Voxel Count Research Articles

Evaluating the impact of downsampling on 3D MRI images segmentation results based on similarity metrics

Medical imaging plays a crucial role in diagnosing patient conditions, with magnetic resonance imaging (MRI) standing as a significant modality for numerous years. However, leveraging convolutional neural network (CNN) architectures like U-Net and its variations for anatomical segmentation demands considerable memory, particularly when working with full 3D image sets. Therefore, downsampling 3D MRIs proves advantageous in reducing memory consumption. Nevertheless, downsampling leads to a reduction in voxel count, potentially impacting the performance of commonly used segmentation metrics. The jaccard similarity index (JSI), dice similarity coefficient (DSC), and structural similarity index (SSIM) are extensively employed in image segmentation contexts. Hence, this study employs all three metrics to assess downsampled images and evaluate the robustness of the metrics when used to evaluate the downsampled 3D MRI images. The results show that JSI and DSC are more robust than SSIM when handling the downsampled data.

Could DTI Unlock the Mystery of Subjective Tinnitus: It's Time for Parameters That Go A Little Out of the Routine.

In this study, it was aimed to assess the microstructural changes in the main central auditory pathway in cases with subjective tinnitus. In total, 101 subjects (52 cases with bilateral subjective non-pulsatile tinnitus and 49 healthy cases as the control group) were included in the study. Participants underwent pure tone audiogram and Diffusion Tensor Imaging-Magnetic Resonance Imaging (DTI-MRI) examination with a 3 Tesla MRI device. The number of tracts, tract length, volume, and quantitative anisotropy (QA) and normalized quantitative anisotropy' (nQA) values were calculated by plotting cochleocortical pathways from the cochlear nerve to ipsilateral and contralateral Heschl's gyrus (HG). In pure tone audiometry, the control group had lower hearing thresholds than cases with tinnitus. Fibres and nQA values from the right cochlear nerve to the right HG were significantly lower in the tinnitus group than in the control group. Cochlear nuclei voxel counts were significantly decreased in the tinnitus group. Both cochlear nucleus volumes were higher in the tinnitus group than in the control group. nQA values in both cochlear nuclei were decreased in the tinnitus group. This study showed that the most commonly affected part in subjective non-pulsatile tinnitus cases is the cochlear nucleus. Therefore, the cochlear nucleus should be evaluated more carefully in cases presenting with subjective tinnitus.

Challenges and possibilities of the manual metallographic serial sectioning process using the example of a quantitative microstructural analysis of graphite in cast iron

Abstract The 3D microstructure analysis presented in this study focuses on the nodular graphite of an EN-GJS grade cast iron. The shape is analyzed based on shape factor and aspect ratio as well as average particle size and distribution. The shape of graphites in cast iron materials is critical to the mechanical properties of these alloys and is essential for the characterization of the material. Metallographic serial sectioning creates a digital twin of the material, which is removed layer by layer and visualized layer by layer under an optical microscope. From this three-dimensional twin, any number of planes can be digitally projected and analyzed. Conventional quantitative 3D analysis examines the voxel count and composition of a body in 3D space. The analysis presented here applies stereological 2D analysis methods to three spatial planes of the material. Two of the planes are digital projections of the twin. The two reconstructed planes of the material are chosen so that the direction vectors of all three planes form an angle of 90° to each other. The methodology is described in detail and the challenges and opportunities of the serial section method presented here are discussed.

The Effect of Ankle Muscles Dry Needling on Brain Activity Map Based on fMRI: a Study Protocol for Randomized Controlled Trial.

Neuromodulation may be one of the underlying mechanisms of dry needling (DN); however, the mechanism has not yet been fully clarified. This randomized controlled trial is designed to evaluate DN stimulation of the tibialis anterior and peroneus longus muscles in chronic ankle instability (CAI) and healthy subjects, employing functional magnetic resonance imaging (fMRI). Clinical study protocol, SPIRIT compliant. Brain Mapping Laboratory. A total of thirty participants aged between 18 and 40 years old will be included in this study. Twenty healthy participants will be randomized into 2 groups (real DN and sham DN). Ten patients with CAI will also be recruited to the third group and receive only real DN for comparison. Real and sham DN. The voxel count, coordinates of peak activation, and peak intensity will be obtained as primary outcomes to report brain map activation. Measurements will be taken before, during, and after DN treatment. The strength of the ankle dorsiflexors, active dorsiflexion range of motion, and McGill pain questionnaire short-form will be used as secondary outcome measures. The results from this study will be published in peer-reviewed journals and disseminated as presentations at national and international congresses. This trial will explore brain responses to real and sham DN in healthy participants and to real DN in CAI patients. Overall, our results will provide preliminary evidence of the neural mechanism of DN.

