Voxel Of Interest Research Articles

MULTI-SCALE BASED EXTRACION OF VEGETATION FROM TERRESTRIAL LiDAR DATA FOR ASSESSING LOCAL LANDSCAPE

Abstract. In this study, we propose a method to accurately extract vegetation from terrestrial three-dimensional (3D) point clouds for estimating landscape index in urban areas. Extraction of vegetation in urban areas is challenging because the light returned by vegetation does not show as clear patterns as man-made objects and because urban areas may have various objects to discriminate vegetation from. The proposed method takes a multi-scale voxel approach to effectively extract different types of vegetation in complex urban areas. With two different voxel sizes, a process is repeated that calculates the eigenvalues of the planar surface using a set of points, classifies voxels using the approximate curvature of the voxel of interest derived from the eigenvalues, and examines the connectivity of the valid voxels. We applied the proposed method to two data sets measured in a residential area in Kyoto, Japan. The validation results were acceptable, with F-measures of approximately 95% and 92%. It was also demonstrated that several types of vegetation were successfully extracted by the proposed method whereas the occluded vegetation were omitted. We conclude that the proposed method is suitable for extracting vegetation in urban areas from terrestrial light detection and ranging (LiDAR) data. In future, the proposed method will be applied to mobile LiDAR data and the performance of the method against lower density of point clouds will be examined.

Glutamate as a mediating transmitter for auditory hallucinations in schizophrenia: A 1H MRS study

This is a 1H MR spectroscopy (MRS) study of glutamate (Glu), measured as Glx, levels in temporal and frontal lobe regions in patients with schizophrenia compared with a healthy control group with the objective of revealing aspects of the underlying neurochemistry of auditory hallucinations. We further compared and correlated Glu(Glx) levels for the patients-only against frequency and severity of auditory hallucinations and the sum of Positive symptoms, and also for frequency and severity of emotional withdrawal, and sum of Negative symptoms. The sample included 23 patients with an ICD-10 and DSM-IV diagnosis of schizophrenia, and 26 healthy control subjects without any known psychiatric or neurological disorders. Symptom scores were obtained from the Positive and Negative Syndrome Scale (PANSS). 1H MRS data were acquired on a 3T MR scanner from two temporal and two frontal voxels, using standard sequences and analysis parameters. The results showed that schizophrenia patients as a group had reduced Glu(Glx) levels in the voxels of interest compared to the healthy control subjects, while increased levels were found for patients with frequent and severe auditory hallucinations, relative to patients with less frequent and severe hallucination. We further found significant positive correlations between frequency and severity of auditory hallucinations, and for sum Positive symptoms, and Glu(Glx) levels in all regions, not seen when the analysis was done for negative symptoms. It is concluded that the results show for the first time that glutamate may be a mediating factor in auditory hallucinations in schizophrenia.

The partial volume effect in the quantification of 1H magnetic resonance spectroscopy in Alzheimer's disease and aging.

1H-MRS variability increases due to normal aging and also as a result of atrophy in grey and white matter caused by neurodegeneration. In this work, an automatic process was developed to integrate data from spectra and high-resolution anatomical images to quantify metabolites, taking into account tissue partial volumes within the voxel of interest avoiding additional spectra acquisitions required for partial volume correction. To evaluate this method, we use a cohort of 135 subjects (47 male and 88 female, aged between 57 and 99 years) classified into 4 groups: 38 healthy participants, 20 amnesic mild cognitive impairment patients, 22 multi-domain mild cognitive impairment patients, and 55 Alzheimer's disease patients. Our findings suggest that knowing the voxel composition of white and grey matter and cerebrospinal fluid is necessary to avoid partial volume variations in a single-voxel study and to decrease part of the variability found in metabolites quantification, particularly in those studies involving elder patients and neurodegenerative diseases. The proposed method facilitates the use of 1H-MRS techniques in statistical studies in Alzheimer's disease, because it provides more accurate quantitative measurements, reduces the inter-subject variability, and improves statistical results when performing group comparisons.

The Amygdala's Neurochemical Ratios after 12 Weeks Administration of 20 mg Long-acting Methylphenidate in Children with Attention Deficit and Hyperactivity Disorder: A Pilot Study Using (1)H Magnetic Resonance Spectroscopy.

