Towards the computer-aided diagnosis of dementia based on the geometric and network connectivity of structural MRI data

We present an intuitive geometric approach for analysing the structure and fragility of T1-weighted structural MRI scans of human brains. Apart from computing characteristics like the surface area and volume of regions of the brain that consist of highly active voxels, we also employ Network Theory in order to test how close these regions are to breaking apart. This analysis is used in an attempt to automatically classify subjects into three categories: Alzheimer’s disease, mild cognitive impairment and healthy controls, for the CADDementia Challenge.

An LDA and probability-based classifier for the diagnosis of Alzheimer's Disease from structural MRI

In this paper a custom classification algorithm based on linear discriminant analysis and probability-based weights is implemented and applied to the hippocampus measurements of structural magnetic resonance images from healthy subjects and Alzheimer’s Disease sufferers; and then attempts to diagnose them as accurately as possible. The classifier works by classifying each measurement of a hippocampal volume as healthy controlsized or Alzheimer’s Disease-sized, these new features are then weighted and used to classify the subject as a healthy control or suffering from Alzheimer’s Disease. The preliminary results obtained reach an accuracy of 85.8% and this is a similar accuracy to state-of-the-art methods such as a Naive Bayes classifier and a Support Vector Machine. An advantage of the method proposed in this paper over the aforementioned state of the art classifiers is the descriptive ability of the classifications it produces. The descriptive model can be of great help to aid a doctor in the diagnosis of Alzheimer’s Disease, or even further the understand of how Alzheimer’s Disease affects the hippocampus.

A Genetic Algorithm for the selection of structural MRI features for classification of Mild Cognitive Impairment and Alzheimer's Disease

This work investigates the problem of feature selection in neuroimaging features from structural MRI brain images for the classification of subjects as healthy controls, suffering from Mild Cognitive Impairment or Alzheimer’s Disease. A Genetic Algorithm wrapper method for feature selection is adopted in conjunction with a Support Vector Machine classifier. In very large feature sets, feature selection is found to be redundant as the accuracy is often worsened when compared to an Support Vector Machine with no feature selection. However, when just the hippocampal subfields are used, feature selection shows a significant improvement of the classification accuracy. Three-class Support Vector Machines and two-class Support Vector Machines combined with weighted voting are also compared with the former and found more useful. The highest accuracy achieved at classifying the test data was 65.5% using a genetic algorithm for feature selection with a three-class Support Vector Machine classifier.

Arterial spin labeling of cerebral perfusion territories using a seperate labeling coil

Purpose: To obtain cerebral perfusion territories of the left, the right. and the posterior circulation in humans with high signal-to-noise ratio (SNR) and robust delineation. Materials and Methods: Continuous arterial spin labeling (CASL) was implemented using a dedicated radio frequency (RF) coil. positioned over the neck, to label the major cerebral feeding arteries in humans. Selective labeling was achieved by flow-driven adiabatic fast passage and by tilting the longitudinal labeling gradient about the Y-axis by theta = +/- 60 degrees. Results: Mean cerebral blood flow (CBF) values in gray matter (GM) and white matter (WM) were 74 +/- 13 mL center dot 100 g(-1) center dot minute(-1) and 14 +/- 13 mL center dot 100 g(-1) center dot minute(-1), respectively (N = 14). There were no signal differences between left and right hemispheres when theta = 0 degrees (P > 0.19), indicating efficient labeling of both hemispheres. When theta = +60 degrees, the signal in GM on the left hemisphere, 0.07 +/- 0.06%, was 92% lower than on the right hemisphere. 0.85 +/- 0.30% (P < 1 x 10(-9)). while for theta = -60 degrees, the signal in the right hemisphere. 0.16 +/- 0.13%, was 82% lower than on the contralateral side. 0.89 +/- 0.22% (P < 1 x 10(-10)). Similar attenuations were obtained in WM. Conclusion: Clear delineation of the left and right cerebral perfusion territories was obtained, allowing discrimination of the anterior and posterior circulation in each hemisphere.

