Segmental Variation Research Articles

Measuring Device for Belt Wedge Profile Parameters of Automobile V-Ribbed Belt Based on Included Angle of Slope Change Algorithm

In order to solve the problem of fast and accurate measurement of belt wedge profile parameters in the production of automobile V-ribbed belt, a measuring scheme for belt wedge profile parameters of automobile V-ribbed belt based on the included angle of slope value change algorithm is proposed, and the corresponding measuring device for belt wedge profile parameters of automobile V-ribbed belt is designed to collect belt wedge profile data points of automobile V-ribbed belt, and by analyzing the characteristics of belt wedge profile of PK type automobile V-ribbed belt, a kind of profile segmentation algorithm based on the included angle of slope change is proposed. By analyzing the characteristics of belt wedge profile of the PK type automobile V-ribbed belt, we propose an algorithm based on the included angle of slope variation for segmentation of belt wedge profile data points of automobile V-ribbed belt, and carry out the profile segmentation of collected belt wedge profile data points, and carry out the arc fitting of the segmented slot bottom data points to calculate the belt wedge profile parameters of measured belt. Based on the included angle of slope variation algorithm, the measurement of the belt wedge profile parameters of automobile V-ribbed belt is realized, and automatic measurement of the belt wedge profile parameters of automobile V-ribbed belt is achieved, which improves the measurement efficiency and ensures the measuring accuracy while avoiding the influence of human factors.

Accuracy of Prostate Segmentation on Synthetic MR Images Generated from CT Scans Using a Novel Generative Adversarial Network for High-Dose-Rate Prostate Brachytherapy

<h3>Purpose/Objective(s)</h3> For patients undergoing prostate HDR brachytherapy at our institution, both CT and MRI scans are acquired to identify catheters and delineate the prostate, respectively. However, acquiring both scans increases the time and hospital resources needed for treatment. We hypothesized that a synthetic MRI (sMRI) could be generated with sufficiently high soft-tissue contrast from the planning CT using a novel Generative Adversarial Network (GAN) framework with the goal to use sMRI to delineate the prostate without requiring the MRI. <h3>Materials/Methods</h3> Our newly developed PixCycleGAN, a hybrid of Pix2Pix and CycleGAN to maximize the advantage of Pix2Pix and to add the strength of CycleGAN training with paired CT-MRI datasets acquired on the same day for previously treated HDR prostate patients. PixCycleGAN is composed of two Residual Network generators and four discriminators with two transformation cycles. To train PixCycleGAN, 58 paired CT and T2-weighted MRI datasets without metal or motion artifacts were selected. Using 10 independent datasets, the image quality of sMRI was evaluated using metrics including the mean absolute error (MAE), mean squared error (MSE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM). The accuracy of prostate segmentation was evaluated by having radiation oncologists manually delineate the prostate on real MRI (rMRI) and sMRI images. All rMRI and sMRI images were presented in a random order to reduce potential bias in segmentation. The dice similarity coefficient (DSC) and Hausdorff distance (HD) were evaluated for the prostate contours delineated on sMRI vs. rMRI. To estimate interobserver variability (IOV), DSC and HD of prostate contours independently delineated on rMRI by two radiation oncologists were calculated. <h3>Results</h3> Qualitatively, sMRI images showed enhanced soft-tissue contrast on the boundary of the prostate and rectum compared to the CT scans. In addition, sMRI contained features from both MRI and CT showing clear visualization of the boundary of the prostate and the catheters. Image quality metrics were: MAE=0.13±0.03, MSE=0.03±0.01, PSNR=69.2±1.42 (dB), SSIM=0.78±0.07. The DSC of the prostate contour was 0.85±0.05 for sMRI vs. rMRI and 0.84±0.07 for IOV (p=0.13). However, the HD for sMRI vs. rMRI (7.77±1.83 mm) was smaller than the HD for IOV (9.53±2.74 mm). <h3>Conclusion</h3> We developed a novel PixCycleGAN framework to generate a sMRI from the planning CT which depicted enhanced soft-tissue contrast at the boundaries of the prostate. The agreement between DSC of the prostate on sMRI vs. rMRI and DSC of IOV shows that the accuracy of prostate segmentation on sMRI is within the segmentation variation on rMRI between two radiation oncologists. In the future, we will evaluate dosimetric impact of using sMRI in prostate segmentation for HDR treatment planning.