ObjectiveRecent pediatric studies have suggested a correlation between decreased amygdala volume and attention deficit and hyperactivity disorder (ADHD) symptoms, including the emotional dysregulation. To investigate the hypothesis that medication treatment of ADHD specifically improves amygdala function, we used 1H magnetic resonance spectroscopy (MRS) to study the effect of 12 weeks of treatment with daily 20 mg long-acting methylphenidate on the Glu/Cr, NAA/Cr, Cho/Cr, and mI/Cr ratios in the amygdala of medication-naïve children with ADHD.MethodsThis was a prospective study, using a pre- and post-test design, on a single group of 21 children (average age 8.52 years, 17 males and 4 females) diagnosed with ADHD. Low Time Echo MRS scans sampled voxels of interest (1.5×1.5×2.0) from both the right and left amygdala.ResultsThere was significant clinical improvement after 12 weeks of treatment with 20 mg long-acting methylphenidate. On 1H MRS, there were no statistical significant differences of NAA/Cr ratio, Cho/Cr ratio, mI/Cr ratio before and after 12 weeks administration of 20 mg long-acting methylphenidate both in the right and left amygdala. In addition, Glu/Cr ratio decreased 14.1% in the right amygdala (p=0.029) and 11.4% in the left amygdala (p=0.008). Standardized mean effect sizes ranged from 0.14-0.32.ConclusionThe findings are consistent with the possibility that hyperglutamatergic processes in the amygdale are related to the hyperactive-impulsive symptoms of ADHD.

Restless legs syndrome and central nervous system gamma-aminobutyric acid: preliminary associations with periodic limb movements in sleep and restless leg syndrome symptom severity

BackgroundPrevious research has demonstrated abnormalities in glutamate and N-acetyl aspartate (NAA) in the thalamus in individuals with restless legs syndrome (RLS) compared with healthy matched controls. However, levels of these transmitters in other RLS-related brain areas and levels of the most common inhibitory neurotransmitter, gamma-aminobutyric acid (GABA), have not been assessed. MethodsThis study examined GABA, glutamate, and NAA levels in the dorsal anterior cingulate cortex (ACC), thalamus and cerebellum with the use of proton magnetic resonance spectroscopy (1H-MRS) at 4 tesla (4 T) and Megapress difference-editing in 18 subjects with RLS and a matched control group without RLS. Actigraphy was performed on the nights before scans to assess periodic limb movements of sleep (PLMS). ResultsLevels of GABA, glutamate, and NAA were no different between RLS and control subjects in any of the three voxels of interest. However, GABA levels were positively correlated with both PLM indices and RLS severity in the thalamus and negatively with both of these measures in the cerebellum in RLS subjects. In addition, NAA levels were higher in the ACC in RLS than in controls. ConclusionOur preliminary data suggest that known cerebellar–thalamic interactions may modulate the intensity of RLS sensory and motor symptoms. In addition, anterior cingulate cortex may be associated with the affective components of the painful symptoms in this disorder.

Zolpidem increases GABA in depressed volunteers maintained on SSRIs

Individuals with major depressive disorder (MDD) often use hypnotics like zolpidem (Ambien®) to improve sleep in addition to their selective serotonin reuptake inhibitor (SSRI) regimen. SSRIs act in part to restore disrupted GABAergic activity, but benzodiazepines and related drugs have been shown to lower GABA in a way that may be counter to these therapeutic effects. The present within-subject, single-blind, placebo-controlled study measured changes in GABA in the anterior cingulate (ACC) and thalamus of volunteers maintained on SSRIs for the treatment of MDD (n=14) following zolpidem (10mg) administration. In addition to neurochemical measurements obtained using proton magnetic resonance spectroscopy (1H MRS) at 4T, a series of questionnaires were administered to assess subjective effects associated with acute zolpidem exposure. Zolpidem elevated GABA levels in both voxels of interest (P<0.05) in the depressed participants, which could imply normalization, given the lower baseline levels associated with depression. The subjective drug experience in the depressed cohort was similar to that reported previously by healthy volunteers, and no relationships existed between GABA increases and the observed behavioral effects. Aside from treating insomnia, using zolpidem in the presence of SSRIs may have some unidentified therapeutic effects for depressed individuals.