Fast, Nondestructive and Precise Biomass Measurements Are Possible Using Lidar-Based Convex Hull and Voxelization Algorithms

Light detection and ranging (lidar) scanning tools are available that can make rapid digital estimations of biomass. Voxelization and convex hull are two algorithms used to calculate the volume of the scanned plant canopy, which is correlated with biomass, often the primary trait of interest. Voxelization splits the scans into regular-sized cubes, or voxels, whereas the convex hull algorithm creates a polygon mesh around the outermost points of the point cloud and calculates the volume within that mesh. In this study, digital estimates of biomass were correlated against hand-harvested biomass for field-grown corn, broom corn, and energy sorghum. Voxelization (r = 0.92) and convex hull (r = 0.95) both correlated well with plant dry biomass. Lidar data were also collected in a large breeding trial with nearly 900 genotypes of energy sorghum. In contrast to the manual harvest studies, digital biomass estimations correlated poorly with yield collected from a forage harvester for both voxel count (r = 0.32) and convex hull volume (r = 0.39). However, further analysis showed that the coefficient of variation (CV, a measure of variability) for harvester-based estimates of biomass was greater than the CV of the voxel and convex-hull-based biomass estimates, indicating that poor correlation was due to harvester imprecision, not digital estimations. Overall, results indicate that the lidar-based digital biomass estimates presented here are comparable or more precise than current approaches.

Intra-Atlas Node Size Effects on Graph Metrics in fMRI Data: Implications for Alzheimer's Disease and Cognitive Impairment.

Network neuroscience, a multidisciplinary field merging insights from neuroscience and network theory, offers a profound understanding of neural network intricacies. However, the impact of varying node sizes on computed graph metrics in neuroimaging data remains underexplored. This study addresses this gap by adopting a data-driven methodology to delineate functional nodes and assess their influence on graph metrics. Using the Neuromark framework, automated independent component analysis is applied to resting state fMRI data, capturing functional network connectivity (FNC) matrices. Global and local graph metrics reveal intricate connectivity patterns, emphasizing the need for nuanced analysis. Notably, node sizes, computed based on voxel counts, contribute to a novel metric termed 'node-metric coupling' (NMC). Correlations between graph metrics and node dimensions are consistently observed. The study extends its analysis to a dataset comprising Alzheimer's disease, mild cognitive impairment, and control subjects, showcasing the potential of NMC as a biomarker for brain disorders. The two key outcomes underscore the interplay between node sizes and resultant graph metrics within a given atlas, shedding light on an often-overlooked source of variability. Additionally, the study highlights the utility of NMC as a valuable biomarker, emphasizing the necessity of accounting for node sizes in future neuroimaging investigations. This work contributes to refining comparative studies employing diverse atlases and advocates for thoughtful consideration of intra-atlas node size in shaping graph metrics, paving the way for more robust neuroimaging research.

Comparison of brain volume estimation via cerebral computed tomography and magnetic resonance imaging

AbstractBackgroundIn clinical routine, cerebral computed tomography (cCT) is easily accessible and more commonly used than brain magnetic resonance imaging (cMRI). However, to the best of our knowledge a fully automated, validated regional specific volume estimation tool is lacking. Therefore, the aim of this study was to investigate the association between tissue classification for gray matter of both modalities.MethodBrain imaging segmentations of cCT and structural MRI of 37 healthy controls and 20 patients with mild cognitive impairment due to Alzheimer’s disease were performed. The intensity and voxel count of normalized gray matter maps (MNI) in regions of interest using the Automated Anatomical Labelling Atlas (AAL3) were computed and compared between cCT and cMRI in both cohorts.ResultBrain Volume, measured by intensity and voxel count of gray matter, is associated between cortical regions acquired by cCT and cMRI. Although correlation coefficients varied between different brain areas, cortical compared to subcortical regions showed a relatively strong correlation.ConclusionThe evaluation of brain atrophy using segmented brain images acquired by cCT may be a simple, cost‐effective, and valuable addition to the clinical workflow, particularly in the context of socioeconomic and health disparities.

