3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset

The effect of fat model variation on muscle fat fraction quantification in a cross-sectional cohort

Spectroscopic imaging, rooted in Dixon's two-echo spin sequence to distinguish water and fat, has evolved significantly in acquisition and processing. Yet precise fat quantification remains a persistent challenge in ongoing research. With adequate phase characterization and correction, the fat composition models will impact measurements of fatty tissue. However, the effect of the used fat model in low-fat regions such as healthy muscle is unknown. In this study, we investigate the effect of assumed fat composition, in terms of chain length and double bond count, on fat fraction quantification in healthy muscle, while addressing phase and relaxometry confounders. For this purpose, we acquired bilateral thigh datasets from 38 healthy volunteers. Fat fractions were estimated using the IDEAL algorithm employing three different fat models fitted with and without the initial phase constrained. After data processing and model fitting, we used a convolutional neural net to automatically segment all thigh muscles and subcutaneous fat to evaluate the fitted parameters. The fat composition was compared with those reported in the literature. Overall, all the observed estimated fat composition values fall within the range of previously reported fatty acid composition based on gas chromatography measurements. All methods and models revealed different estimates of the muscle fat fractions in various evaluated muscle groups. Lateral differences changed from 0.5% to 5.3% in the hamstring muscle groups depending on the chosen method. The lowest observed left-right differences in each muscle group were all for the fat model estimating the number of double bonds with the initial phase unconstrained. With this model, the left-right differences were 0.64% ± 0.31%, 0.50% ± 0.27%, and 0.50% ± 0.40% for the quadriceps, hamstrings, and adductors muscle groups, respectively. Our findings suggest that a fat model estimating double bond numbers while allowing separate phases for each chemical species, given some assumptions, yields the best fat fraction estimate for our dataset.

Prospective longitudinal cohort study of quantitative muscle magnetic resonance imaging in a healthy control population

Quantitative muscle magnetic resonance imaging (qMRI) is a valuable methodology for assessing muscular injuries and neuromuscular disorders. Notably, muscle diffusion tensor imaging (DTI) gives insights into muscle microstructural and macrostructural characteristics. However, the long-term reproducibility and robustness of these measurements remain relatively unexplored. The purpose of this prospective longitudinal cohort study was to assess the long-term robustness and range of variation of qMRI parameters, especially DTI metrics, in the lower extremity muscles of healthy controls under real-life conditions. Twelve volunteers (seven females, age 44.1 ± 12.1 years, body mass index 23.3 ± 2.0 kg/m2) underwent five leg muscle MRI sessions every 20 ± 4 weeks over a total period of 1.5 years. A multiecho gradient-echo Dixon-based sequence, a multiecho spin-echo T2-mapping sequence, and a spin-echo echo planar imaging diffusion-weighted sequence were acquired bilaterally with a Philips 3-T Achieva MR System using a 16-channel torso coil. Fifteen leg muscles were segmented in both lower extremities. qMRI parameters, including fat fraction (FF), water T2 relaxation time, and the diffusion metrics fractional anisotropy (FA) and mean diffusivity (MD), were evaluated. Coefficients of variance (wsCV) and intraclass correlation coefficients (ICCs) were calculated to assess the reproducibility of qMRI parameters. The standard error of measurement (SEM) and the minimal detectable change (MDC) were calculated to determine the range of variation. All tests were applied to all muscles and, subsequently, to each muscle separately. wsCV showed good reproducibility (≤ 10%) for all qMRI parameters in all muscles. The ICCs revealed excellent agreement between time points (FF = 0.980, water T2 = 0.941, FA = 0.952, MD = 0.948). Random measurement errors assessed by SEM and the MDC were low (< 12%). In conclusion, in this study, we showed that qMRI parameters in healthy volunteers living normal lives are stable over 18 months, thereby defining a benchmark for the expected range of variation over time.