MO‐C‐17A‐02: A Novel Method for Evaluating Hepatic Stiffness Based On 4D‐MRI and Deformable Image Registration

Purpose:Quantitative imaging of hepatic stiffness has significant potential in radiation therapy, ranging from treatment planning to response assessment. This study aims to develop a novel, noninvasive method to quantify liver stiffness with 3D strains liver maps using 4D‐MRI and deformable image registration (DIR).Methods:Five patients with liver cancer were imaged with an institutionally developed 4D‐MRI technique under an IRB‐approved protocol. Displacement vector fields (DVFs) across the liver were generated via DIR of different phases of 4D‐MRI. Strain tensor at each voxel of interest (VOI) was computed from the relative displacements between the VOI and each of the six adjacent voxels. Three principal strains (E1, E2 and E3) of the VOI were derived as the eigenvalue of the strain tensor, which represent the magnitudes of the maximum and minimum stretches. Strain tensors for two regions of interest (ROIs) were calculated and compared for each patient, one within the tumor (ROI1) and the other in normal liver distant from the heart (ROI2).Results:3D strain maps were successfully generated fort each respiratory phase of 4D‐MRI for all patients. Liver deformations induced by both respiration and cardiac motion were observed. Differences in strain values adjacent to the distant from the heart indicate significant deformation caused by cardiac expansion during diastole. The large E1/E2 (∼2) and E1/E2 (∼10) ratios reflect the predominance of liver deformation in the superior‐inferior direction. The mean E1 in ROI1 (0.12±0.10) was smaller than in ROI2 (0.15±0.12), reflecting a higher degree of stiffness of the cirrhotic tumor.Conclusion:We have successfully developed a novel method for quantitatively evaluating regional hepatic stiffness based on DIR of 4D‐MRI. Our initial findings indicate that liver strain is heterogeneous, and liver tumors may have lower principal strain values than normal liver. Thorough validation of our method is warranted in future studies.NIH (1R21CA165384‐01A1)

MO‐C‐17A‐08: Evaluation of Lung Deformation Using Three Dimensional Strain Maps

Purpose:To develop a systematic approach to generate three dimensional (3D) strain maps of lung using the displacement vector field (DVF) during the respiratory deformation, and to demonstrate its application in evaluating deformable image registration (DIR).Methods:A DVF based strain tensor at each voxel of interest (VOI) was calculated from the relative displacements between the VOI and each of the six nearest neighbors. The maximum and minimum stretches of a VOI can be determined by the principal strains (E1, E2 and E3), which are the eigenvalues and the corresponding strain tensors. Two healthy volunteers enrolled in this study under IRB‐approved protocol, each was scanned using 3D Hyperpolarized He‐3 tagging‐MRI and 3D proton‐MRI with TrueFISP sequence at the endof‐ inhalation (EOI) and the end‐of‐exhalation (EOE) phases. 3D DVFs of tagging‐ and proton‐MRI were obtained by the direct measurements of the tagging grid trajectory and by the DIR method implemented in commercial software.Results:3D strain maps were successfully generated for all DVFs. The principal strain E1s were calculated as 0.43±0.05 and 0.17±0.25 for tagging‐MRI and proton‐MRI, respectively. The large values of E1 indicate the predominant lung motion in the superior‐inferior (SI) direction. Given that the DVFs from the tagging images are considered as the ground truth, the discrepancies in the DIR‐based strain maps suggest the inaccuracy of the DIR algorithm. In the E1 maps of tagging‐MRI for subject 1, the fissures were distinguishable by the larger values (0.49±0.02) from the adjacent tissues (0.41±0.03) due to the larger relative displacement between the lung lobes.Conclusion:We have successfully developed a methodology to generate DVF‐based 3D strain maps of lung. It can potentially enable us to better understand the pulmonary biomechanics and to evaluate and improve the DIR algorithms for the lung deformation. We are currently studying more subjects to evaluate this tool.