Comparison of brain volume estimation via cerebral computed tomography and magnetic resonance imaging

AbstractBackgroundIn clinical routine, cerebral computed tomography (cCT) is easily accessible and more commonly used than brain magnetic resonance imaging (cMRI). However, to the best of our knowledge a fully automated, validated regional specific volume estimation tool is lacking. Therefore, the aim of this study was to investigate the association between tissue classification for gray matter of both modalities.MethodBrain imaging segmentations of cCT and structural MRI of 37 healthy controls and 20 patients with mild cognitive impairment due to Alzheimer’s disease were performed. The intensity and voxel count of normalized gray matter maps (MNI) in regions of interest using the Automated Anatomical Labelling Atlas (AAL3) were computed and compared between cCT and cMRI in both cohorts.ResultBrain Volume, measured by intensity and voxel count of gray matter, is associated between cortical regions acquired by cCT and cMRI. Although correlation coefficients varied between different brain areas, cortical compared to subcortical regions showed a relatively strong correlation.ConclusionThe evaluation of brain atrophy using segmented brain images acquired by cCT may be a simple, cost‐effective, and valuable addition to the clinical workflow, particularly in the context of socioeconomic and health disparities.

A new 3D phase unwrapping method by region partitioning and local polynomial modeling in abdominal quantitative susceptibility mapping.

Accurate phase unwrapping is a critical prerequisite for successful applications in phase-related MRI, including quantitative susceptibility mapping (QSM) and susceptibility weighted imaging. However, many existing 3D phase unwrapping algorithms face challenges in the presence of severe noise, rapidly changing phase, and open-end cutline. In this study, we introduce a novel 3D phase unwrapping approach utilizing region partitioning and a local polynomial model. Initially, the method leverages phase partitioning to create initial regions. Noisy voxels connecting areas within these regions are excluded and grouped into residual voxels. The connected regions within the region of interest are then reidentified and categorized into blocks and residual voxels based on voxel count thresholds. Subsequently, the method sequentially performs inter-block and residual voxel phase unwrapping using the local polynomial model. The proposed method was evaluated on simulation and in vivo abdominal QSM data, and was compared with the classical Region-growing, Laplacian_based, Graph-cut, and PRELUDE methods. Simulation experiments, conducted under different signal-to-noise ratios and phase change levels, consistently demonstrate that the proposed method achieves accurate unwrapping results, with mean error ratios not exceeding 0.01%. In contrast, the error ratios of Region-growing (N/A, 84.47%), Laplacian_based (20.65%, N/A), Graph-cut (2.26%, 20.71%), and PRELUDE (4.28%, 10.33%) methods are all substantially higher than those of the proposed method. In vivo abdominal QSM experiments further confirm the effectiveness of the proposed method in unwrapping phase data and successfully reconstructing susceptibility maps, even in scenarios with significant noise, rapidly changing phase, and open-end cutline in a large field of view. The proposed method demonstrates robust and accurate phase unwrapping capabilities, positioning it as a promising option for abdominal QSM applications.

Quantitative T1 brain mapping in early relapsing-remitting multiple sclerosis: longitudinal changes, lesion heterogeneity and disability