Structural consequences of sequence variation in mammalian prion β2α2 loop segments.

Sequence variation in the β2α2 loop, residues 165-175 of the mammalian prion protein (PrP), influences its structure. To better understand the consequences of sequence variation in this region of the protein, we biochemically and biophysically interrogate natural and artificial sequence variants of the β2α2 loop of mammalian PrP. Using microcrystal electron diffraction (MicroED), we determine atomic resolution structures of segments encompassing residues 168-176 from the β2α2 loop of PrP with sequences corresponding to human, mouse/cow, bank vole/hamster, rabbit/pig/guinea pig, and naked mole rat (elk-T174S) β2α2 loops, as well as synthetic β2α2 loop sequences. This collection of structures presents two dominant amyloid packing polymorphisms. In the first polymorph, denoted "clasped", side chains within a sheet form polar clasps by facing each other on the same strand, exemplified by the mouse/cow, human, and bank vole/hamster sequences. Because its stability is derived from within a strand and through polar ladders within a sheet, the sequence requirements for the mating strand are less restrictive. A second polymorph, denoted "interdigitated," has sidechains interdigitate across mating sheets, exemplified by the elk, naked mole rat (elk T174S), and rabbit sequences. The two types of packing present distinct networks of stabilizing hydrogen bonds. The identity of residue 174 appears to strongly influence the packing adopted in these peptides, but consideration of the overall sequence of a given segment is needed to understand the stability of its assemblies. Incorporation of these β2α2 loop sequences into an 85 residue recombinant segment encoding wild-type bank vole PrP94-178 demonstrates that even single residue substitutions could impact fibril morphology as evaluated by negative stain electron microscopy. This is in line with recent findings supporting the accessibility of different structural geometries by varied mammalian prion sequences, and indicates that sequence-specific polymorphisms may be influenced by residues in the β2α2 loop.

Effect of Normal Breathing on the Movement of CSF in the Spinal Subarachnoid Space.

Forced respirations reportedly have an effect on CSF movement in the spinal canal. We studied respiratory-related CSF motion during normal respiration. Six healthy subjects breathed at their normal rate with a visual guide to ensure an unchanging rhythm. Respiratory-gated phase-contrast MR flow images were acquired at 5 selected axial planes along the spine. At each spinal level, we computed the flow rate voxelwise in the spinal canal, together with the associated stroke volume. From these data, we computed the periodic volume changes of spinal segments. A phantom was used to quantify the effect of respiration-related magnetic susceptibility changes on the velocity data measured. At each level, CSF moved cephalad during inhalation and caudad during expiration. While the general pattern of fluid movement was the same in the 6 subjects, the flow rates, stroke volumes, and spine segment volume changes varied among subjects. Peak flow rates ranged from 0.60 to 1.59 mL/s in the cervical region, 0.46 to 3.17 mL/s in the thoracic region, and 0.75 to 3.64 mL/s in the lumbar region. The differences in flow rates along the canal yielded cyclic volume variations of spine segments that were largest in the lumbar spine, ranging from 0.76 to 3.07 mL among subjects. In the phantom study, flow velocities oscillated periodically during the respiratory cycle by up to 0.02 cm/s or 0.5%. Respiratory-gated measurements of the CSF motion in the spinal canal showed cyclic oscillatory movements of spinal fluid correlated to the breathing pattern.