MO‐G‐18C‐07: Improving T2 Determination and Quantification of Lipid Methylene Protons in Proton Magnetic Resonance Spectroscopy at 3 T

Purpose:To improve proton magnetic resonance spectroscopy (MRS) transverse relaxation (T2) determination and quantification of lipid methylene chain (1.3 ppm) protons by rewinding their J‐coupling evolution.Methods:MRS experiments were performed on four lipid phantoms, namely, almond, corn, sunflower and oleic acid, using a 3 T Philips MRI scanner with a transmit/receive birdcage head coil. Two PRESS (Point RESolved Spectroscopy) pulse sequences were used. The first PRESS sequence employed standard bandwidth (BW) (∼550 Hz) RF (radiofrequency) refocussing pulses, while the second used refocussing pulses of narrow BW (∼50 Hz) designed to rewind J‐coupling evolution of the methylene protons in the voxel of interest. Signal was acquired with each sequence from a 5×5×5 mm3 voxel, with a repetition time (TR) of 3000 ms, and with echo times (TE) of 100 to 200 ms in steps of 20 ms. 2048 sample points were measured with a 2000 Hz sampling bandwidth. Additionally, 30 mm outer volume suppression slabs were used to suppress signal outside the voxel of interest. The frequency of the RF pulses was set to that of the methylene resonance. Methylene peak areas were calculated and fitted in MATLAB to a monexponentially decaying function of the form M0exp(‐TE/T2), where M0 is the extrapolated area when TE = 0 ms and yields a measure of concentration.Results:The determined values of M0 and T2 increased for all fatty acids when using the PRESS sequence with narrow BW refocussing pulses. M0 and T2 values increased by an average amount (over all the phantoms) of 31% and 14%, respectively.Conclusion:This investigation has demonstrated that J‐coupling interactions of lipid methylene protons causes non‐negligible signal losses which, if not accounted for, Result in underestimations of their levels and T2 values when performing MRS measurements.Funded by the Natural Sciences and Engineering Research Council of Canada and the Canadian Breast Cancer Foundation – Prairies.NWT

Diffusion tensor imaging for target volume definition in glioblastoma multiforme

Diffusion tensor imaging (DTI) is an MR-based technique that may better detect the peritumoural region than MRI. Our aim was to explore the feasibility of using DTI for target volume delineation in glioblastoma patients. MR tensor tracts and maps of the isotropic (p) and anisotropic (q) components of water diffusion were coregistered with CT in 13 glioblastoma patients. An in-house image processing program was used to analyse water diffusion in each voxel of interest in the region of the tumour. Tumour infiltration was mapped according to validated criteria and contralateral normal brain was used as an internal control. A clinical target volume (CTV) was generated based on the T1-weighted image obtained using contrast agent (T1Gd), tractography and the infiltration map. This was compared to a conventional T2-weighted CTV (T2-w CTV). Definition of a diffusion-based CTV that included the adjacent white matter tracts proved highly feasible. A statistically significant difference was detected between the DTI-CTV and T2-w CTV volumes (p < 0.005, t = 3.480). As the DTI-CTVs were smaller than the T2-w CTVs (tumour plus peritumoural oedema), the pq maps were not simply detecting oedema. Compared to the clinical planning target volume (PTV), the DTI-PTV showed a trend towards volume reduction. These diffusion-based volumes were smaller than conventional volumes, yet still included sites of tumour recurrence. Extending the CTV along the abnormal tensor tracts in order to preserve coverage of the likely routes of dissemination, whilst sparing uninvolved brain, is a rational approach to individualising radiotherapy planning for glioblastoma patients.