ObjectivesTo quantify brain microstructural changes in recently diagnosed relapsing-remitting multiple sclerosis (RRMS) using longitudinal T1 measures, and determine their associations with clinical disability.MethodsSeventy-nine people with recently diagnosed (< 6 months) RRMS were recruited from a single-centre cohort sub-study, and underwent baseline and 1-year brain MRI, including variable flip angle T1 mapping. Median T1 was measured in white matter lesions (WML), normal-appearing white matter (NAWM), cortical/deep grey matter (GM), thalami, basal ganglia and medial temporal regions. Prolonged T1 (≥ 2.00 s) and supramedian T1 (relative to cohort WML values) WML voxel counts were also measured. Longitudinal change was assessed with paired t-tests and compared with Bland-Altman limits of agreement from healthy control test-retest data. Regression analyses determined relationships with Expanded Disability Status Scale (EDSS) score and dichotomised EDSS outcomes (worsening or stable/improving).ResultsSixty-two people with RRMS (mean age 37.2 ± 10.9 [standard deviation], 48 female) and 11 healthy controls (age 44 ± 11, 7 female) contributed data. Prolonged and supramedian T1 WML components increased longitudinally (176 and 463 voxels, respectively; p < .001), and were associated with EDSS score at baseline (p < .05) and follow-up (supramedian: p < .01; prolonged: p < .05). No cohort-wide median T1 changes were found; however, increasing T1 in WML, NAWM, cortical/deep GM, basal ganglia and thalami was positively associated with EDSS worsening (p < .05).ConclusionT1 is sensitive to brain microstructure changes in early RRMS. Prolonged WML T1 components and subtle changes in NAWM and GM structures are associated with disability.Clinical relevance statementMRI T1 brain mapping quantifies disability-associated white matter lesion heterogeneity and subtle microstructural damage in normal-appearing brain parenchyma in recently diagnosed RRMS, and shows promise for early objective disease characterisation and stratification.Key Points• Quantitative T1 mapping detects brain microstructural damage and lesion heterogeneity in recently diagnosed relapsing-remitting multiple sclerosis.• T1 increases in lesions and normal-appearing parenchyma, indicating microstructural damage, are associated with worsening disability.• Brain T1 measures are objective markers of disability-relevant pathology in early multiple sclerosis.Graphical abstract

A novel analysis of the formation and resorption changes in dental hard tissue using longitudinal in vivo micro computed tomography.

This study was to investigate the new analysis manner of dental hard tissue change using in vivo micro-computed tomography (CT) in rat. Scanning, registration, analyzing, and presenting method to track longitudinal in vivo micro-CT data on dental hard tissues were validated in murine models: formative, dentin thickness after direct pulp capping with mineral trioxide aggregate; resorptive, development of apical bone rarefaction in apical periodontitis model. Serial in vivo micro-CT scans were analyzed through rigid-registration, active-contouring, deformable-registration, and motion vector-based quantitative analyses. The rate and direction of hard tissue formation after direct pulp capping was datafied by tracing coordinate shift of fiducial points on pulp chamber outline in formative model. The development of apical periodontitis could be monitored with voxel counts, and quantitatively analyzed in terms of lesion size, bone loss, and mineral density in resorptive model. This study supports the application of longitudinal in vivo micro-CT for resorptive- and formative-phase specific monitoring of dental hard tissues.

An In-House 3D Voxel Dosimetric Tool to Compare Predictive and Post- Treatment Dosimetry in 90Y Radioembolization: A Proof of Concept.

The aim of this study was to implement an in-house dosimetric tool to assess tumour- absorbed doses in pre and post-dosimetry for 90Y radioembolization with resin spheres. To perform dosimetric calculations we set up a dosimetric procedure and developed homemade software to calculate tumour absorbed dose and dose volume histograms (DVHs). The method is based on a simplified voxel dosimetry for an estimated 3D absorbed dose and it can be applied to both 99mTc-MAA SPECT/CT and 90Y PET/CT acquisitions for pre and post-dosimetry. We tested the software performance in a retrospective study using the data of 22 patients with hepatocellular carcinoma who underwent radioembolization with 90Y resin spheres in the period 2016-2021. The software calculates tumour doses (mean, minimum and maximum doses) from voxel counts and dose-volume histograms (DVH_spect, DVH_pet) for both 99mTc-MAA SPECT/CT and 90Y PET/CT imaging. DVH_spect and DVH_pet data were analyzed and compared with the aim to assess an agreement between them. Concordance between dosimetric data were evaluated with the Wilcoxon Signed Ranked test, descriptive statistical analysis and Pearson correlation coefficient. The mean administrated activity was 1313 MBq (range 444 MBq - 2200 MBq). Tumour volumes ranged from 75 mL to 1012 mL. The mean absorbed dose for tumour volume was 161 ± 66 Gy (Dm_spect) and 173 ± 79 Gy (Dm_pet). From Wilcoxon Signed Rank Test the differences between the dosimetric data extrapolated from DVH_spect and DVH_pet results were not significant with α = 0.05 (two-sided test). A good linear correlation was found between 99mTc-MAA and 90Y dosimetric data (Pearson correlation coefficient 0.887 p < 0.001). Generally, DVHs calculated on 99mTc-MAA SPECT/CT and 90Y PET/CT gave comparable results, some discrepancies were observed particularly with those patients where SPECT and PET imaging presented a visual mismatching. A simplified 3D dosimetry methodology was implemented and tested retrospectively on patient data treated with 90Y resin spheres. Even if the clinical feasibility of our approach has to be further validated on an extended patient cohort, the preliminary results of our study highlight the potential of the implemented dosimetric tool for tumour dose assessment.