Arterial Spin Labeling Measurements of Cerebral Perfusion Territories in Experimental Ischemic Stroke

Collateral circulation, defined as the supplementary vascular network that maintains cerebral blood flow (CBF) when the main vessels fail, constitutes one important defense mechanism of the brain against ischemic stroke. In the present study, continuous arterial spin labeling (CASL) was used to quantify CBF and obtain perfusion territory maps of the major cerebral arteries in spontaneously hypertensive rats (SHRs) and their normotensive Wistar-Kyoto (WKY) controls. Results show that both WKY and SHR have complementary, yet significantly asymmetric perfusion territories. Right or left dominances were observed in territories of the anterior (ACA), middle and posterior cerebral arteries, and the thalamic artery. Magnetic resonance angiography showed that some of the asymmetries were correlated with variations of the ACA. The leptomeningeal circulation perfusing the outer layers of the cortex was observed as well. Significant and permanent changes in perfusion territories were obtained after temporary occlusion of the right middle cerebral artery in both SHR and WKY, regardless of their particular dominance. However, animals with right dominance presented a larger volume change of the left perfusion territory (23 +/- 9%) than animals with left dominance (7 +/- 5%, P<0.002). The data suggest that animals with contralesional dominance primarily safeguard local CBF values with small changes in contralesional perfusion territory, while animals with ipsilesional dominance show a reversal of dominance and a substantial increase in contralesional perfusion territory. These findings show the usefulness of CASL to probe the collateral circulation.

Near-Term Fetal Hypoxia-Ischemia in Rabbits MRI Can Predict Muscle Tone Abnormalities and Deep Brain Injury

Background and Purpose-The pattern of antenatal brain injury varies with gestational age at the time of insult. Deep brain nuclei are often injured at older gestational ages. Having previously shown postnatal hypertonia after preterm fetal rabbit hypoxia-ischemia, the objective of this study was to investigate the causal relationship between the dynamic regional pattern of brain injury on MRI and the evolution of muscle tone in the near-term rabbit fetus. Methods-Serial MRI was performed on New Zealand white rabbit fetuses to determine equipotency of fetal hypoxia-ischemia during uterine ischemia comparing 29 days gestation (E29, 92% gestation) with E22 and E25. E29 postnatal kits at 4, 24, and 72 hours after hypoxia-ischemia underwent T2- and diffusion-weighted imaging. Quantitative assessments of tone were made serially using a torque apparatus in addition to clinical assessments. Results-Based on the brain apparent diffusion coefficient, 32 minutes of uterine ischemia was selected for E29 fetuses. At E30, 58% of the survivors manifested hind limb hypotonia. By E32, 71% of the hypotonic kits developed dystonic hypertonia. Marked and persistent apparent diffusion coefficient reduction in the basal ganglia, thalamus, and brain stem was predictive of these motor deficits. Conclusions-MRI observation of deep brain injury 6 to 24 hours after near-term hypoxia-ischemia predicts dystonic hypertonia postnatally. Torque-displacement measurements indicate that motor deficits in rabbits progressed from initial hypotonia to hypertonia, similar to human cerebral palsy, but in a compressed timeframe. The presence of deep brain injury and quantitative shift from hypo-to hypertonia may identify patients at risk for developing cerebral palsy. (Stroke. 2012;43:2757-2763.)

Assessment of myocardial infarction by CT angiography and cardiovascular MRI in patients with cocaine-associated chest pain: a pilot study

Objectives: Cocaine is a commonly used illicit drug that leads to the most emergency department (ED) visits. Chest pain is the most common presentation, reported in 40% of patients. Our aim was to evaluate the incidence of previous myocardial infarction among young cocaine users (18-40 years) with cocaine-associated chest pain by the assessment of myocardial fibrosis by cardiovascular MRI. Second, we also intended to evaluate the coronary tree by CT angiography (CTA). Methods: 24 cocaine users (22 males) who frequently complained about cocaine-associated chest pain underwent CTA and cardiovascular MRI. Mean age of patients was 29.7 years and most of them (79%) had frequently used inhalatory cocaine. Results: The calcium score turned out to be positive in only one patient (Agatston=54). Among the coronary segments evaluated, only one patient had calcified plaques at the anterior descending coronary artery (proximal and medium segments). Assessment of regional ventricular function by the evaluation of 17 segments was normal in all patients. None of the patients showed myocardial delayed enhancement, indicative of myocardial fibrosis. CTA therefore confirmed the low cardiovascular risk of these patients, since most of them (96%) had no atherosclerosis detected by this examination. Only one patient (4%) had coronary atherosclerosis detected, without significant coronary stenosis. Conclusion: Cardiovascular MR did not detect the presence of delayed enhancement indicative of myocardial fibrosis among young cocaine users with low cardiovascular risk who had complained of cocaine-associated chest pain.