Voxel-based partial volume correction for accurate quantitative voxel values

Purpose: The accuracy of voxelized information in emission imaging is limited by spatial resolution (FWHM = 2.35σ) producing biases for objects smaller than 3 FWHM. If the signal distribution is non‐uniform within 3σ of the voxel of interest then equilibrium does not exist and partial volume effect (PVE) compromises voxel accuracy. We propose a mathematical model to improve the accuracy of quantitative images of arbitrary distribution by bounding true voxel signal and estimating PVE for each voxel. Methods : A monotonically increasing parametric dataset is created for each voxel of an emission image by radial integration from the voxel center to radius = 6σ. Each cumulative integration plot from r = 3σ to 6σ is fit to a function A*4π /3*r 3 + B*ΔV derived assuming a local uniform signal distribution (A) where ΔV is the voxel volume. The constant BΔV represents the converged within 3σ integral of PVE. B > 0 implies spill‐out, B 99% in simulated 1D & 2D datasets. For the tumor, the original maximum value was 10kBq/ml. We obtained A = 3.5kBq/ml and B = 14kBq/ml for a total of 17.7kBq/ml. This yields (A+B)/original = 1.8 indicating substantial spill‐out of ~80% and a large error for the original voxel value. For a voxel in the center of the bladder, the original value was 46kBq/ml with A = 44kBq/ml, B = 7kBq/ml. (A+B)/original = 1.11 indicating near‐equilibrium at center of bladder and low spill-out of ~11% as expected. Local signal images (A) resemble low‐pass filtered original image and (B) shows the magnitude and direction of PVE. Conclusion : A new mathematical model to estimate the accuracy of voxels in quantitative images of arbitrary distribution has been developed. Analysis of additional patients is underway. ------------------------------------- Cite this article as: Mikell J, Kappadath SC. Voxel-based partial volume correction for accurate quantitative voxel values. Int J Cancer Ther Oncol 2014; 2(2):020229. DOI: 10.14319/ijcto.0202.29

Model‐based dose calculations for COMS eye plaque brachytherapy using an anatomically realistic eye phantom

To investigate the effects of the composition and geometry of ocular media and tissues surrounding the eye on dose distributions for COMS eye plaque brachytherapy with(125)I, (103)Pd, or (131)Cs seeds, and to investigate doses to ocular structures. An anatomically and compositionally realistic voxelized eye model with a medial tumor is developed based on a literature review. Mass energy absorption and attenuation coefficients for ocular media are calculated. Radiation transport and dose deposition are simulated using the EGSnrc Monte Carlo user-code BrachyDose for a fully loaded COMS eye plaque within a water phantom and our full eye model for the three radionuclides. A TG-43 simulation with the same seed configuration in a water phantom neglecting the plaque and interseed effects is also performed. The impact on dose distributions of varying tumor position, as well as tumor and surrounding tissue media is investigated. Each simulation and radionuclide is compared using isodose contours, dose volume histograms for the lens and tumor, maximum, minimum, and average doses to structures of interest, and doses to voxels of interest within the eye. Mass energy absorption and attenuation coefficients of the ocular media differ from those of water by as much as 12% within the 20-30 keV photon energy range. For all radionuclides studied, average doses to the tumor and lens regions in the full eye model differ from those for the plaque in water by 8%-10% and 13%-14%, respectively; the average doses to the tumor and lens regions differ between the full eye model and the TG-43 simulation by 2%-17% and 29%-34%, respectively. Replacing the surrounding tissues in the eye model with water increases the maximum and average doses to the lens by 2% and 3%, respectively. Substituting the tumor medium in the eye model for water, soft tissue, or an alternate melanoma composition affects tumor dose compared to the default eye model simulation by up to 16%. In the full eye model simulations, the average dose to the lens is larger by 7%-9% than the dose to the center of the lens, and the maximum dose to the optic nerve is 17%-22% higher than the dose to the optic disk for all radionuclides. In general, when normalized to the same prescription dose at the tumor apex, doses delivered to all structures of interest in the full eye model are lowest for(103)Pd and highest for (131)Cs, except for the tumor where the average dose is highest for (103)Pd and lowest for (131)Cs. The eye is not radiologically water-equivalent, as doses from simulations of the plaque in the full eye model differ considerably from doses for the plaque in a water phantom and from simulated TG-43 calculated doses. This demonstrates the importance of model-based dose calculations for eye plaque brachytherapy, for which accurate elemental compositions of ocular media are necessary.

A method for unwrapping highly wrapped multi‐echo phase images at very high field: UMPIRE

To develop a method of unwrapping phase images from multi-echo scans that works even where there are several wraps between echoes, and which generates unwrapped phase images in addition to phase difference (PD) images. The difference between the echo spacings in an acquisition with three unevenly spaced echoes (an imposed delay) can be selected such that the phase evolution in that time is in the range -π to +π in all voxels of interest. Under this condition, an image of the difference between the phase evolutions in the two inter-echo periods, an estimate of ΔB0 , is free of wraps. This ΔB0 estimate can be used to identify and remove receiver phase offsets and wraps in phase images. The approach was tested on simulated data and high-resolution in vivo brain data acquired from six subjects at 7 Tesla. The method generated wrap-free phase images. It was able to remove more wraps than is possible with PD imaging and was faster and more reliable than spatial unwrapping. Unwrapping Multi-echo Phase Images with iRregular Echo spacings (UMPIRE) is conceptually simple, fast, reliable, and requires no fitting, thresholds, or operator intervention.