A novel mean shape based post-processing method for enhancing deep learning lower-limb muscle segmentation accuracy

This study aims at improving the lower-limb muscle segmentation accuracy of deep learning approaches based on Magnetic Resonance Imaging (MRI) scans, crucial for the diagnostic and therapeutic processes in musculoskeletal diseases. In general, segmentation methods such as U-Net deep learning neural networks can achieve good Dice Similarity Coefficient (DSC) values, e.g. around 0.83 to 0.91 on various cohorts. Some generic post-processing strategies have been studied to incorporate connectivity constraints into the resulting masks for the purpose of further improving the segmentation accuracy. In this paper, a novel mean shape (MS) based post-processing method is proposed, utilizing Statistical Shape Modelling (SSM) to fine-tune the segmentation output, taking into consideration the muscle anatomical shape. The methodology was compared to existing post-processing techniques and a commercial semi-automatic tool on MRI scans from two cohorts of post-menopausal women (10 Training, 8 Testing, voxel size 1.0x1.0x1.0 mm3). The MS based method obtained a mean DSC of 0.83 across the different analysed muscles and the best performance for the Hausdorff Distance (HD, 20.6 mm) and the Average Symmetric Surface Distance (ASSD, 2.1 mm). These findings highlight the feasibility and potential of using anatomical mean shape in post-processing of human lower-limb muscle segmentation task and indicate that the proposed method can be popularized to other biological organ segmentation mission.

Positional contrastive learning for improved thigh muscle segmentation in MR images

The accurate segmentation of individual muscles is essential for quantitative MRI analysis of thigh images. Deep learning methods have achieved state-of-the-art results in segmentation, but they require large numbers of labeled data to perform well. However, labeling individual thigh muscles slice by slice for numerous volumes is a laborious and time-consuming task, which limits the availability of annotated datasets. To address this challenge, self-supervised learning (SSL) emerges as a promising technique to enhance model performance by pretraining the model on unlabeled data. A recent approach, called positional contrastive learning, exploits the information given by the axial position of the slices to learn features transferable on the segmentation task. The aim of this work was to propose positional contrastive SSL for the segmentation of individual thigh muscles from MRI acquisitions in a population of elderly healthy subjects and to evaluate it on different levels of limited annotated data. An unlabeled dataset of 72 T1w MRI thigh acquisitions was available for SSL pretraining, while a labeled dataset of 52 volumes was employed for the final segmentation task, split into training and test sets. The effectiveness of SSL pretraining to fine-tune a U-Net architecture for thigh muscle segmentation was compared with that of a randomly initialized model (RND), considering an increasing number of annotated volumes (S = 1, 2, 5, 10, 20, 30, 40). Our results demonstrated that SSL yields substantial improvements in Dice similarity coefficient (DSC) when using a very limited number of labeled volumes (e.g., for S  = 1, DSC 0.631 versus 0.530 for SSL and RND, respectively). Moreover, enhancements are achievable even when utilizing the full number of labeled subjects, with DSC = 0.927 for SSL and 0.924 for RND. In conclusion, positional contrastive SSL was effective in obtaining more accurate thigh muscle segmentation, even with a very low number of labeled data, with a potential impact of speeding up the annotation process in clinics.

Quantitative muscle magnetic resonance imaging in limb-girdle muscular dystrophy type R1 (LGMDR1): A prospective longitudinal cohort study

Limb-girdle muscular dystrophy (LGMD) type R1 (LGMDR1) is the most common subtype of LGMD in Europe. Prospective longitudinal data, including clinical assessments and new biomarkers such as quantitative magnetic resonance imaging (qMRI), are needed to evaluate the natural course of the disease and therapeutic options. We evaluated eight thigh and seven leg muscles of 13 LGMDR1 patients (seven females, mean age 36.7 years, body mass index 23.9 kg/m2) and 13 healthy age- and gender-matched controls in a prospective longitudinal design over 1 year. Clinical assessment included testing for muscle strength with quick motor function measure (QMFM), gait analysis and patient questionnaires (neuromuscular symptom score, activity limitation [ACTIVLIM]). MRI scans were performed on a 3-T MRI scanner, including a Dixon-based sequence, T2 mapping and diffusion tensor imaging. The qMRI values of fat fraction (FF), water T2 relaxation time (T2), fractional anisotropy, mean diffusivity, axial diffusivity and radial diffusivity were analysed. Within the clinical outcome measures, significant deterioration between baseline and follow-up was found for ACTIVLIM (p = 0.029), QMFM (p = 0.012). Analysis of qMRI parameters of the patient group revealed differences between time points for both FF and T2 when analysing all muscles (FF: p < 0.001; T2: p = 0.016). The highest increase of fat replacement was found in muscles with an FF of between 10% and 50% at baseline. T2 in muscles with low-fat replacement increased significantly. No significant differences were found for the diffusion metrics. Significant correlations between qMRI metrics and clinical assessments were found at baseline and follow-up, while only T2 changes in thigh muscles correlated with changes in ACTIVLIM over time (ρ = -0.621, p < 0.05). Clinical assessments can show deterioration of the general condition of LGMDR1 patients. qMRI measures can give additional information about underlying pathophysiology. Further research is needed to establish qMRI outcome measures for clinical trials.