Assessment of understory vegetation in a plantation forest of the southeastern United States using terrestrial laser scanning

Forests of the southeastern United States are home to large timber industries with substantial contributions to global round wood production and paper products. Despite the success of plantations and the large timber industries in this area, pine growth remains constrained due to the competition between planted pine and the species in the understory. Moreover, effective control of this interspecies competition had shown a significant two- to four-times increase in stand productivity. Thus, this study aims to evaluate the use of laser scanning derived data from Terrestrial Laser Scanner (TLS) to assess understory vegetation biomass, as conventional methods utilizing optical imagery have yet to be effective in quantifying and mapping evergreen understory in the southeastern coastal forests of the United States. For this study, we utilized TLS to scan the entire forest profile within 60 sample plots of an operational loblolly pine plantation in Nassau County, Florida, and collected understory biomass data through destructive sampling. We compared three TLS-based volume estimation methods for predicting understory biomass and applied the Adaptive Least Absolute Shrinkage and Selection Operator (ALASSO) regression method to derive an optimal model by integrating the most efficient volume estimation methods and other TLS-derived standard metrics. Our study identifies the 20th percentile of the echo height and a 3D volume metric based on mean height and understory cover as the most significant explanatory variables for the optimal model. The model exhibits high accuracy, with Adjusted R-squared (Adj. R2) of 0.80 and Root Mean Square Error (RMSE) of 234.8 g per square meter (g/m2). Additionally, the mean height and understory-based volume estimation method outperformed other methods, such as voxel count and three dimensional (3D) alpha hall-based method, with Adj. R2 of 0.79, 0.47, and 0.57, and RMSE values of 288, 448.6, and 413 g/m2, respectively, when used as a single variable in the model. The resulting model successfully predicted and quantified understory vegetation, showcasing TLS's potential to accurately capture biomass variation, particularly in evergreen-dominated pine plantation forests in the southeastern United States coastal regions. As a tool for monitoring understory, TLS can be used to aid plantation forest managers in identifying areas that require control measures for enhanced management practices.

Identifying sex-specific risk architectures for predicting amyloid deposition using neural networks

In older adults without dementia, White Matter Hyperintensities (WMH) in MRI have been shown to be highly associated with cerebral amyloid deposition, measured by the Pittsburgh compound B (PiB) PET. However, the relation to age, sex, and education in explaining this association is not well understood. We use the voxel counts of regional WMH, age, one-hot encoded sex, and education to predict the regional PiB using a multilayer perceptron with only rectilinear activations using mean squared error. We then develop a novel, robust metric to understand the relevance of each input variable for prediction. Our observations indicate that sex is the most relevant predictor of PiB and that WMH is not relevant for prediction. These results indicate that there is a sex-specific risk architecture for Aβ deposition.

Estimation of Age from Pulp Chamber Volume of First Molars and Second Premolars using CBCT in Central Indian Population

Introduction: Teeth form a suitable parameter for age estimation as they disintegrate very slowly and are impervious to injuries. CBCT imaging not only allows observation of changes within teeth but permits observations of living specimens. Pulp chamber size may show a significant relationship with chronological age. Aim: To estimate the age from pulp chamber volume of first molars and second premolars using CBCT and compare it to the actual age. Materials and Methods: The current study comprised 115 subjects which were divided into four age groups. DICOM CBCT images were exported to ITK SNAP 3.8 software. Pulp chamber volumes were measured using voxel counting and image segmentation tools. Results: Molars predict the age with an accuracy of 55% and 52.1%, whereas premolars predict the age with an accuracy of 35% and 33.4% in males and females, respectively. Discriminant equations are derived for maxillary and mandibular molars and premolars irrespective of age groups, age-wise, and gender-wise groups for age estimation. Conclusion: Maxillary molars and mandibular premolars are found as strong predictors of age. CBCT imaging has proved as an excellent tool with precision and accuracy and is recommended for use in age estimation.