Whole-body MRI: comprehensive evaluation on a 48-channel 3T MRI system in less than 40 minutes. Preliminary results

OBJECTIVE: To evaluate a comprehensive MRI protocol that investigates for cancer, vascular disease, and degenerative/inflammatory disease from the head to the pelvis in less than 40 minutes on a new generation 48-channel 3T system. MATERIALS AND METHODS: All MR studies were performed on a 48-channel 3T MR scanner. A 20-channel head/neck coil, two 18-channel body arrays, and a 32-channel spine array were employed. A total of 4 healthy individuals were studied. The designed protocol included a combination of single-shot T2-weighted sequences, T1-weighted 3D gradient-echo pre- and post-gadolinium. All images were retrospectively evaluated by two radiologists independently for overall image quality. RESULTS: The image quality for cancer was rated as excellent in the liver, pancreas, kidneys, lungs, pelvic organs, and brain, and rated as fair in the colon and breast. For vascular diseases ratings were excellent in the aorta, major branch vessel origins, inferior vena cava, portal and hepatic veins, rated as good in pulmonary arteries, and as poor in the coronary arteries. For degenerative/inflammatory diseases ratings were excellent in the brain, liver and pancreas. The inter-observer agreement was excellent. CONCLUSION: A comprehensive and time efficient screening for important categories of disease processes may be achieved with high quality imaging in a new generation 48-channel 3T system.

Evaluation of renal allograft function early after transplantation with diffusion-weighted MR imaging

To determine the inter-patient variability of apparent diffusion coefficients (ADC) and concurrent micro-circulation contributions from diffusion-weighted MR imaging (DW-MRI) in renal allografts early after transplantation, and to obtain initial information on whether these measures are altered in histologically proven acute allograft rejection (AR).

Ventilation-perfusion ratio in perflubron during partial liquid ventilation

BACKGROUND: Functional magnetic resonance imaging (fMRI) of fluorine-19 allows for the mapping of oxygen partial pressure within perfluorocarbons in the alveolar space (Pao(2)). Theoretically, fMRI-detected Pao(2) can be combined with the Fick principle approach, i.e., a mass balance of oxygen uptake by ventilation and delivery by perfusion, to quantify the ventilation-perfusion ratio (Va/Q) of a lung region: The mixed venous blood and the inspiratory oxygen fraction, which are equal for all lung regions, are measured. In addition, the local expiratory oxygen fraction and the end capillary oxygen content, both of which may differ between the lung regions, are calculated using the fMRI-detected Pao(2). We investigated this approach by numerical simulations and applied it to quantify local Va/Q in the perfluorocarbons during partial liquid ventilation. METHODS: Numerical simulations were performed to analyze the sensitivity of the Va/Q calculation and to compare this approach with another one proposed by Rizi et al. in 2004 (Magn Reson Med 2004;52:65-72). Experimentally, the method was used during partial liquid ventilation in 7 anesthetized pigs. The Pao(2) distribution in intraalveolar perflubron was measured by fluorine-19 MRI. Respiratory gas fractions together with arterial and mixed venous blood samples were taken to quantify oxygen partial pressure and content. Using the Fick principle, the local Va/Q was estimated. The impact of gravity (nondependent versus dependent) of perflubron dose (10 vs 20 mL/kg body weight) and of inspired oxygen fraction (Fio(2)) (0.4-1.0) on Va/Q was examined. RESULTS: In numerical simulations, the Fick principle proved to be appropriate over the Va/Q range from 0.02 to 2.5. Va/Q values were in acceptable agreement with the method published by Rizi et al. In the experimental setting, low mean Va/Q values were found in perflubron (confidence interval [CI] 0.08-0.29 with 20 mL/kg perflubron). At this dose, Va/Q in the nondependent lung was higher (CI 0.18-0.39) than in the dependent lung regions (CI 0.06-0.16; P = 0.006; Student t test). Differences depending on Fio(2) or perflubron dose were, however, small. CONCLUSION: The results show that derivation of Va/Q from local Po(2) measurements using fMRI in perflubron is feasible. The low detected Va/Q suggests that oxygen transport into the perflubron-filled alveolar space is significantly restrained.

DW-MRI of the urogenital tract: applications in oncology

Diffusion-weighted magnetic resonance imaging (DW-MRI) appears to hold promise as a non-invasive imaging modality in the detection of early microstructural and functional changes of different organs. DW-MRI is an imaging technique with a high sensitivity for the detection of a large variety of diseases in the urogenital tract. In kidneys, DW-MRI has shown promise for the characterization of solid lesions. Also in focal T1 hyperintense lesions DW-MRI was able to differentiate hemorrhagic cysts from tumours according to the lower apparent diffusion coefficient (ADC) values reported for renal cell carcinomas. Promising results were also published for the detection of prostate cancer. DW-MRI applied in addition to conventional T2-weighted imaging has been found to improve tumour detection. On a 3 T magnetic resonance unit ADC values were reported to be lower for tumours compared with the normal-appearing peripheral zone. The combined approach of T2-weighted imaging and DW-MRI also showed promising results for the detection of recurrent tumour in patients after radiation therapy. DW-MRI may improve the performance of conventional T2-weighted and contrast-enhanced MRI in the preoperative work-up of bladder cancer, as it may help in distinguishing superficial from muscle invasive bladder cancer, which is critical for patient management. Another challenging application of DW-MRI in the urogenital tract is the detection of pelvic lymph node metastases. As the ADC is generally reduced in malignant tumours and increased under inflammatory conditions, reduced ADC values were expected in patients with lymph node metastases.