Changes in Functional Integration with the Non-Epileptic Temporal Lobe of Patients with Unilateral Mesiotemporal Epilepsy

PurposeTo investigate epilepsy-induced changes in effective connectivity between the non-epileptic amygdalo-hippocampal complex (AHC) and the rest of the brain in patients with unilateral mesiotemporal lobe epilepsy (MTLE) associated with hippocampal sclerosis (HS).MethodsThirty-three patients with unilateral MTLE associated with HS (20 females, mean age: 36 years, 19 left HS) and 33 adult controls matched for age and gender underwent 18F-Fluorodeoxyglucose positron emission tomography (FDG-PET). Right-HS patients' FDG-PET data were flipped to obtain a left–epileptic–focus–lateralized group of patients. Voxels of interest (VOI) were selected within the cytoarchitectonic probabilistic maps of the non-epileptic AHC (probability level = 100%, SPM8 Anatomy toolbox v1.7). Patients and controls were compared using VOI metabolic activity as covariate of interest to search for epilepsy-induced changes in the contribution of the non-epileptic AHC to the level of metabolic activity in other brain areas. Age, gender, duration of epilepsy, seizure type and frequency were used as covariates of no-interest for connectivity analyses.Key findingsSignificant decrease in effective connectivity was found between the non-epileptic AHC and ventral prefrontal cortical areas bilaterally, as well as with the temporal pole and the posterior cingulate cortex contralateral to HS. Significant increase in connectivity was found between the non-epileptic AHC and midline structures, such as the anterior cingulate and dorsal medial prefrontal cortices, as well as the temporo-parietal junction bilaterally. Connectivity analyses also revealed a preserved positive connectivity between the non-epileptic and the epileptic AHC in the patients' group.SignificanceThis study evidences epilepsy-induced changes in connectivity between the non-epileptic AHC and some limbic and default mode network areas. These changes in connectivity probably account for emotional, cognitive and decision-making impairments frequently observed in MTLE patients. The preserved neurometabolic connectivity between the non-epileptic and the epileptic AHC in MTLE patients is pivotal to explain the epilepsy-induced changes found in this study.

문맥 정보를 이용한 분류 기반 무릎 뼈 검출 기법

본 논문에서는 영상 내의 문맥 특징(context feature)과 외형 특징(appearance feature)을 함께 학습함으로써 의료영상 내의 비슷한 외형 특징을 가지는 장기들을 자동으로 검출하는 기법을 제안한다. 기존 검출 기법들은 외형 특징 정보만을 학습하여 분류기(classifier)를 생성하였기 때문에 의료영상 내에 외형이 비슷한 장기들이 다수 포함되어 있는 경우 검출 오류가 발생하였다. 제안하는 기법은 외형 특징을 이용하여 학습된 분류기를 통해 얻은 확률 값들을 바탕으로 관심 복셀(voxel) 주변의 확률 분포 특징을 반복적으로 학습함으로써 문맥 정보를 포함하는 분류기를 생성한다. 또한, 실험 단계(test stage)에서 '지역 기반 투표 방식'(region based voting scheme)을 도입함으로써 효율성과 정확성을 향상시킨다. 제안하는 기법의 성능 평가를 위해 SKI10 무릎 관절 데이터 셋 내에서 외형 특징이 비슷한 대퇴골(femur)과 경골(tibia)을 검출하는 실험을 진행하였다. 실험 결과를 통해 제안하는 기법이 외형 특징만을 이용했던 검출 기법에 비해 개선된 검출 성능을 보이고 있음을 확인할 수 있었다. In this paper, we propose a method that automatically detects organs having similar appearances in medical images by learning both context and appearance features. Since only the appearance feature is used to learn the classifier in most existing detection methods, detection errors occur when the medical images include multiple organs having similar appearances. In the proposed method, based on the probabilities acquired by the appearance-based classifier, new classifier containing the context feature is created by iteratively learning the characteristics of probability distribution around the interest voxel. Furthermore, both the efficiency and the accuracy are improved through 'region based voting scheme' in test stage. To evaluate the performance of the proposed method, we detect femur and tibia which have similar appearance from SKI10 knee joint dataset. The proposed method outperformed the detection method only using appearance feature in aspect of overall detection performance.