Robust whole-tumour 3D volumetric CT-based radiomics approach for predicting the WHO/ISUP grade of a ccRCC tumour

ABSTRACT Clear Cell Renal Cell Carcinoma (ccRCC) is a common and major type of cancer. It is clinically vital to know the nuclear grade of a ccRCC tumour pre-surgically and non-invasively as this may impact therapeutic actions. We hypothesise that using three-dimensional (3D) radiomic features and ensemble learning models can increase the discrimination power between grades. The hybrid feature selection method was used to reduce the number of features. Besides the actual tumour volume, five additional volumes of interests (VOIs) were created to consider peritumour regions and test the robustness of the model against variations in segmentation. The best result was acquired when Synthetic Minority Oversampling Technique (SMOTE) was used in combination with Light Gradient Boosting Machine (LightGBM). The area under the curve (AUC) for this model was 0.89 ± 0.02. The results were 0.86 and 0.83 when the contour was made smaller and larger by 2 mm. We conclude that a ccRCC tumour grade can be predicted from 3D CT images with a high reliability, despite the inadequacy of a dataset. The algorithm is moderately robust against deviations in segmentation by observers. The features from the peritumour area did not result in any significant improvement.

T1 values and extracellular volume fraction in asymptomatic subjects: variations in left ventricular segments and correlation with cardiovascular risk factors

To evaluate variations in pre-contrast (preT1) and post-contrast (postT1) myocardial T1 values and extracellular volume fraction (ECV) according to left ventricular (LV) segments and to find correlations between them and cardiovascular risk factors. The 233 asymptomatic subjects (210 men, 23 women; aged 54.1 ± 6.0 years) underwent cardiac magnetic resonance imaging with preT1 and postT1 mapping on a 1.5-T scanner. T1 values and ECVs were evaluated according to LV segments, age, sex, and estimated glomerular filtration rate (eGFR). Based on the presence of hypertension (HTN) and diabetes mellitus (DM), subjects were subdivided into the control, HTN, DM, and HTN and DM (HTN-DM) groups. T1 values and ECV showed significant differences between septal and lateral segments at the mid-ventricular and basal levels (p ≤ 0.003). In subgroup analysis, the HTN-DM group showed a significantly higher ECV (0.260 ± 0.023) than the control (0.240 ± 0.021, p = 0.011) and HTN (0.241 ± 0.024, p = 0.041) groups. Overall postT1 and ECV of the LV had significant correlation with eGFR (r = 0.19, p = 0.038 for postT1; r = − 0.23, p = 0.011 for ECV). Septal segments show higher preT1 and ECV but lower postT1 than lateral segments at the mid-ventricular and basal levels. ECV is significantly affected by HTN, DM, and eGFR, even in asymptomatic subjects.

Characterization of replication and variations in genome segments of a bat reovirus, BatMRV/B19-02, by RNA-seq in infected Vero-E6 cells.

Mammalian orthoreoviruses (MEVs) that can cause enteric, respiratory, and encephalitic infections have been identified in a wide variety of mammalian species. Here, we report a novel MRV type 1 strain detected in Miniopterus schreibersii that may have resulted from reassortment events. Using next-generation RNA sequencing (RNA-seq), we found that the ratios of the RNA levels of the 10 reovirus segments in infected cells were constant during the late stages of infection. We also discovered that the relative abundance of each segment differed. Notably, the relative abundance of M2 (encoding the µ1 protein) and S4 (encoding the σ3 protein) RNAs was higher than that of the others throughout the infection. Additionally, massive junctions were identified. These results support the hypothesis that defective genome segments are generated and that cross-family recombination occurs. These data may further the study of gene function, viral replication, and virus evolution.

Overall Survival Prediction of Glioma Patients With Multiregional Radiomics.