Evaluation of Neuromuscular Diseases and Complaints by Quantitative Muscle MRI

Background: Quantitative muscle MRI (qMRI) is a promising tool for evaluating and monitoring neuromuscular disorders (NMD). However, the application of different imaging protocols and processing pipelines restricts comparison between patient cohorts and disorders. In this qMRI study, we aim to compare dystrophic (limb-girdle muscular dystrophy), inflammatory (inclusion body myositis), and metabolic myopathy (Pompe disease) as well as patients with post-COVID-19 conditions suffering from myalgia to healthy controls. Methods: Ten subjects of each group underwent a 3T lower extremity muscle MRI, including a multi-echo, gradient-echo, Dixon-based sequence, a multi-echo, spin-echo (MESE) T2 mapping sequence, and a spin-echo EPI diffusion-weighted sequence. Furthermore, the following clinical assessments were performed: Quick Motor Function Measure, patient questionnaires for daily life activities, and 6-min walking distance. Results: Different involvement patterns of conspicuous qMRI parameters for different NMDs were observed. qMRI metrics correlated significantly with clinical assessments. Conclusions: qMRI metrics are suitable for evaluating patients with NMD since they show differences in muscular involvement in different NMDs and correlate with clinical assessments. Still, standardisation of acquisition and processing is needed for broad clinical use.

The Impact of Fatty Infiltration on MRI Segmentation of Lower Limb Muscles in Neuromuscular Diseases: A Comparative Study of Deep Learning Approaches

BACKGROUND

Deep learning methods have been shown to be useful for segmentation of lower limb muscle MRIs of healthy subjects but, have not been sufficiently evaluated on neuromuscular disease (NDM) patients.

PURPOSE

Evaluate the influence of fat infiltration on convolutional neural network (CNN) segmentation of MRIs from NMD patients.

STUDY TYPE

Retrospective study.

SUBJECTS

Data were collected from a hospital database of 67 patients with NMDs and 14 controls (age: 53 ± 17 years, sex: 48 M, 33 F). Ten individual muscles were segmented from the thigh and six from the calf (20 slices, 200 cm section).

FIELD STRENGTH/SEQUENCE

A 1.5 T. Sequences: 2D T1 -weighted fast spin echo. Fat fraction (FF): three-point Dixon 3D GRE, magnetization transfer ratio (MTR): 3D MT-prepared GRE, T2: 2D multispin-echo sequence.

ASSESSMENT

U-Net 2D, U-Net 3D, TransUNet, and HRNet were trained to segment thigh and leg muscles (101/11 and 95/11 training/validation images, 10-fold cross-validation). Automatic and manual segmentations were compared based on geometric criteria (Dice coefficient [DSC], outlier rate, absence rate) and reliability of measured MRI quantities (FF, MTR, T2, volume).

STATISTICAL TESTS

Bland-Altman plots were chosen to describe agreement between manual vs. automatic estimated FF, MTR, T2 and volume. Comparisons were made between muscle populations with an FF greater than 20% (G20+) and lower than 20% (G20-).

RESULTS

The CNNs achieved equivalent results, yet only HRNet recognized every muscle in the database, with a DSC of 0.91 ± 0.08, and measurement biases reaching -0.32% ± 0.92% for FF, 0.19 ± 0.77 for MTR, -0.55 ± 1.95 msec for T2, and - 0.38 ± 3.67 cm3 for volume. The performances of HRNet, between G20- and G20+ decreased significantly.

DATA CONCLUSION

HRNet was the most appropriate network, as it did not omit any muscle. The accuracy obtained shows that CNNs could provide fully automated methods for studying NMDs. However, the accuracy of the methods may be degraded on the most infiltrated muscles (>20%).

EVIDENCE LEVEL

4.

TECHNICAL EFFICACY

Stage 1.