ICA-based artifact correction improves spatial localization of adaptive spatial filters in MEG

Beamformers are one of the most common inverse models currently used in the estimation of source activity from magnetoencephelography (MEG) data. They rely on a minimization of total power while constraining the gain in the voxel of interest, resulting in the suppression of background noise. Nonetheless, in cases where background noise is strong compared to the source of interest, or when many sources are present, the ability of the beamformer to detect and accurately localize weak sources is reduced. In visual paradigms, two main background sources can substantially impact an accurate estimation of weaker sources. Ocular artifacts are orders of magnitude higher than neural sources making it difficult for the beamformer to effectively suppress them. Primary visual activations also result in strong signals that can impede localization of weak sources. In this paper, we systematically evaluated how neural (visual) and non-neural (eye, heart) sources affect the localization accuracy of frontal and medial temporal sources in visual tasks. These sources are of tremendous interest in learning and memory studies as well as in clinical settings (Alzheimer's/epilepsy) and are typically difficult to localize robustly in MEG. Empirical data from two tasks - active learning and control - were used to evaluate our analysis techniques. Global field power calculations showed multiple time periods where active learning was significantly different from response selection with dominant sources converging to the eyes. Extensive leakage of eye activity into frontal and visual that evoked responses into parietal cortices was also observed. Contributions from ocular activity to the reconstructed time series were indiscernible from task-based recruitment of frontal sources in the original data. Removing artifacts (eye movements, cardiac, and muscular) by means of independent component analysis (ICA) led to a significant improvement in detection and localization of frontal and medial temporal sources. We verified our results by using simulations of sources placed in frontal and medial temporal regions with various types of background noise (eye, heart, and visual). We report that the detection and localization accuracy of frontal and medial temporal sources with beamformer techniques is highly dependent on the magnitude and location of background sources and that removing artifacts can substantially improve the beamformer's performance.

Intraoperative magnetic resonance spectroscopy for identification of residual tumor during low-grade glioma surgery

The authors had previously shown that 3-T intraoperative MRI (ioMRI) detects residual tumor tissue during low-grade glioma and that it helps to increase the extent of resection. In a proportion of their cases, however, the ioMRI disclosed T2-hyperintense areas at the tumor resection border after the initial resection attempt and prompted a differential diagnosis between residual tumor and nontumoral changes. To guide this differential diagnosis the authors used intraoperative long-TE single-voxel proton MR spectroscopy (ioMRS) and tested the correlation of these findings with findings from pathological examination of resected tissue. Patients who were undergoing surgery for hemispheric or insular WHO Grade II gliomas and were found to have T2 changes around the resection cavity at the initial ioMRI were prospectively examined with ioMRS and biopsies were taken from corresponding localizations. In 14 consecutive patients, the ioMRS diagnosis in 20 voxels of interest was tested against the histopathological diagnosis. Intraoperative diffusion-weighted imaging (ioDWI) was also performed, as a part of the routine imaging, to rule out surgically induced changes, which could also appear as T2 hyperintensity. Presence of tumor was documented in 14 (70%) of the 20 T2-hyperintense areas by histopathological examination. The sensitivity of ioMRS for identifying residual tumor was 85.7%, the specificity was 100%, the positive predictive value was 100%, and the negative predictive value was 75%. The specificity of ioDWI for surgically induced changes was high (100%), but the sensitivity was only 60%. This is the first clinical series to indicate that ioMRS can be used to differentiate residual tumor from nontumoral changes around the resection cavity, with high sensitivity and specificity.

