression, spotting the disease’s onset can have a significant impact in the medical field. Deep learning (DL) has revolutionized medical image classification by automating feature engineering, removing the requirement for human experts in feature extraction. DL-based solutions are highly accurate but demand a lot of training data, which poses a common challenge. Transfer learning (TL) has gained attention for its knack for handling limited data and expediting model training. This study uses TL to classify AD using T1-weighted 3D Magnetic Resonance Imaging (MRI) from the Alzheimer’s Disease Neuroimaging (ADNI) database. Four modified pre-trained deep neural networks (DNN), VGG16, MobileNet, DenseNet121, and NASNetMobile, are trained and evaluated on the ADNI dataset. The 3-1-3 weight averaging technique and fine-tuning improve the performance of the classification models. The evaluated accuracies for AD classification are VGG16: 98.75%; MobileNet: 97.5%; DenseNet: 97.5%; and NASNetMobile: 96.25%. The receiver operating characteristic (ROC), precision–recall (PR), and Kolmogorov-Smirnov (KS) statistic plots validate the effectiveness of the modified pre-trained model. Modified VGG16 excels with area under the curve (AUC) values of 0.99 for ROC and 0.998 for PR curves. The proposed approach shows effective AD classification by achieving high accuracy using the 3-1-3 weight averaging technique and fine-tuning.

Turbulence in surgical suction heads as detected by MRI.

Blood loss is common during surgical procedures, especially in open cardiac surgery. Allogenic blood transfusion is associated with increased morbidity and mortality. Blood conservation programs in cardiac surgery recommend re-transfusion of shed blood directly or after processing, as this decreases transfusion rates of allogenic blood. But aspiration of blood from the wound area is often associated with increased hemolysis, due to flow induced forces, mainly through development of turbulence. We evaluated magnetic resonance imaging (MRI) as a qualitative tool for detection of turbulence. MRI is sensitive to flow; this study uses velocity-compensated T1-weighted 3D MRI for turbulence detection in four geometrically different cardiotomy suction heads under comparable flow conditions (0-1250mL/min). Our standard control suction head Model A showed pronounced signs of turbulence at all flow rates measured, while turbulence was only detectable in our modified Models 1-3 at higher flow rates (Models 1 and 3) or not at all (Model 2). The comparison of flow performance of surgical suction heads with different geometries via acceleration-sensitized 3D MRI revealed significant differences in turbulence development between our standard control Model A and the modified alternatives (Models 1-3). As flow conditions during measurement have been comparable, the specific geometry of the respective suction heads must have been the main factor responsible. The underlying mechanisms and causative factors can only be speculated about, but as other investigations have shown, hemolytic activity is positively associated with degree of turbulence. The turbulence data measured in this study correlate with data from other investigations about hemolysis induced by surgical suction heads. The experimental MRI technique used showed added value for further elucidating the underlying physical phenomena causing blood damage due to non-physiological flow.

Automated Brain Metastases Detection Framework for T1-Weighted Contrast-Enhanced 3D MRI.

Brain Metastases (BM) complicate 20-40% of cancer cases. BM lesions can present as punctate (1mm) foci, requiring high-precision Magnetic Resonance Imaging (MRI) in order to prevent inadequate or delayed BM treatment. However, BM lesion detection remains challenging partly due to their structural similarities to normal structures (e.g., vasculature).We propose a BM-detection framework using a single-sequence gadolinium-enhanced T1-weighted 3D MRI dataset. The framework focuses on the detection of smaller (<15mm) BM lesions and consists of: (1) candidate-selection stage, using Laplacian of Gaussian approach for highlighting parts of an MRI volume holding higher BM occurrence probabilities, and (2) detection stage that iteratively processes cropped region-of-interest volumes centered by candidates using a custom-built 3D convolutional neural network ("CropNet"). Data is augmented extensively during training via a pipeline consisting of random gamma correction and elastic deformation stages; the framework thereby maintains its invariance for a plausible range of BM shape and intensity representations.This approach is tested using five-fold cross-validation on 217 datasets from 158 patients, with training and testing groups randomized per patient to eliminate learning bias. The BM database included lesions with a mean diameter of ∼5.4mm and a mean volume of ∼160mm3. For 90% BM-detection sensitivity, the framework produced on average 9.12 false-positive BM detections per patient (standard deviation of 3.49); for 85% sensitivity, the average number of false-positives declined to 5.85. Comparative analysis showed that the framework produces comparable BM-detection accuracy with the state-of-art approaches validated for significantly larger lesions.

