Dataset Information

Fully automated, real-time 3D ultrasound segmentation to estimate first trimester placental volume using deep learning.

ABSTRACT: We present a new technique to fully automate the segmentation of an organ from 3D ultrasound (3D-US) volumes, using the placenta as the target organ. Image analysis tools to estimate organ volume do exist but are too time consuming and operator dependant. Fully automating the segmentation process would potentially allow the use of placental volume to screen for increased risk of pregnancy complications. The placenta was segmented from 2,393 first trimester 3D-US volumes using a semiautomated technique. This was quality controlled by three operators to produce the "ground-truth" data set. A fully convolutional neural network (OxNNet) was trained using this ground-truth data set to automatically segment the placenta. OxNNet delivered state-of-the-art automatic segmentation. The effect of training set size on the performance of OxNNet demonstrated the need for large data sets. The clinical utility of placental volume was tested by looking at predictions of small-for-gestational-age babies at term. The receiver-operating characteristics curves demonstrated almost identical results between OxNNet and the ground-truth). Our results demonstrated good similarity to the ground-truth and almost identical clinical results for the prediction of SGA.

SUBMITTER: Looney P

PROVIDER: S-EPMC6124436 | biostudies-literature | 2018 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Fully automated, real-time 3D ultrasound segmentation to estimate first trimester placental volume using deep learning.

Looney Pádraig P Stevenson Gordon N GN Nicolaides Kypros H KH Plasencia Walter W Molloholli Malid M Natsis Stavros S Collins Sally L SL

JCI insight 20180607 11

We present a new technique to fully automate the segmentation of an organ from 3D ultrasound (3D-US) volumes, using the placenta as the target organ. Image analysis tools to estimate organ volume do exist but are too time consuming and operator dependant. Fully automating the segmentation process would potentially allow the use of placental volume to screen for increased risk of pregnancy complications. The placenta was segmented from 2,393 first trimester 3D-US volumes using a semiautomated tec ...[more]

PMID: 29875312

Similar Datasets

Project description:IntroductionBladder magnetic resonance imaging (MRI) has been recently integrated in the diagnosis pathway of bladder cancer. However, automatic recognition of suspicious lesions is still challenging. Thus, development of a solution for proper delimitation of the tumor and its separation from the healthy tissue is of primordial importance. As a solution to this unmet medical need, we aimed to develop an artificial intelligence-based decision support system, which automatically segments the bladder wall and the tumor as well as any suspect area from the 3D MRI images.MaterialsWe retrospectively assessed all patients diagnosed with bladder cancer, who underwent MRI at our department (n=33). All examinations were performed using a 1.5 Tesla MRI scanner. All images were reviewed by two radiologists, who performed manual segmentation of the bladder wall and all lesions. First, the performance of our fully automated end-to-end segmentation model based on a 3D U-Net architecture (by considering various depths of 4, 5 or 6 blocks) trained in two data augmentation scenarios (on 5 and 10 augmentation datasets per original data, respectively) was tested. Second, two learning setups were analyzed by training the segmentation algorithm with 7 and 14 MRI original volumes, respectively.ResultsWe obtained a Dice-based performance over 0.878 for automatic segmentation of bladder wall and tumors, as compared to manual segmentation. A larger training dataset using 10 augmentations for 7 patients could further improve the results of the U-Net-5 model (0.902 Dice coefficient at image level). This model performed best in terms of automated segmentation of bladder, as compared to U-Net-4 and U-Net-6. However, in this case increased time for learning was needed as compared to U-Net-4. We observed that an extended dataset for training led to significantly improved segmentation of the bladder wall, but not of the tumor.ConclusionWe developed an intelligent system for bladder tumors automated diagnostic, that uses a deep learning model to segment both the bladder wall and the tumor. As a conclusion, low complexity networks, with less than five-layers U-Net architecture are feasible and show good performance for automatic 3D MRI image segmentation in patients with bladder tumors.