Diffusion-weighted MR imaging of native and transplanted kidneys

Applications of diffusion-weighted (DW) magnetic resonance (MR) imaging outside the brain have gained increasing importance in recent years. Owing to technical improvements in MR imaging units and faster sequences, the need for noninvasive imaging without contrast medium administration, mainly in patients with renal insufficiency, can be met successfully by applying this technique. DW MR imaging is quantified by the apparent diffusion coefficient (ADC), which provides information on diffusion and perfusion simultaneously. By using a biexponential fitting process of the DW MR imaging data, these two entities can be separated, because this type of fitting process can serve as an estimate of both the perfusion fraction and the true diffusion coefficient. DW MR imaging can be applied for functional evaluation of the kidneys in patients with acute or chronic renal failure. Impairment of renal function is accompanied by a decreased ADC. Acute ureteral obstruction leads to perfusion and diffusion changes in the affected kidney, and renal artery stenosis results in a decreased ADC. In patients with pyelonephritis, diffuse or focal changes in signal intensity are seen on the high-b-value images, with increased signal intensity corresponding to low signal intensity on the ADC map. The feasibility and reproducibility of DW MR imaging in patients with transplanted kidneys have already been demonstrated, and initial results seem to be promising for the assessment of allograft deterioration. Overall, performance of renal DW MR imaging, presuming that measurements are of high quality, will further boost this modality, particularly for early detection of diffuse renal conditions, as well as more accurate characterization of focal renal lesions.

Diffusion-weighted magnetic resonance imaging detects local recurrence after radical prostatectomy: initial experience

Current conventional cross-sectional imaging techniques, such as contrast-enhanced computed tomography and magnetic resonance imaging (MRI), are largely inaccurate in detecting local recurrence after radical prostatectomy. We report on five patients with biochemical recurrence after radical retropubic prostatectomy and pelvic lymph node dissection for whom local recurrence could only be detected with diffusion-weighted (DW) MRI. Prior to DW-MRI, all patients had negative digital rectal examinations, negative or equivocal conventional cross-sectional imaging, and negative bone scans. All suspicious lesions on DW-MRI imaging were histologically proved to be local recurrences of prostate cancer after either transrectal ultrasound-guided or transurethral biopsy. These results should encourage other centres to test our findings.

Gadolinium diethylenetriaminepentaacetate enhancement kinetics in the menisci of asymptomatic subjects: a first step towards a dedicated dGEMRIC (delayed gadolinium-enhanced MRI of cartilage)-like protocol for biochemical imaging of the menisci

It was our aim to investigate the gadolinium diethylenetriaminepentaacetate (Gd-DTPA(2-) ) enhancement kinetics in the menisci of the knee joint over a prolonged period of time. Six asymptomatic volunteers (four men and two women; mean age, 25 ± 2.4 years) were enrolled. Sagittal, T(1) -weighted, spin-echo MR sequences of the right knee joint were obtained at 3 T. Imaging was performed before (baseline), 1 h after and in half-hour intervals up to 9 h after the intravenous administration of 0.2 mmol/kg of Gd-DTPA(2-) . To measure the rates of contrast enhancement relative to the baseline, regions of interest that covered the anterior and posterior horns of the medial and lateral meniscus were defined on each of two adjacent sections, and enhancement curves were constructed. An enhancement peak between 2.5 and 4.5 h after Gd-DTPA(2-) administration was observed, and analysis of variance also revealed no significant difference (p=0.94), in terms of enhancement, within this time interval. Pair-wise, post hoc testing also revealed no significant differences between 2.5 and 3, 3 and 3.5, 3.5 and 4, and 4 and 4.5 h post Gd-DTPA(2-) application. Our preliminary data therefore suggest that the time window suitable for a dGEMRIC (delayed gadolinium-enhanced MRI of cartilage)-like T(1) mapping of the menisci is relatively short, and lies between 2.5 and 4.5 h after Gd-DTPA(2-) injection.

Potential and limitations of diffusion-weighted magnetic resonance imaging in kidney, prostate, and bladder cancer including pelvic lymph node staging: a critical analysis of the literature

Diagnosis, staging, and treatment monitoring are still suboptimal for most genitourinary tumours. Diffusion-weighted magnetic resonance imaging (DW-MRI) has already shown promise as a noninvasive imaging modality in the early detection of microstructural and functional changes in several pathologies of various organs.