Continuous monitoring of dough fermentation and bread baking by magnetic resonance microscopy

The consumer quality of baked products is closely related with dough structure properties. These are developed during dough fermentation and finalized during its baking. In this study, magnetic resonance microscopy (MRM) was employed in a study of dough fermentation and baking. A small hot air oven was installed inside a 2.35-T horizontal bore superconducting magnet. Four different samples of commercial bread mixes for home baking were used to prepare small samples of dough that were inserted in the oven and allowed to rise at 33°C for 112 min; this was followed by baking at 180°C for 49 min. The entire process was followed by dynamic T1-weighted 3D magnetic resonance imaging with 7 min of temporal resolution and 0.23×0.23×1.5 mm3 of spatial resolution. Acquired images were analyzed to determine time courses of dough pore distribution, dough volume and bread crust thickness. Image analysis showed that both the number of dough pores and the normalized dough volume increased in a sigmoid-like fashion during fermentation and decreased during baking due to the bread crust formation. The presented magnetic resonance method was found to be efficient in analysis of dough structure properties and in discrimination between different dough types.

Mapping of the habenulo-interpeduncular pathway in living mice using manganese-enhanced 3D MRI

This magnetic resonance imaging (MRI) study describes mapping of the habenulo-interpeduncular pathway in living mice based on manganese-induced contrast. Six hours after intracerebroventricular microinjection of MnCl2, T1-weighted 3D MRI (2.35 T) at 117 mum isotropic resolution revealed a continuous pattern of anterograde labeling from the habenula via the fasciculus retroflexus to the interpeduncular nucleus. Alternatively, the less invasive systemic administration of MnCl2 allowed for monitoring of the dynamic uptake pattern of respective neural components with even higher reproducibility across animals. Time courses covered the range from 42 min to 24 h after injection. In conclusion, manganese-enhanced MRI may open new ways for functional assessments of the habenulo-interpeduncular system in animal models with cognitive impairment.

In vivo 3D MRI staining of the mouse hippocampal system using intracerebral injection of MnCl 2

The morphology and function of the hippocampal system of C57BL/6J mice ( n = 8) was studied in vivo using T1-weighted 3D magnetic resonance imaging (MRI) (117 μm isotropic resolution) after bilateral injection of MnCl 2 (0.25 μl, 5 or 200 mM) into the posterior hippocampal formation. The neuronal uptake of the T1-shortening Mn 2+ ions resulted in a pronounced MRI signal enhancement within the CA3 subfield and dentate gyrus with milder increases in CA1 and subiculum. This finding is in line with differences in the excitability of hippocampal neurons previously reported using electrophysiologic recordings. The subsequent axonal transport of Mn 2+ highlighted the principal extrinsic projections from the posterior hippocampal formation via the fimbria and the precommissural fornix to the dorsal part of the lateral septal nucleus. A strong MRI signal enhancement was also observed in the ventral hippocampal commissure. A time-course analysis revealed unsaturated conditions of Mn 2+ accumulation at about 2 h after injection and optimal contrast-to-noise ratios at about 6 h after injection. The present results using Mn 2+-enhanced 3D MRI open new ways for studying the role of the hippocampal system in specific aspects of learning and memory in normal and mutant mice.

Structural brain abnormalities among relatives of patients with schizophrenia: implications for linkage studies

Several studies suggest that the nonschizophrenic relatives of schizophrenic patients exhibit structural brain abnormalities that may be manifestations of genes that predispose to schizophrenia. In this work, we examine the utility of such measures for linkage analyses. Subjects were 45 nonpsychotic first-degree adult relatives of schizophrenic patients and 48 normal controls. Sixty contiguous 3-mm coronal, T1-weighted 3D magnetic resonance images of the entire brain were acquired on a 1.5-T magnet. We used factor analysis to derive MRI-based phenotypes for analysis. The factor analyses produced three factors that significantly discriminated relatives from controls. We used a linear combination of the three factor scores to derive an MRI phenotype. A receiver operating characteristic (ROC) analysis of this phenotype estimated an area under the curve (AUC) statistic of 0.85. The phenotype also discriminated nonpsychotic relatives having two schizophrenic relatives from those having only one. The nonpsychotic relatives of schizophrenic patients show deviant values on MRI measures of brain structure and the distribution of these deviations among relatives and controls suggests that if these results can be replicated, an MRI-derived phenotype could be useful for genetic linkage and association analyses.