Abstract TMP39: Identifying Cerebrovascular Reserve Capacity by Arterial Spin Labeling MR in Patients with Moyamoya Disease

Background— The diagnostic accuracy to identify the cerebrocascular reserve capacity (CVRC) by arterial spin labeling (ASL) MR in patients with moyamoya disease is unclear. Methods— We conducted a prospective study to examine the accuracy of ASL as compared with single-photon emission CT (SPECT) in patients with moyamoya disease. Preoperative ASL and SPECT images from 61 adult patients with moyamoya disease were analyzed. For 122 hemispheres from the patients, voxels of interest consisting of bilateral ICA and MCA territories were applied to the spatially normalized cerebral blood flow maps from basal and acetazolamide challenge ASL and SPECT. The correlation between cerebrovascular reserve capacity (CVRC) from ASL and SPECT was analyzed by a using Pearson correlation analysis and the area under the receiver-operating-characteristic curve (AUC) was used to evaluate diagnostic accuracy relative to that of SPECT for various cutoff points by CVRC. Results— Significant relationships were observed between CVRC from ASL and SPECT for the internal carotid artery (ICA) and middle cerebral artery (MCA) (correlation coefficient r = .7093, P &lt; .0001 and r = .7319, P &lt; .0001, respectively) (Table). The diagnostic accuracies of ASL for detecting ICA and MCA areas, in which CVRC are &lt; -5.0%, revealed an AUC of 0.96 (95% confidence interval [CI], 0.91 to 0.99) and 0.96 (95% CI, 0.91 to 0.99, respectively). Conclusions— ASL can identify the reduced CVRC with good performance and has the potential as a noninvasive imaging tool for determining CVRC in patients with moyamoya disease. Table. Correlations between CBF and CVRI values from ASL and SPECT

An NMR Phantom Mimicking Intramyocellular (IMCL) and Extramyocellular Lipids (EMCL)

Intramyocellular lipids (IMCL) play an important role in muscle metabolism. 1H magnetic resonance spectroscopy is the method of choice for non-invasive assessment of IMCL. However, IMCL quantitation is hampered by the larger overlapping resonances of extramyocellular lipids (EMCL). A phantom that mimics EMCL and IMCL, i.e., the 0.2-ppm resonance splitting, would be useful for testing acquisition strategies and post-processing algorithms. Here, we propose a phantom that consists of a cylindrical bottle filled with dairy cream and sunflower oil. Similar to EMCL, the oil (CH2)n protons resonate at 1.5 ppm; similar to IMCL, the spherical shape of droplets in cream results in (CH2)n protons resonating at 1.3 ppm. The relative amount of IMCL versus EMCL can be easily controlled in a systematic and exact fashion by displacing the voxel of interest across the cream–oil interface. This phantom is of simple construction and made of inexpensive and readily available materials, and should be of value in testing both acquisition and spectral analysis strategies in the context of ICML/ECML studies.

De-noising, phase ambiguity correction and visualization techniques for complex-valued ICA of group fMRI data

Analysis of functional magnetic resonance imaging (fMRI) data in its native, complex form has been shown to increase the sensitivity both for data-driven techniques, such as independent component analysis (ICA), and for model-driven techniques. The promise of an increase in sensitivity and specificity in clinical studies, provides a powerful motivation for utilizing both the phase and magnitude data; however, the unknown and noisy nature of the phase poses a challenge. In addition, many complex-valued analysis algorithms, such as ICA, suffer from an inherent phase ambiguity, which introduces additional difficulty for group analysis. We present solutions for these issues, which have been among the main reasons phase information has been traditionally discarded, and show their effectiveness when used as part of a complex-valued group ICA algorithm application. The methods we present thus allow the development of new fully complex data-driven and semi-blind methods to process, analyze, and visualize fMRI data.We first introduce a phase ambiguity correction scheme that can be either applied subsequent to ICA of fMRI data or can be incorporated into the ICA algorithm in the form of prior information to eliminate the need for further processing for phase correction. We also present a Mahalanobis distance-based thresholding method, which incorporates both magnitude and phase information into a single threshold, that can be used to increase the sensitivity in the identification of voxels of interest. This method shows particular promise for identifying voxels with significant susceptibility changes but that are located in low magnitude (i.e. activation) areas. We demonstrate the performance gain of the introduced methods on actual fMRI data.