Project description:PurposeThis study aimed to develop a fully automated deep learning ciliary body segmentation and assessment approach in three-dimensional ultrasound biomicroscopy (3D-UBM) images.MethodsEach 3D-UBM eye volume was aligned to the optic axis via multiplanar reformatting. Ciliary muscle and processes were manually annotated, and Deeplab-v3+ models with different loss functions were trained to segment the ciliary body (ciliary muscle and processes) in both en face and radial images.ResultsWe trained and tested the models on 4320 radial and 3864 en face images from 12 cadaver eye volumes. Deep learning models trained on radial images with Dice loss achieved the highest mean F1-score (0.89) for ciliary body segmentation. For three-class segmentation (ciliary muscle, processes, and background), radial images with Dice loss achieved the highest mean F1-score (0.75 for the ciliary process and 0.82 for the ciliary muscle). Part of the ciliary muscle (10.9%) was misclassified as the ciliary process and vice versa, which occurred owing to the difficulty in differentiating the ciliary muscle-processes border, even by experts. Deep learning segmentation made further editing by experts at least seven times faster than a fully manual approach. In eight cadaver eyes, the average ciliary muscle, process, and body volumes were 56 ± 9, 43 ± 13, and 99 ± 18 mm3, respectively. The average surface area of the ciliary muscle, process, and body were 346 ± 45, 363 ± 83, and 709 ± 80 mm2, respectively. We performed transscleral cyclophotocoagulation in cadaver eyes to shrink the ciliary processes. Both manual and automated measurements from deep learning segmentation show a decrease in volume, surface area, and 360° cross-sectional area measurements.ConclusionsThe proposed deep learning segmentation of the ciliary body and 3D measurements showed transscleral cyclophotocoagulation-related changes in the ciliary body.Translational relevanceAutomated ciliary body assessment using 3D-UBM has the translational potential for ophthalmic treatment planning and monitoring.

Project description:PurposeTo generate fully automated and fast 4D-flow MRI-based 3D segmentations of the aorta using deep learning for reproducible quantification of aortic flow, peak velocity, and dimensions.MethodsA total of 1018 subjects with aortic 4D-flow MRI (528 with bicuspid aortic valve, 376 with tricuspid aortic valve and aortic dilation, 114 healthy controls) comprised the data set. A convolutional neural network was trained to generate 3D aortic segmentations from 4D-flow data. Manual segmentations served as the ground truth (N = 499 training, N = 101 validation, N = 418 testing). Dice scores, Hausdorff distance, and average symmetrical surface distance were calculated to assess performance. Aortic flow, peak velocity, and lumen dimensions were quantified at the ascending, arch, and descending aorta and compared using Bland-Altman analysis. Interobserver variability of manual analysis was assessed on a subset of 40.ResultsConvolutional neural network segmentation required 0.438 ± 0.355 seconds versus 630 ± 254 seconds for manual analysis and demonstrated excellent performance with a median Dice score of 0.951 (0.930-0.966), Hausdorff distance of 2.80 (2.13-4.35), and average symmetrical surface distance of 0.176 (0.119-0.290). Excellent agreement was found for flow, peak velocity, and dimensions with low bias and limits of agreement less than 10% difference versus manual analysis. For aortic volume, limits of agreement were moderate within 16.3%. Interobserver variability (median Dice score: 0.950; Hausdorff distance: 2.45; and average symmetrical surface distance: 0.145) and convolutional neural network-based analysis (median Dice score: 0.953-0.959; Hausdorff distance: 2.24-2.91; and average symmetrical surface distance: 0.145-1.98 to observers) demonstrated similar reproducibility.ConclusionsDeep learning enabled fast and automated 3D aortic segmentation from 4D-flow MRI, demonstrating its potential for efficient clinical workflows. Future studies should investigate its utility for other vasculature and multivendor applications.