48. Medial temporal lobe memory system dysfunction in relatives of patients with schizophrenia

Growing evidence indicates that nonpsychotic relatives of patients with schizophrenia have subtle brain abnormalities in limbic, paralimbic and cortical regions. Verbal memory deficits in adult relatives have also been demonstrated. In this study we present more comprehensive morphometric analyses of the limbic and paralimbic cortices, to test the hypothesis that brain volume reductions, in a distributed memory encoding system, are associated with verbal memory deficits, and that they are more severe in nonpsychotic relatives with two schizophrenic family members (“multiplex”) as compared to one member with the illness (“simplex”). Subjects were 45 nonpsychotic first degree relatives of patients with schizophrenia and 48 matched normal controls. Sixty contiguous 3 millimeter coronal, T1-weighted 3D magnetic resonance images of the entire brain were acquired on a 1.5 Tesla magnet. Cortical and subcortical gray and white matter were segmented using a semi-automated intensity contour mapping algorithm. After segmenting the hippocampus and amygdala as a continuous volume, we parcellated the hippocampus from the amygdala. We used a new method to parcellate the fornix and mamillary bodies, along with the hippocampus, and the parahippocampal gyrus. Findings showed that the relatives had significant volumetric reductions in the left hippocampus, and parahippocampal gyri bilaterally, and that these were more severe in the relatives from “multiplex” families. Verbal memory deficits were significantly associated with hippocampal and parahippocampal volumes in the direction of smaller volumes, worse memory performance. These new data on cognitive and brain endophenotypes provide clues to the pathophysiology of schizophrenia.

Thalamic and amygdala–hippocampal volume reductions in first-degree relatives of patients with schizophrenia: an MRI-based morphometric analysis

Background: Schizophrenia is characterized by subcortical and cortical brain abnormalities. Evidence indicates that some nonpsychotic relatives of schizophrenic patients manifest biobehavioral abnormalities, including brain abnormalities. The goal of this study was to determine whether amygdala–hippocampal and thalamic abnormalities are present in relatives of schizophrenic patients. Methods: Subjects were 28 nonpsychotic, and nonschizotypal, first-degree adult relatives of schizophrenics and 26 normal control subjects. Sixty contiguous 3 mm coronal, T1-weighted 3D magnetic resonance images of the brain were acquired on a 1.5 Tesla magnet. Cortical and subcortical gray and white matter and cerebrospinal fluid (CSF) were segmented using a semi-automated intensity contour mapping algorithm. Analyses of covariance of the volumes of brain regions, controlling for expected intellectual (i.e., reading) ability and diagnosis, were used to compare groups. Results: The main findings were that relatives had significant volume reductions bilaterally in the amygdala–hippocampal region and thalamus compared to control subjects. Marginal differences were noted in the pallidum, putamen, cerebellum, and third and fourth ventricles. Conclusions: Results support the hypothesis that core components of the vulnerability to schizophrenia include structural abnormalities in the thalamus and amygdala–hippocampus. These findings require further work to determine if the abnormalities are an expression of the genetic liability to schizophrenia.

Reduced subcortical brain volumes in nonpsychotic siblings of schizophrenic patients: a pilot magnetic resonance imaging study.

Substantial evidence suggests that nonpsychotic relatives of schizophrenia patients manifest subtle abnormalities in communication, eye movements, event-related potentials, and neuropsychological processes of attention, reasoning, and memory. We sought to determine whether adult relatives without psychosis or schizophrenia spectrum diagnoses might also have structural brain abnormalities, particularly in subcortical regions found to be impaired in patients with schizophrenia itself. Subjects were six sisters of schizophrenic patients and eleven normal female controls. Sixty contiguous 3 mm coronal, T1-weighted 3D magnetic resonance images (MRI) of the entire brain were acquired on a 1.5 Tesla magnet. Cortical and subcortical gray and white matter was segmented using a semiautomated intensity contour mapping algorithm. Volumes were adjusted for total brain volumes. Adjusted gray matter subcortical volumes were significantly smaller in relatives than in controls in total hippocampus, right amygdala, right putamen, left thalamus, and brainstem. Relatives had significantly enlarged left and total inferior lateral ventricles. These results, though preliminary, suggest that some never-psychotic relatives of schizophrenic patients have abnormal brain structure. If replicated in a larger sample including both sexes, these results would suggest that the genetic liability to schizophrenia is also expressed as structural brain abnormalities.