Radiomics-guided prediction of overall survival (OS) in brain gliomas is seen as a significant problem in Neuro-oncology. The ultimate goal is to develop a robust MRI-based approach (i.e., a radiomics model) that can accurately classify a novel subject as a short-term survivor, a medium-term survivor, or a long-term survivor. The BraTS 2020 challenge provides radiological imaging and clinical data (178 subjects) to develop and validate radiomics-based methods for OS classification in brain gliomas. In this study, we empirically evaluated the efficacy of four multiregional radiomic models, for OS classification, and quantified the robustness of predictions to variations in automatic segmentation of brain tumor volume. More specifically, we evaluated four radiomic models, namely, the Whole Tumor (WT) radiomics model, the 3-subregions radiomics model, the 6-subregions radiomics model, and the 21-subregions radiomics model. The 3-subregions radiomics model is based on a physiological segmentation of whole tumor volume (WT) into three non-overlapping subregions. The 6-subregions and 21-subregions radiomic models are based on an anatomical segmentation of the brain tumor into 6 and 21 anatomical regions, respectively. Moreover, we employed six segmentation schemes – five CNNs and one STAPLE-fusion method – to quantify the robustness of radiomic models. Our experiments revealed that the 3-subregions radiomics model had the best predictive performance (mean AUC = 0.73) but poor robustness (RSD = 1.99) and the 6-subregions and 21-subregions radiomics models were more robust (RSD 1.39) with lower predictive performance (mean AUC 0.71). The poor robustness of the 3-subregions radiomics model was associated with highly variable and inferior segmentation of tumor core and active tumor subregions as quantified by the Hausdorff distance metric (4.4−6.5mm) across six segmentation schemes. Failure analysis revealed that the WT radiomics model, the 6-subregions radiomics model, and the 21-subregions radiomics model failed for the same subjects which is attributed to the common requirement of accurate segmentation of the WT volume. Moreover, short-term survivors were largely misclassified by the radiomic models and had large segmentation errors (average Hausdorff distance of 7.09mm). Lastly, we concluded that while STAPLE-fusion can reduce segmentation errors, it is not a solution to learning accurate and robust radiomic models.

Fish-inspired segment models for undulatory steady swimming

Many aquatic animals swim by undulatory body movements and understanding the diversity of these movements could unlock the potential for designing better underwater robots. Here, we analyzed the steady swimming kinematics of a diverse group of fish species to investigate whether their undulatory movements can be represented using a series of interconnected multi-segment models, and if so, to identify the key factors driving the segment configuration of the models. Our results show that the steady swimming kinematics of fishes can be described successfully using parsimonious models, 83% of which had fewer than five segments. In these models, the anterior segments were significantly longer than the posterior segments, and there was a direct link between segment configuration and swimming kinematics, body shape, and Reynolds number. The models representing eel-like fishes with elongated bodies and fishes swimming at high Reynolds numbers had more segments and less segment length variability along the body than the models representing other fishes. These fishes recruited their anterior bodies to a greater extent, initiating the undulatory wave more anteriorly. Two shape parameters, related to axial and overall body thickness, predicted segment configuration with moderate to high success rate. We found that head morphology was a good predictor of its segment length. While there was a large variation in head segments, the length of tail segments was similar across all models. Given that fishes exhibited variable caudal fin shapes, the consistency of tail segments could be a result of an evolutionary constraint tuned for high propulsive efficiency. The bio-inspired multi-segment models presented in this study highlight the key bending points along the body and can be used to decide on the placement of actuators in fish-inspired robots, to model hydrodynamic forces in theoretical and computational studies, or for predicting muscle activation patterns during swimming.

PO-1595 Automated stability study on mpMRI prostate radiomics features to variations in segmentation

PO-1595 Automated stability study on mpMRI prostate radiomics features to variations in segmentation

Myocardial electrophysiological and mechanical changes caused by moderate hypothermia-A clinical study.