Automated MRI quantification of volumetric per-muscle fat fractions in the proximal leg of patients with muscular dystrophies

Muscular dystrophies (MD) are a class of rare genetic diseases resulting in progressive muscle weakness affecting specific muscle groups, depending on the type of disease. Disease progression is characterized by the gradual replacement of muscle tissue by fat, which can be assessed with fat-sensitive magnetic resonance imaging (MRI) and objectively evaluated by quantifying the fat fraction percentage (FF%) per muscle. Volumetric quantification of fat replacement over the full 3D extent of each muscle is more precise and potentially more sensitive than 2D quantification in few selected slices only, but it requires an accurate 3D segmentation of each muscle individually, which is time consuming when this has to be performed manually for a large number of muscles. A reliable, largely automated approach for 3D muscle segmentation is thus needed to facilitate the adoption of fat fraction quantification as a measure of MD disease progression in clinical routine practice, but this is challenging due to the variable appearance of the images and the ambiguity in the discrimination of the contours of adjacent muscles, especially when the normal image contrast is affected and diminished by the fat replacement. To deal with these challenges, we used deep learning to train AI-models to segment the muscles in the proximal leg from knee to hip in Dixon MRI images of healthy subjects as well as patients with MD. We demonstrate state-of-the-art segmentation results of all 18 muscles individually in terms of overlap (Dice score, DSC) with the manual ground truth delineation for images of cases with low fat infiltration (mean overall FF%: 11.3%; mean DSC: 95.3% per image, 84.4-97.3% per muscle) as well as with medium and high fat infiltration (mean overall FF%: 44.3%; mean DSC: 89.0% per image, 70.8-94.5% per muscle). In addition, we demonstrate that the segmentation performance is largely invariant to the field of view of the MRI scan, is generalizable to patients with different types of MD and that the manual delineation effort to create the training set can be drastically reduced without significant loss of segmentation quality by delineating only a subset of the slices.

Quantitative muscle magnetic resonance imaging depicts microstructural abnormalities but no signs of inflammation or dystrophy in post-COVID-19 condition

BACKGROUND AND PURPOSE

Post-COVID-19 condition (PCC) has high impact on quality of life, with myalgia and fatigue affecting at least 25% of PCC patients. This case-control study aims to noninvasively assess muscular alterations via quantitative muscle magnetic resonance imaging (MRI) as possible mechanisms for ongoing musculoskeletal complaints and premature exhaustion in PCC.

METHODS

Quantitative muscle MRI was performed on a 3 Tesla MRI scanner of the whole legs in PCC patients compared to age- and sex-matched healthy controls, including a Dixon sequence to determine muscle fat fraction (FF), a multi-echo spin-echo sequence for quantitative water mapping reflecting putative edema, and a diffusion-weighted spin-echo echo-planar imaging sequence to assess microstructural alterations. Clinical examination, nerve conduction studies, and serum creatine kinase were performed in all patients. Quantitative muscle MRI results were correlated to the results of the 6-min walk test and standardized questionnaires assessing quality of life, fatigue, and depression.

RESULTS

Twenty PCC patients (female: n = 15, age = 48.8 ± 10.1 years, symptoms duration = 13.4 ± 4.2 months, body mass index [BMI] = 28.8 ± 4.7 kg/m² ) were compared to 20 healthy controls (female: n = 15, age = 48.1 ± 11.1 years, BMI = 22.9 ± 2.2 kg/m² ). Neither FF nor T2 revealed signs of muscle degeneration or inflammation in either study groups. Diffusion tensor imaging (DTI) revealed reduced mean, axial, and radial diffusivity in the PCC group.

CONCLUSIONS

Quantitative muscle MRI did not depict any signs of ongoing inflammation or dystrophic process in the skeletal muscles in PCC patients. However, differences observed in muscle DTI depict microstructural abnormalities, which may reflect potentially reversible fiber hypotrophy due to deconditioning. Further longitudinal and interventional studies should prove this hypothesis.

Muscle quantitative MRI as a novel biomarker in hereditary transthyretin amyloidosis with polyneuropathy: a cross-sectional study

BACKGROUND

The development of reproducible and sensitive outcome measures has been challenging in hereditary transthyretin (ATTRv) amyloidosis. Recently, quantification of intramuscular fat by magnetic resonance imaging (MRI) has proven as a sensitive marker in patients with other genetic neuropathies. The aim of this study was to investigate the role of muscle quantitative MRI (qMRI) as an outcome measure in ATTRv.