Serotonin Reuptake Inhibitors are Associated with Less Amyloid‐β Burden Spatial Extent in Military Veterans with only PTSD, but not with TBI or Comorbid PTSD/TBI: Preliminary PET Findings from ADNI‐DOD Project

AbstractBackgroundSome but not all studies suggested an association of serotonin reuptake inhibitors (SRIs) with a reduction in “magnitude” of amyloid‐β (Aβ) plaque burden. Using PET images from ADNI‐DOD and a voxel‐based Majority Count Algorithm (MCA), we demonstrated that cognitively unimpaired veterans with post‐traumatic stress disorder (PTSD) had significantly more cerebral voxels with lower rather than higher florbetapir standard uptake value ratios (voxel‐based P&lt;0.05), reflecting Aβ burden reduction in “spatial extent” [Chen, AAIC 2018]). We postulated that this finding might relate their SRI use. Here, we investigate the impact of SRI use on the spatial extent of Aβ burden among ADNI‐DOD veterans with PTSD, comorbid PTSD and traumatic brain injury (TBI), and TBI.MethodsStatistical Brain Mapping (SPM12) adjusted for age, MMSE, education and APOE4 allele, a voxel‐based P&lt;0.05 threshold, and MCA which is free of multi‐comparisons (Stern, NEJM, 2019) were used to compare the greater versus lower number of cerebral voxels with higher Aβ burden in SRI non‐users versus users in 6 groups: 1) PTSD (n=76), 2) PTSD or TBI (n=115), 3) comorbid PTSD/TBI or TBI (n=126), 4) PTSD, PTSD/TBI or TBI (n=202), 5) PTSD/TBI (n=87) and 6) TBI (n=39).ResultsSRI non‐users had a significantly larger number of cerebral voxels with higher Aβ burden than users in the first four groups: 1) 28,539 in non‐users versus 70 in users (voxel count ratio VCR=551) in PTSD; 2) 3,573 versus 21 (VCR=170) in PTSD or TBI; 3) 22,927 versus 1,162 (VCR=20) in comorbid PTSD/TBI or TBI; and 4) 7,569 versus 1,481 (VCR=5) in PTSD, comorbid PTSD/TBI or TBI, MCA p&lt;0.001 for all 4 groups. In contrast, 5) SRI non‐users and users did not differ in the number of cerebral voxels with higher Aβ burden in PTSD/TBI (3,209 versus 4,027, VCR=1, MCA p=0.68); and 6) non‐users had a significantly smaller number of cerebral voxels than users with higher Aβ burden in TBI (60 versus 11,715 VCR=0.005, MCA p&lt;0.001).ConclusionThis study showed SRIs’ association with spatially less extensive Aβ burden in PTSD, more extensive Aβ burden in TBI, and intermediate/opposing effects in other combinations. Finding conformation and mechanism clarification are needed.

Mandibular Teeth as Predictors in Forensic Age Estimation: A Cone-Beam Computed Tomography-Based Pulp Volume Regression Study.

Age estimation is integral to science of forensic odontology and plays an important role in human identification. Pulp volume estimation using cone-beam computed tomography (CBCT) to derive age can be very useful as it is noninvasive and can be applied on individuals of all ages. The study aimed to estimate the chronological age-based pulp and tooth volume (PTV) ratios in adults from Uttar Pradesh, India, using CBCT and voxel counting dental software. Thirty-six CBCT scans were allocated into three Groups, I, II, and III (12 in each group) randomly to study the pulp volume of lateral incisor, canine, and first premolar, respectively. PTV was calculated by image segmentation method using Xelis Dental software. A correlation between chronological age with PTV was established using derived regression equations. All the data were subjected to statistical analysis using Statistical Package for Social Sciences version 16.0. Estimated age and standard error were determined. Standard errors of estimates were 11.24 years (LI), 7.78 years (C), 9.89 years (PM) group, lowest being for canines. The estimated age was compared with the chronological age. The difference between the chronological age and the estimated age by the derived equation for the present study was statistically nonsignificant (P = 1.00). The results of the study show the feasibility of calculation of PTV ratios on CBCT to estimate the age for the set population with canine as the best predictor of age for the present study population.

A deep learning model for CT-based kidney volume determination in dogs and normal reference definition.