Moderate hypothermia has been used to improve outcomes in comatose out‐of‐hospital cardiac arrest survivors during the past two decades, although the effects remain controversial. We have recently shown in an experimental study that myocardial electrophysiological and mechanical relationships were altered during moderate hypothermia. Electromechanical window positivity increased, and electrical dispersion of repolarization decreased, both of which are changes associated with decreased arrhythmogenicity in clinical conditions. Mechanical dispersion, a parameter also linked to arrhythmic risk, remained unaltered. Whether corresponding electrophysiological and mechanical changes occur in humans during moderate hypothermia, has not been previously explored. Twenty patients with normal left ventricular function were included. Measurements were obtained at 36 and 32°C prior to ascending aortic repair while on partial cardiopulmonary bypass and at 36°C after repair. Registrations were performed in the presence of both spontaneous and comparable paced heart rate during standardized loading conditions. The following electrical and mechanical parameters were explored: (1) Electromechanical window, measured as time difference between mechanical and electrical systole, (2) dispersion of repolarization from ECG T‐wave, and (3) mechanical dispersion, measured as segmental variation in time to peak echocardiographic strain. At moderate hypothermia, mechanical systolic prolongation (425 ± 43–588 ± 67 ms, p < 0.001) exceeded electrical systolic prolongation (397 ± 49–497 ± 79 ms, p < 0.001), whereby, electromechanical window positivity increased (29 ± 30–86 ± 50 ms, p < 0.001). Dispersion of repolarization and mechanical dispersion remained unchanged. Corresponding electrophysiological and mechanical relationships were present at comparable paced heart rates. After rewarming, the increased electromechanical window was reversed in the presence of both spontaneous and paced heart rates. Moderate hypothermia increased electromechanical window positivity, while dispersion of repolarization and mechanical dispersion remained unchanged. This impact of hypothermia may be clinically relevant for selected groups of patients after cardiac arrest.

Global U-net with amalgamation of inception model and improved kernel variation for MRI brain image segmentation

Global U-net with amalgamation of inception model and improved kernel variation for MRI brain image segmentation

Epistemic and aleatoric uncertainties reduction with rotation variation for medical image segmentation with ConvNets

The deep convolutional neural network (ConvNet) achieves significant segmentation performance on medical images of various modalities. However, the isolated errors in a large testing set with various tumor conditions are not acceptable in clinical practice. This is usually caused in inadequate training and noise inherent during data collection, which are recognized as epistemic and aleatoric uncertainties in deep learning-based approaches. In this paper, we analyze the two types of uncertainties in medical image segmentation tasks and propose a reduction method by training models with data augmentation. The shelter zones in images are reduced with 2D imaging on surfaces of different angles from 3D organs. Rotation transformation and noise are estimated by Monte Carlo simulation with prior parameter distributions, and the aleatoric uncertainty is quantized in this process. Experiments on segmentation of computed tomography images demonstrate that overconfident incorrect predictions are reduced through uncertainty reduction and that our method outperforms prediction baselines based on epistemic and aleatoric estimation.

Acute ST-segment elevation myocardial infarction: to be or not to be?

Calcium has a significant effect on cells in the myocardium, affecting conduction, intracellular signaling and contraction of muscle fibers. In fact, calcium levels could affect myocardial action potential and affect heart conduction. Hypocalcemia manifest on ECG with arrythmias or QT prolongation, in some extraordinary cases, it could present with variations in the ST segment, which requires the study of differential diagnoses of cardiac origin, such as acute myocardial infarction with ST elevation (STEMI) or pericarditis. We expose a case of a 28-year-old male patient who arrived at the emergency department with chest pain and was misdiagnosed with STEMI, after a complex process finally it was a severe hypocalcemia mimicking acute myocardial infarction.

Basal Segmental Strain as a Marker of Cardiac Involvement in Anderson-Fabry Disease

Anderson-Fabry disease (AFD), an X-linked genetic disease, can lead to cardiac involvement, manifesting notably as increased left ventricular (LV) wall thickness and myocardial fibrosis in the 4th-5th decade. We sought to evaluate and validate assessment of global and segmental variations in subclinical LV function by 2D speckle tracking in AFD patients versus healthy controls.