METHODS

Calf- and thigh-centered multi-echo T2-weighted spin-echo and gradient-echo sequences were obtained in patients with ATTRv amyloidosis with polyneuropathy (n = 24) and healthy controls (n = 12). Water T2 (wT2) and fat fraction (FF) were calculated. Neurological assessment was performed in all ATTRv subjects. Quantitative MRI parameters were correlated with clinical and neurophysiological measures of disease severity.

RESULTS

Quantitative imaging revealed significantly higher FF in lower limb muscles in patients with ATTRv amyloidosis compared to controls. In addition, wT2 was significantly higher in ATTRv patients. There was prominent involvement of the posterior compartment of the thighs. Noticeably, FF and wT2 did not exhibit a length-dependent pattern in ATTRv patients. MRI biomarkers correlated with previously validated clinical outcome measures, Polyneuropathy Disability scoring system, Neuropathy Impairment Score (NIS) and NIS-lower limb, and neurophysiological parameters of axonal damage regardless of age, sex, treatment and TTR mutation.

CONCLUSIONS

Muscle qMRI revealed significant difference between ATTRv and healthy controls. MRI biomarkers showed high correlation with clinical and neurophysiological measures of disease severity making qMRI as a promising tool to be further investigated in longitudinal studies to assess its role at monitoring onset, progression, and therapy efficacy for future clinical trials on this treatable condition.

DeepACSA: Automatic Segmentation of Cross-Sectional Area in Ultrasound Images of Lower Limb Muscles Using Deep Learning

PURPOSE

Muscle anatomical cross-sectional area (ACSA) can be assessed using ultrasound and images are usually evaluated manually. Here, we present DeepACSA, a deep learning approach to automatically segment ACSA in panoramic ultrasound images of the human rectus femoris (RF), vastus lateralis (VL), gastrocnemius medialis (GM) and lateralis (GL) muscles.

METHODS

We trained three muscle-specific convolutional neural networks (CNN) using 1772 ultrasound images from 153 participants (age = 38.2 yr, range = 13-78). Images were acquired in 10% increments from 30% to 70% of femur length for RF and VL and at 30% and 50% of muscle length for GM and GL. During training, CNN performance was evaluated using intersection-over-union scores. We compared the performance of DeepACSA to manual analysis and a semiautomated algorithm using an unseen test set.

RESULTS

Comparing DeepACSA analysis of the RF to manual analysis with erroneous predictions removed (3.3%) resulted in intraclass correlation (ICC) of 0.989 (95% confidence interval = 0.983-0.992), mean difference of 0.20 cm 2 (0.10-0.30), and SEM of 0.33 cm 2 (0.26-0.41). For the VL, ICC was 0.97 (0.96-0.968), mean difference was 0.85 cm 2 (-0.4 to 1.31), and SEM was 0.92 cm 2 (0.73-1.09) after removal of erroneous predictions (7.7%). After removal of erroneous predictions (12.3%), GM/GL muscles demonstrated an ICC of 0.98 (0.96-0.99), a mean difference of 0.43 cm 2 (0.21-0.65), and an SEM of 0.41 cm 2 (0.29-0.51). Analysis duration was 4.0 ± 0.43 s (mean ± SD) for analysis of one image in our test set using DeepACSA.

CONCLUSIONS

DeepACSA provides fast and objective segmentation of lower limb panoramic ultrasound images comparable with manual segmentation. Inaccurate model predictions occurred predominantly on low-quality images, highlighting the importance of high-quality image for accurate prediction.