Kidney volume is associated with renal function and the severity of renal diseases, thus accurate assessment of the kidney is important. Although the voxel count method is reported to be more accurate than several methods, its laborious and time-consuming process is considered as a main limitation. In need of a new technology that is fast and as accurate as the manual voxel count method, the aim of this study was to develop the first deep learning model for automatic kidney detection and volume estimation from computed tomography (CT) images of dogs. A total of 182,974 image slices from 386 CT scans of 211 dogs were used to develop this deep learning model. Owing to the variance of kidney size and location in dogs compared to humans, several processing methods and an architecture based on UNEt Transformers which is known to show promising results for various medical image segmentation tasks including this study. Combined loss function and data augmentation were applied to elevate the performance of the model. The Dice similarity coefficient (DSC) which shows the similarity between manual segmentation and automated segmentation by deep-learning model was 0.915 ± 0.054 (mean ± SD) with post-processing. Kidney volume agreement analysis assessing the similarity between the kidney volume estimated by manual voxel count method and the deep-learning model was r = 0.960 (p < 0.001), 0.95 from Lin's concordance correlation coefficient (CCC), and 0.975 from the intraclass correlation coefficient (ICC). Kidney volume was positively correlated with body weight (BW), and insignificantly correlated with body conditions score (BCS), age, and sex. The correlations between BW, BCS, and kidney volume were as follows: kidney volume = 3.701 × BW + 11.962 (R 2 = 0.74, p < 0.001) and kidney volume = 19.823 × BW/BCS index + 10.705 (R 2 = 0.72, p < 0.001). The deep learning model developed in this study is useful for the automatic estimation of kidney volume. Furthermore, a reference range established in this study for CT-based normal kidney volume considering BW and BCS can be helpful in assessment of kidney in dogs.

White Matter Abnormalities and Cognitive Deficit After Mild Traumatic Brain Injury: Comparing DTI, DKI, and NODDI.

White matter (WM) disruption is an important determinant of cognitive impairment after mild traumatic brain injury (mTBI), but traditional diffusion tensor imaging (DTI) shows some limitations in assessing WM damage. Diffusion kurtosis imaging (DKI) and neurite orientation dispersion and density imaging (NODDI) show advantages over DTI in this respect. Therefore, we used these three diffusion models to investigate complex WM changes in the acute stage after mTBI. From 32 mTBI patients and 31 age-, sex-, and education-matched healthy controls, we calculated eight diffusion metrics based on DTI (fractional anisotropy, axial diffusivity, radial diffusivity, and mean diffusivity), DKI (mean kurtosis), and NODDI (orientation dispersion index, volume fraction of intracellular water (Vic), and volume fraction of the isotropic diffusion compartment). We used tract-based spatial statistics to identify group differences at the voxel level, and we then assessed the correlation between diffusion metrics and cognitive function. We also performed subgroup comparisons based on loss of consciousness. Patients showed WM abnormalities and cognitive deficit. And these two changes showed positive correlation. The correlation between Vic of the splenium of the corpus callosum and Digit Symbol Substitution Test scores showed the smallest p-value (p = 0.000, r = 0.481). We concluded that WM changes, especially in the splenium of the corpus callosum, correlate to cognitive deficit in this study. Furthermore, the high voxel count of NODDI results and the consistency of mean kurtosis and the volume fraction of intracellular water in previous studies and our study showed the functional complementarity of DKI and NODDI to DTI.

An integrated cluster-wise significance measure for fMRI analysis.

Cluster‐wise inference is widely used in fMRI analysis. The cluster‐level statistic is often obtained by counting the number of intra‐cluster voxels which surpass a voxel‐level statistical significance threshold. This measure can be sub‐optimal regarding the power and false‐positive error rate because the suprathreshold voxel count neglects the voxel‐wise significance levels and ignores the dependence between voxels. This article aims to provide a new Integrated Cluster‐wise significance Measure (ICM) for cluster‐level significance determination in cluster‐wise fMRI analysis by integrating cluster extent, voxel‐level significance (e.g., p values), and activation dependence between within‐cluster voxels. We develop a computationally efficient strategy for ICM based on probabilistic approximation theories. Consequently, the computational load for ICM‐based cluster‐wise inference (e.g., permutation tests) is affordable. We validate the proposed method via extensive simulations and then apply it to two fMRI data sets. The results demonstrate that ICM can improve the power with well‐controlled family‐wise error (FWE).