Slab metamorphism and interface earthquakes in Peru: Implications from three-dimensional hydrothermal variation in the subducted Nazca plate

Convergent plate dehydration is of significance in determining the coupling strength and seismicity activity at variable subduction contacts. Characterized by decadal M > 7.5 devastating earthquakes offshore and a lack of a magmatic arc in north-central Peru, the Peruvian megathrust remains enigmatic in terms of the along-strike heterogeneity of the thermal regime, slab dehydration, and seismotectonics. Hydrothermally controlled petrological metamorphism of multiple facies within the subducted oceanic lithosphere and subsequent fluid channeling and migration lead to fault instability and seismic and aseismic slips on the megathrusts. Based on 3D thermal modeling, we estimate the features of the Peruvian subduction hydrothermal regime. Our results show that (1) the shallow seismogenic zone (M > 7.5 earthquakes) along coastal Peru exhibits high slab dehydration rates and gradients and is associated with fault segmentation and the updip migration of fluids controlling the occurrence of great megathrust earthquakes; (2) the low slab dehydration rate (<0.01 wt%/km) in north Peru contributes to the absent volcanism in north Peru, while the high slab dehydration rate (>0.02 wt%/km) beneath south Peru facilitate the recurrence of magmatic volcanism; (3) the subduction-related slip behaviors at a certain depth resulting from heterogeneous variations in the coupled plate interface are determined by slab dehydration and petrological metamorphism in the subducted Nazca plate. The variation in slab dehydration contributes to the seismotectonic segmentation and slip variation associated with changes in the pore fluid pressure on the subducted megathrust.

Formation, segmentation and deep crustal structure variations along the Algerian margin from the SPIRAL seismic experiment

The Algerian continental margin in the western Mediterranean formed as a back-arc basin and is today reactivated by the convergence between the African and Eurasian plates. It is one of the very rare examples of passive margins undergoing inversion expressed in a moderate seismicity of the margin and is possibly giving way for future subduction. With the objective to better know the deep structure of the margin and its associated basin, the origin of its seismicity and to study the mechanism of reactivation, five existing wide-angle seismic profiles along the margin are revisited. They were located offshore Mostaganem, Tipasa, Greater Kabylia, Jijel and Annaba. These profiles show that the basin is underlain by a 5 km thick crust of oceanic magmatic origin, possibly created at non-continuous small accretionary ridge segments. The continent-ocean transition zone is narrow, except at the easternmost profile, possibly due to an opening including a shear movement. No high velocity zone in the lower crust corresponding to mantle rocks has been imaged at the Algerian margin. The continental crust is thinned in a narrow and strongly segmented manner. It is widest (70 km) in the central segment offshore Greater Kabylia where there is a wider zone of distal thinned continental crust than on the other margin segments. The thickest crust detected during this survey corresponds to the African continental crust and the Kabylides blocks and is about 22–25 km thick. This reduced thickness in comparison with unthinned continental crust might be due to the influence of earlier subduction at the margin, in form of erosion by the subducting slab.

Validation of 4D Flow based relative pressure maps in aortic flows.

While the clinical gold standard for pressure difference measurements is invasive catheterization, 4D Flow MRI is a promising tool for enabling a non-invasive quantification, by linking highly spatially resolved velocity measurements with pressure differences via the incompressible Navier-Stokes equations. In this work we provide a validation and comparison with phantom and clinical patient data of pressure difference maps estimators. We compare the classical Pressure Poisson Estimator (PPE) and the new Stokes Estimator (STE) against catheter pressure measurements under a variety of stenosis severities and flow intensities. Specifically, we use several 4D Flow data sets of realistic aortic phantoms with different anatomic and hemodynamic severities and two patients with aortic coarctation. The phantom data sets are enriched by subsampling to lower resolutions, modification of the segmentation and addition of synthetic noise, in order to study the sensitivity of the pressure difference estimators to these factors. Overall, the STE method yields more accurate results than the PPE method compared to catheterization data. The superiority of the STE becomes more evident at increasing Reynolds numbers with a better capacity of capturing pressure gradients in strongly convective flow regimes. The results indicate an improved robustness of the STE method with respect to variation in lumen segmentation. However, with heuristic removal of the wall-voxels, the PPE can reach a comparable accuracy for lower Reynolds' numbers.