Quantitative muscle MRI captures early muscle degeneration in calpainopathy

To evaluate differences in qMRI parameters of muscle diffusion tensor imaging (mDTI), fat-fraction (FF) and water T2 time in leg muscles of calpainopathy patients (LGMD R1/D4) compared to healthy controls, to correlate those findings to clinical parameters and to evaluate if qMRI parameters show muscle degeneration in not-yet fatty infiltrated muscles. We evaluated eight thigh and seven calf muscles of 19 calpainopathy patients and 19 healthy matched controls. MRI scans were performed on a 3T MRI including a mDTI, T2 mapping and mDixonquant sequence. Clinical assessment was done with manual muscle testing, patient questionnaires (ACTIVLIM, NSS) as well as gait analysis. Average FF was significantly different in all muscles compared to controls (p < 0.001). In muscles with less than 8% FF a significant increase of FA (p < 0.005) and decrease of RD (p < 0.004) was found in high-risk muscles of calpainopathy patients. Water T2 times were increased within the low- and intermediate-risk muscles (p ≤ 0.045) but not in high-risk muscles (p = 0.062). Clinical assessments correlated significantly with qMRI values: QMFM vs. FF: r = - 0.881, p < 0.001; QMFM versus FA: r = - 0.747, p < 0.001; QMFM versus MD: r = 0.942, p < 0.001. A good correlation of FF and diffusion metrics to clinical assessments was found. Diffusion metrics and T2 values are promising candidates to serve as sensitive early and non-invasive methods to capture early muscle degeneration in non-fat-infiltrated muscles in calpainopathies.

Label efficient segmentation of single slice thigh CT with two-stage pseudo labels

Purpose: Muscle, bone, and fat segmentation from thigh images is essential for quantifying body composition. Voxelwise image segmentation enables quantification of tissue properties including area, intensity, and texture. Deep learning approaches have had substantial success in medical image segmentation, but they typically require a significant amount of data. Due to the high cost of manual annotation, training deep learning models with limited human label data is desirable, but it is a challenging problem. Approach: Inspired by transfer learning, we proposed a two-stage deep learning pipeline to address the thigh and lower leg segmentation issue. We studied three datasets, 3022 thigh slices and 8939 lower leg slices from the BLSA dataset and 121 thigh slices from the GESTALT study. First, we generated pseudo labels for thigh based on approximate handcrafted approaches using CT intensity and anatomical morphology. Then, those pseudo labels were fed into deep neural networks to train models from scratch. Finally, the first stage model was loaded as the initialization and fine-tuned with a more limited set of expert human labels of the thigh. Results: We evaluated the performance of this framework on 73 thigh CT images and obtained an average Dice similarity coefficient (DSC) of 0.927 across muscle, internal bone, cortical bone, subcutaneous fat, and intermuscular fat. To test the generalizability of the proposed framework, we applied the model on lower leg images and obtained an average DSC of 0.823. Conclusions: Approximated handcrafted pseudo labels can build a good initialization for deep neural networks, which can help to reduce the need for, and make full use of, human expert labeled data.

Robustness and stability of volume-based tractography in a multicenter setting

Muscle diffusion tensor imaging (mDTI)-based tractography is a promising tool with which to detect subclinical changes in muscle injuries and to evaluate pathophysiology in neuromuscular diseases. Classic region of interest (ROI)-based tractography is very time-consuming and requires an examiner with extensive experience. (Semi)automatic approaches such as volume-based tractography (VBT) can diminish this problem but its robustness and stability are unknown. The aim of the current study was to assess the performance of VBT in a multicenter setting and to evaluate semiautomatic segmentation approaches in the analysis of VBT-derived data in terms of the comparability of the outcome measures. Five traveling volunteers underwent 3-T mDTI of seven calf muscles of both legs at six different MR sites. Tract properties and diffusion metrics were calculated using VBT. Within-subject coefficients of variance (wsCVs) and intraclass correlation coefficients (ICCs) were calculated to assess the multicenter reproducibility of tract properties such as tract density (TD), mean tract length, volume and tract propagation angle, and diffusion metrics such as fractional anisotropy, mean diffusivity, axial diffusivity (λ₁ ) and radial diffusivity in traveling subjects. Furthermore, 50 individual datasets from five different centers (10 datasets per center) were pooled to assess the feasibility of VBT with manual and semiautomatic segmentation. To assess the differences of tract properties and diffusion metrics between segmentation approaches an ANOVA was performed, and ICC and Bland-Altman plots were analyzed. wsCVs and ICCs showed good reproducibility of the tract properties TD and volume, as well as diffusion metrics. ANOVA showed no significant differences between manual and semiautomatic approaches. ICCs were excellent (≥ 0.992) and Bland-Altman analysis did not reveal any systemic bias between the methods. Tract properties and diffusion metrics derived from VBT showed good comparability among centers. Semiautomatic approaches revealed excellent agreement with gold standard of manual segmentation. These findings suggest that pooling data from different centers to construct a reference database for tractography results is feasible using semiautomatic segmentation approaches.