scholarly journals Three-Dimensional Semantic Segmentation of Pituitary Adenomas Based on the Deep Learning Framework-nnU-Net: A Clinical Perspective

Micromachines ◽  
2021 ◽  
Vol 12 (12) ◽  
pp. 1473
Author(s):  
Xujun Shu ◽  
Yijie Zhou ◽  
Fangye Li ◽  
Tao Zhou ◽  
Xianghui Meng ◽  
...  
Keyword(s):  
Deep Learning ◽  
Clinical Practice ◽  
Pituitary Adenomas ◽  
Three Dimensional ◽  
Cost Effective ◽  
Semantic Segmentation ◽  
Dice Similarity Coefficient ◽  
Learning Framework ◽  
Two Phases ◽  
Practice Methods

This study developed and evaluated nnU-Net models for three-dimensional semantic segmentation of pituitary adenomas (PAs) from contrast-enhanced T1 (T1ce) images, with aims to train a deep learning-based model cost-effectively and apply it to clinical practice. Methods: This study was conducted in two phases. In phase one, two models were trained with nnUNet using distinct PA datasets. Model 1 was trained with 208 PAs in total, and model 2 was trained with 109 primary nonfunctional pituitary adenomas (NFPA). In phase two, the performances of the two models were investigated according to the Dice similarity coefficient (DSC) in the leave-out test dataset. Results: Both models performed well (DSC > 0.8) for PAs with volumes > 1000 mm3, but unsatisfactorily (DSC < 0.5) for PAs < 1000 mm3. Conclusions: Both nnU-Net models showed good segmentation performance for PAs > 1000 mm3 (75% of the dataset) and limited performance for PAs < 1000 mm3 (25% of the dataset). Model 2 trained with fewer samples was more cost-effective. We propose to combine the use of model-based segmentation for PA > 1000 mm3 and manual segmentation for PA < 1000 mm3 in clinical practice at the current stage.

scholarly journals Automated Segmentation of Infarct Lesions in T1-Weighted MRI Scans Using Variational Mode Decomposition and Deep Learning

Sensors ◽  
2021 ◽  
Vol 21 (6) ◽  
pp. 1952
Author(s):  
May Phu Paing ◽  
Supan Tungjitkusolmun ◽  
Toan Huy Bui ◽  
Sarinporn Visitsattapongse ◽  
Chuchart Pintavirooj
Keyword(s):  
Deep Learning ◽  
Brain Infarction ◽  
Three Dimensional ◽  
Dice Similarity Coefficient ◽  
Automated Segmentation ◽  
Variational Mode Decomposition ◽  
Brain Scans ◽  
Automated Method ◽  
Mode Decomposition ◽  
Segmentation Task

Automated segmentation methods are critical for early detection, prompt actions, and immediate treatments in reducing disability and death risks of brain infarction. This paper aims to develop a fully automated method to segment the infarct lesions from T1-weighted brain scans. As a key novelty, the proposed method combines variational mode decomposition and deep learning-based segmentation to take advantages of both methods and provide better results. There are three main technical contributions in this paper. First, variational mode decomposition is applied as a pre-processing to discriminate the infarct lesions from unwanted non-infarct tissues. Second, overlapped patches strategy is proposed to reduce the workload of the deep-learning-based segmentation task. Finally, a three-dimensional U-Net model is developed to perform patch-wise segmentation of infarct lesions. A total of 239 brain scans from a public dataset is utilized to develop and evaluate the proposed method. Empirical results reveal that the proposed automated segmentation can provide promising performances with an average dice similarity coefficient (DSC) of 0.6684, intersection over union (IoU) of 0.5022, and average symmetric surface distance (ASSD) of 0.3932, respectively.

Automated Pancreas Segmentation And Volumetry Using Deep Neural Network On Computed Tomography

2021 ◽  
Author(s):  
Sang-Heon Lim ◽  
Young Jae Kim ◽  
Yeon-Ho Park ◽  
Doojin Kim ◽  
Kwang Gi Kim ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Deep Learning ◽  
Automatic Segmentation ◽  
External Validation ◽  
Three Dimensional ◽  
Semantic Segmentation ◽  
Quantitative Information ◽  
Similarity Coefficients ◽  
Internal Validation ◽  
Pancreas Segmentation

Abstract Pancreas segmentation is necessary for observing lesions, analyzing anatomical structures, and predicting patient prognosis. Therefore, various studies have designed segmentation models based on convolutional neural networks for pancreas segmentation. However, the deep learning approach is limited by a lack of data, and studies conducted on a large computed tomography dataset are scarce. Therefore, this study aims to perform deep-learning-based semantic segmentation on 1,006 participants and evaluate the automatic segmentation performance of the pancreas via four individual three-dimensional segmentation networks. In this study, we performed internal validation with 1,006 patients and external validation using the Cancer Imaging Archive (TCIA) pancreas dataset. We obtained mean precision, recall, and dice similarity coefficients of 0.869, 0.842, and 0.842, respectively, for internal validation via a relevant approach among the four deep learning networks. Using the external dataset, the deep learning network achieved mean precision, recall, and dice similarity coefficients of 0.779, 0.749, and 0.735, respectively. We expect that generalized deep-learning-based systems can assist clinical decisions by providing accurate pancreatic segmentation and quantitative information of the pancreas for abdominal computed tomography.

scholarly journals A computationally efficient approach to segmentation of the aorta and coronary arteries using deep learning

2021 ◽  
Author(s):  
Wing Keung Cheung ◽  
Robert Bell ◽  
Arjun Nair ◽  
Leon Menezies ◽  
Riyaz Patel ◽  
...  
Keyword(s):  
Deep Learning ◽  
Coronary Arteries ◽  
Automatic Segmentation ◽  
Three Dimensional ◽  
Regions Of Interest ◽  
Dice Similarity Coefficient ◽  
Processing Unit ◽  
Two Dimensional ◽  
Computed Tomography Images ◽  
Graphical Processing

AbstractA fully automatic two-dimensional Unet model is proposed to segment aorta and coronary arteries in computed tomography images. Two models are trained to segment two regions of interest, (1) the aorta and the coronary arteries or (2) the coronary arteries alone. Our method achieves 91.20% and 88.80% dice similarity coefficient accuracy on regions of interest 1 and 2 respectively. Compared with a semi-automatic segmentation method, our model performs better when segmenting the coronary arteries alone. The performance of the proposed method is comparable to existing published two-dimensional or three-dimensional deep learning models. Furthermore, the algorithmic and graphical processing unit memory efficiencies are maintained such that the model can be deployed within hospital computer networks where graphical processing units are typically not available.

Abstract 224: A Deep Learning Approach to Left-Ventricular Chamber Quantification for Fully Automated Three Dimensional Strain Analysis in Cardiotoxicity

2020 ◽  
Vol 127 (Suppl_1) ◽  
Author(s):  
Bryant M Baldwin ◽  
Shane Joseph ◽  
Xiaodong Zhong ◽  
Ranya Kakish ◽  
Cherie Revere ◽  
...  
Keyword(s):  
Deep Learning ◽  
Interpolation Method ◽  
Three Dimensional ◽  
Strain Analysis ◽  
Correlation Coefficients ◽  
Semantic Segmentation ◽  
Left Ventricular ◽  
Breast Cancer Patients ◽  
Left Ventricular Chamber ◽  
Longitudinal Strains

This study investigated MRI and semantic segmentation-based deep-learning (SSDL) automation for left-ventricular chamber quantifications (LVCQ) and low longitudinal strain (LLS) determination, thus eliminating user-bias by providing an automated tool to detect cardiotoxicity (CT) in breast cancer patients treated with antineoplastic agents. Displacement Encoding with Stimulated Echoes-based (DENSE) myocardial images from 26 patients were analyzed with the tool’s Convolution Neural Network with underlying Resnet-50 architecture. Quantifications based on the SSDL tool’s output were for LV end-diastolic diameter (LVEDD), ejection fraction (LVEF), and mass (LVM) (see figure for phase sequence). LLS was analyzed with Radial Point Interpolation Method (RPIM) with DENSE phase-based displacements. LVCQs were validated by comparison to measurements obtained with an existing semi-automated vendor tool (VT) and strains by 2 independent users employing Bland-Altman analysis (BAA) and interclass correlation coefficients estimated with Cronbach’s Alpha (C-Alpha) index. F1 score for classification accuracy was 0.92. LVCQs determined by SSDL and VT were 4.6 ± 0.5 vs 4.6 ± 0.7 cm (C-Alpha = 0.93 and BAA = 0.5 ± 0.5 cm) for LVEDD, 58 ± 5 vs 58 ± 6 % (0.90, 1 ± 5%) for LVEF, 119 ± 17 vs 121 ± 14 g (0.93, 5 ± 8 g) for LV mass, while LLS was 14 ± 4 vs 14 ± 3 % (0.86, 0.2 ± 6%). Hence, equivalent LV dimensions, mass and strains measured by VT and DENSE imaging validate our unique automated analytic tool. Longitudinal strains in patients can then be analyzed without user bias to detect abnormalities for the indication of cardiotoxicity and the need for therapeutic intervention even if LVEF is not affected.

scholarly journals Deep learning segmentation of gadolinium-enhancing lesions in multiple sclerosis

2020 ◽  
pp. 135245852092136 ◽  
Author(s):  
Ivan Coronado ◽  
Refaat E Gabr ◽  
Ponnada A Narayana
Keyword(s):  
Multiple Sclerosis ◽  
Deep Learning ◽  
Network Performance ◽  
Three Dimensional ◽  
Proton Density ◽  
Dice Similarity Coefficient ◽  
Lesion Volume ◽  
Imaging Data ◽  
Post Contrast ◽  
Fluid Attenuated Inversion Recovery

Objective: The aim of this study is to assess the performance of deep learning convolutional neural networks (CNNs) in segmenting gadolinium-enhancing lesions using a large cohort of multiple sclerosis (MS) patients. Methods: A three-dimensional (3D) CNN model was trained for segmentation of gadolinium-enhancing lesions using multispectral magnetic resonance imaging data (MRI) from 1006 relapsing–remitting MS patients. The network performance was evaluated for three combinations of multispectral MRI used as input: (U5) fluid-attenuated inversion recovery (FLAIR), T2-weighted, proton density-weighted, and pre- and post-contrast T1-weighted images; (U2) pre- and post-contrast T1-weighted images; and (U1) only post-contrast T1-weighted images. Segmentation performance was evaluated using the Dice similarity coefficient (DSC) and lesion-wise true-positive (TPR) and false-positive (FPR) rates. Performance was also evaluated as a function of enhancing lesion volume. Results: The DSC/TPR/FPR values averaged over all the enhancing lesion sizes were 0.77/0.90/0.23 using the U5 model. These values for the largest enhancement volumes (>500 mm3) were 0.81/0.97/0.04. For U2, the average DSC/TPR/FPR values were 0.72/0.86/0.31. Comparable performance was observed with U1. For all types of input, the network performance degraded with decreased enhancement size. Conclusion: Excellent segmentation of enhancing lesions was observed for enhancement volume ⩾70 mm3. The best performance was achieved when the input included all five multispectral image sets.

scholarly journals Voxel-based 3D Point Cloud Semantic Segmentation: Unsupervised Geometric and Relationship Featuring vs Deep Learning Methods

2019 ◽  
Vol 8 (5) ◽  
pp. 213 ◽  
Author(s):  
Florent Poux ◽  
Roland Billen
Keyword(s):  
Deep Learning ◽  
Point Cloud ◽  
Semantic Representation ◽  
Structural Connectivity ◽  
Three Dimensional ◽  
Strong Support ◽  
Semantic Segmentation ◽  
Point Clouds ◽  
Learning Methods ◽  
Cloud Data

Automation in point cloud data processing is central in knowledge discovery within decision-making systems. The definition of relevant features is often key for segmentation and classification, with automated workflows presenting the main challenges. In this paper, we propose a voxel-based feature engineering that better characterize point clusters and provide strong support to supervised or unsupervised classification. We provide different feature generalization levels to permit interoperable frameworks. First, we recommend a shape-based feature set (SF1) that only leverages the raw X, Y, Z attributes of any point cloud. Afterwards, we derive relationship and topology between voxel entities to obtain a three-dimensional (3D) structural connectivity feature set (SF2). Finally, we provide a knowledge-based decision tree to permit infrastructure-related classification. We study SF1/SF2 synergy on a new semantic segmentation framework for the constitution of a higher semantic representation of point clouds in relevant clusters. Finally, we benchmark the approach against novel and best-performing deep-learning methods while using the full S3DIS dataset. We highlight good performances, easy-integration, and high F1-score (> 85%) for planar-dominant classes that are comparable to state-of-the-art deep learning.

scholarly journals A Novel Object-Based Deep Learning Framework for Semantic Segmentation of Very High-Resolution Remote Sensing Data: Comparison with Convolutional and Fully Convolutional Networks

2019 ◽  
Vol 11 (6) ◽  
pp. 684 ◽  
Author(s):  
Maria Papadomanolaki ◽  
Maria Vakalopoulou ◽  
Konstantinos Karantzalos
Keyword(s):  
Deep Learning ◽  
State Of The Art ◽  
Semantic Segmentation ◽  
Novel Object ◽  
Convolutional Networks ◽  
Learning Framework ◽  
Fully Convolutional Networks ◽  
Object Based ◽  
Deep Networks ◽  
Very High

Deep learning architectures have received much attention in recent years demonstrating state-of-the-art performance in several segmentation, classification and other computer vision tasks. Most of these deep networks are based on either convolutional or fully convolutional architectures. In this paper, we propose a novel object-based deep-learning framework for semantic segmentation in very high-resolution satellite data. In particular, we exploit object-based priors integrated into a fully convolutional neural network by incorporating an anisotropic diffusion data preprocessing step and an additional loss term during the training process. Under this constrained framework, the goal is to enforce pixels that belong to the same object to be classified at the same semantic category. We compared thoroughly the novel object-based framework with the currently dominating convolutional and fully convolutional deep networks. In particular, numerous experiments were conducted on the publicly available ISPRS WGII/4 benchmark datasets, namely Vaihingen and Potsdam, for validation and inter-comparison based on a variety of metrics. Quantitatively, experimental results indicate that, overall, the proposed object-based framework slightly outperformed the current state-of-the-art fully convolutional networks by more than 1% in terms of overall accuracy, while intersection over union results are improved for all semantic categories. Qualitatively, man-made classes with more strict geometry such as buildings were the ones that benefit most from our method, especially along object boundaries, highlighting the great potential of the developed approach.

scholarly journals BUILDING OUTLINE EXTRACTION FROM AERIAL IMAGERY AND DIGITAL SURFACE MODEL WITH A FRAME FIELD LEARNING FRAMEWORK

2021 ◽  
Vol XLIII-B2-2021 ◽  
pp. 487-493
Author(s):  
X. Sun ◽  
W. Zhao ◽  
R. V. Maretto ◽  
C. Persello
Keyword(s):  
Deep Learning ◽  
Semantic Segmentation ◽  
Aerial Images ◽  
Surface Model ◽  
Digital Surface Model ◽  
Frame Field ◽  
Learning Framework ◽  
Elevation Data ◽  
3D Information ◽  
Direction Information

Abstract. Deep learning-based semantic segmentation models for building delineation face the challenge of producing precise and regular building outlines. Recently, a building delineation method based on frame field learning was proposed by Girard et al. (2020) to extract regular building footprints as vector polygons directly from aerial RGB images. A fully convolution network (FCN) is trained to learn simultaneously the building mask, contours, and frame field followed by a polygonization method. With the direction information of the building contours stored in the frame field, the polygonization algorithm produces regular outlines accurately detecting edges and corners. This paper investigated the contribution of elevation data from the normalized digital surface model (nDSM) to extract accurate and regular building polygons. The 3D information provided by the nDSM overcomes the aerial images’ limitations and contributes to distinguishing the buildings from the background more accurately. Experiments conducted in Enschede, the Netherlands, demonstrate that the nDSM improves building outlines’ accuracy, resulting in better-aligned building polygons and prevents false positives. The investigated deep learning approach (fusing RGB + nDSM) results in a mean intersection over union (IOU) of 0.70 in the urban area. The baseline method (using RGB only) results in an IOU of 0.58 in the same area. A qualitative analysis of the results shows that the investigated model predicts more precise and regular polygons for large and complex structures.

Sign in / Sign up

Export Citation Format

Share Document