Characteristics analysis of segmental and regional lumbar spontaneous compensation post thoracic fusion in Lenke 1 and 2 adolescent idiopathic scoliosis

ObjectiveTo explore the characteristics of compensation of unfused lumbar region post thoracic fusion in Lenke 1 and 2 adolescent idiopathic scoliosis.BackgroundPreserving lumbar mobility in the compensation is significant in controlling pain and maintaining its functions. The spontaneous correction of the distal unfused lumbar curve after STF has been widely reported, but previous study has not concentrated on the characteristics of compensation of unfused lumbar region post thoracic fusion.MethodA total of 51 Lenke 1 and2 AIS patients were included, whose lowest instrumented vertebrae was L1 from January 2013 to December 2019. For further analysis, demographic data and coronal radiographic films were collected before surgery, at immediate erect postoperatively and final follow-up. The wedge angles of each unfused distal lumbar segments were measured, and the variations in each disc segment were calculated at the immediate postoperative review and final follow-up. Meanwhile, the unfused lumbar curve was divided into upper and lower parts, and we calculated their curve angles and compensations.ResultsThe current study enrolled 41 females (80.4%) and 10 males (19.6%). Thirty-six patients were Lenke type 1, while 15 patients were Lenke type 2. The average main thoracic Cobb angle and thoracolumbar/lumbar Cobb angle were 44.1 ± 7.7°and 24.1 ± 9.3°, preoperatively. At the final follow-up, the disc wedge angle variation of L1/2, L2/3, L3/4, L4/5 and L5/S1 was 3.84 ± 5.96°, 3.09 ± 4.54°, 2.30 ± 4.53°, − 0.12 ± 3.89° and − 1.36 ± 2.80°, respectively. The compensation of upper and lower coronal lumbar curves at final follow-up were 9.22 ± 10.39° and − 1.49 ± 5.14°, respectively.ConclusionWhen choosing L1 as the lowest instrumented vertebrae, the distal unfused lumbar segments’ compensation showed a decreasing trend from the proximal end to the distal end. The adjacent L1/2 and L2/3 discs significantly contributed to this compensation.

Relationship between Circle of Willis Variations and Cerebral or Cervical Arteries Stenosis Investigated by Computer Tomography Angiography and Multitask Convolutional Neural Network.

Circle of Willis (CoW) is the most critical collateral pathway that supports the redistribution of blood supply in the brain. The variation of CoW is closely correlated with cerebral hemodynamic and cerebral vessel-related diseases. But what is responsible for CoW variation remains unclear. Moreover, the visual evaluation for CoW variation is highly time-consuming. In the present study, based on the computer tomography angiography (CTA) dataset from 255 patients, the correlation between the CoW variations with age, gender, and cerebral or cervical artery stenosis was investigated. A multitask convolutional neural network (CNN) was used to segment cerebral arteries automatically. The results showed the prevalence of variation of the anterior communicating artery (Aco) was higher in the normal senior group than in the normal young group and in females than in males. The changes in the prevalence of variations of individual segments were not demonstrated in the population with stenosis of the afferent and efferent arteries, so the critical factors for variation are related to genetic or physiological factors rather than pathological lesions. Using the multitask CNN model, complete cerebral and cervical arteries could be segmented and reconstructed in 120 seconds, and an average Dice coefficient of 78.2% was achieved. The segmentation accuracy for precommunicating part of anterior cerebral artery and posterior cerebral artery, the posterior communicating arteries, and Aco in CoW was 100%, 99.2%, 94%, and 69%, respectively. Artificial intelligence (AI) can be considered as an adjunct tool for detecting the CoW, particularly related to reducing workload and improving the accuracy of the visual evaluation. The study will serve as a basis for the following research to determine an individual's risk of stroke with the aid of AI.