Project description:ObjectivesTo investigate the clinical utility of fully-automated 3D organ segmentation in assessing hepatic steatosis on pre-contrast and post-contrast CT images using magnetic resonance spectroscopy (MRS)-proton density fat fraction (PDFF) as reference standard.Materials and methodsThis retrospective study analyzed 362 adult potential living liver donors with abdominal CT scans and MRS-PDFF. Using a deep learning-based tool, mean volumetric CT attenuation of the liver and spleen were measured on pre-contrast (liver(L)_pre and spleen(S)_pre) and post-contrast (L_post and S_post) images. Agreements between volumetric and manual region-of-interest (ROI)-based measurements were assessed using the intraclass correlation coefficient (ICC) and Bland-Altman analysis. Diagnostic performances of volumetric parameters (L_pre, liver-minus-spleen (L-S)_pre, L_post, and L-S_post) were evaluated for detecting MRS-PDFF ≥ 5% and ≥ 10% using receiver operating characteristic (ROC) curve analysis and compared with those of ROI-based parameters.ResultsAmong the 362 subjects, 105 and 35 had hepatic steatosis with MRS-PDFF ≥ 5% and ≥ 10%, respectively. Volumetric and ROI-based measurements revealed ICCs of 0.974, 0.825, 0.992, and 0.962, with mean differences of -4.2 HU, -3.4 HU, -1.2 HU, and -7.7 HU for L_pre, S_pre, L_post, and S_post, respectively. Volumetric L_pre, L-S_pre, L_post, and L-S_post yielded areas under the ROC curve of 0.813, 0.813, 0.734, and 0.817 for MRS-PDFF ≥ 5%; and 0.901, 0.915, 0.818, and 0.868 for MRS-PDFF ≥ 10%, comparable with those of ROI-based parameters (0.735-0.818; and 0.816-0.895, Ps = 0.228-0.911).ConclusionAutomated 3D segmentation of the liver and spleen in CT scans can provide volumetric CT attenuation-based parameters to detect and grade hepatic steatosis, applicable to pre-contrast and post-contrast images.Clinical relevance statementVolumetric CT attenuation-based parameters of the liver and spleen, obtained through automated segmentation tools from pre-contrast or post-contrast CT scans, can efficiently detect and grade hepatic steatosis, making them applicable for large population data collection.Key points• Automated organ segmentation enables the extraction of CT attenuation-based parameters for the target organ. • Volumetric liver and spleen CT attenuation-based parameters are highly accurate in hepatic steatosis assessment. • Automated CT measurements from pre- or post-contrast imaging show promise for hepatic steatosis screening in large cohorts.

Project description:The quantification of new or enlarged lesions from follow-up MRI scans is an important surrogate of clinical disease activity in patients with multiple sclerosis (MS). Not only is manual segmentation time consuming, but inter-rater variability is high. Currently, only a few fully automated methods are available. We address this gap in the field by employing a 3D convolutional neural network (CNN) with encoder-decoder architecture for fully automatic longitudinal lesion segmentation. Input data consist of two fluid attenuated inversion recovery (FLAIR) images (baseline and follow-up) per patient. Each image is entered into the encoder and the feature maps are concatenated and then fed into the decoder. The output is a 3D mask indicating new or enlarged lesions (compared to the baseline scan). The proposed method was trained on 1809 single point and 1444 longitudinal patient data sets and then validated on 185 independent longitudinal data sets from two different scanners. From the two validation data sets, manual segmentations were available from three experienced raters, respectively. The performance of the proposed method was compared to the open source Lesion Segmentation Toolbox (LST), which is a current state-of-art longitudinal lesion segmentation method. The mean lesion-wise inter-rater sensitivity was 62%, while the mean inter-rater number of false positive (FP) findings was 0.41 lesions per case. The two validated algorithms showed a mean sensitivity of 60% (CNN), 46% (LST) and a mean FP of 0.48 (CNN), 1.86 (LST) per case. Sensitivity and number of FP were not significantly different (p < 0.05) between the CNN and manual raters. New or enlarged lesions counted by the CNN algorithm appeared to be comparable with manual expert ratings. The proposed algorithm seems to outperform currently available approaches, particularly LST. The high inter-rater variability in case of manual segmentation indicates the complexity of identifying new or enlarged lesions. An automated CNN-based approach can quickly provide an independent and deterministic assessment of new or enlarged lesions from baseline to follow-up scans with acceptable reliability.

Dataset Information

Fully automated, real-time 3D ultrasound segmentation to estimate first trimester placental volume using deep learning.

Publications

Fully automated, real-time 3D ultrasound segmentation to estimate first trimester placental volume using deep learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets