Dataset Information

Technical Note: Deriving ventilation imaging from 4DCT by deep convolutional neural network.

ABSTRACT: PURPOSE:Ventilation images can be derived from four-dimensional computed tomography (4DCT) by analyzing the change in HU values and deformable vector fields between different respiration phases of computed tomography (CT). As deformable image registration (DIR) is involved, accuracy of 4DCT-derived ventilation image is sensitive to the choice of DIR algorithms. To overcome the uncertainty associated with DIR, we develop a method based on deep convolutional neural network (CNN) to derive ventilation images directly from the 4DCT without explicit image registration. METHODS:A total of 82 sets of 4DCT and ventilation images from patients with lung cancer were used in this study. In the proposed CNN architecture, the CT two-channel input data consist of CT at the end of exhale and the end of inhale phases. The first convolutional layer has 32 different kernels of size 5 × 5 × 5, followed by another eight convolutional layers each of which is equipped with an activation layer (ReLU). The loss function is the mean-squared-error (MSE) to measure the intensity difference between the predicted and reference ventilation images. RESULTS:The predicted images were comparable to the label images of the test data. The similarity index, correlation coefficient, and Gamma index passing rate averaged over the tenfold cross validation were 0.880 ± 0.035, 0.874 ± 0.024, and 0.806 ± 0.014, respectively. CONCLUSIONS:The results demonstrate that deep CNN can generate ventilation imaging from 4DCT without explicit deformable image registration, reducing the associated uncertainty.

SUBMITTER: Zhong Y

PROVIDER: S-EPMC7098066 | biostudies-literature | 2019 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Technical Note: Deriving ventilation imaging from 4DCT by deep convolutional neural network.

Zhong Yuncheng Y Vinogradskiy Yevgeniy Y Chen Liyuan L Myziuk Nick N Castillo Richard R Castillo Edward E Guerrero Thomas T Jiang Steve S Wang Jing J

Medical physics 20190312 5

<h4>Purpose</h4>Ventilation images can be derived from four-dimensional computed tomography (4DCT) by analyzing the change in HU values and deformable vector fields between different respiration phases of computed tomography (CT). As deformable image registration (DIR) is involved, accuracy of 4DCT-derived ventilation image is sensitive to the choice of DIR algorithms. To overcome the uncertainty associated with DIR, we develop a method based on deep convolutional neural network (CNN) to derive ...[more]

PMID: 30714159

Similar Datasets

Project description:PurposeThe computed tomography (CT)-derived ventilation imaging methodology employs deformable image registration (DIR) to recover respiratory motion-induced volume changes from an inhale/exhale CT image pair, as a surrogate for ventilation. The Integrated Jacobian Formulation (IJF) and Mass Conserving Volume Change (MCVC) numerical methods for volume change estimation represent two classes of ventilation methods, namely transformation based and intensity (Hounsfield Unit) based, respectively. Both the IJF and MCVC methods utilize subregional volume change measurements that satisfy a specified uncertainty tolerance. In previous publications, the ventilation images resulting from this numerical strategy demonstrated robustness to DIR variations. However, the reduced measurement uncertainty comes at the expense of measurement resolution. The purpose of this study was to examine the spatial correlation between robust CT-ventilation images and single photon emission CT-ventilation (SPECT-V).MethodsPreviously described implementations of IJF and MCVC require the solution of a large scale, constrained linear least squares problem defined by a series of robust subregional volume change measurements. We introduce a simpler parameterized implementation that reduces the number of unknowns while increasing the number of data points in the resulting least squares problem. A parameter sweep of the measurement uncertainty tolerance, τ , was conducted using the 4DCT and SPECT-V images acquired for 15 non-small cell lung cancer patients prior to radiotherapy. For each test case, MCVC and IJF CT-ventilation images were created for 30 different uncertainty parameter values, uniformly sampled from the range 0.01,0.25 . Voxel-wise Spearman correlation between the SPECT-V and the resulting CT-ventilation images was computed.ResultsThe median correlations between MCVC and SPECT-V ranged from 0.20 to 0.48 across the parameter sweep, while the median correlations for IJF and SPECT-V ranged between 0.79 and 0.82. For the optimal IJF tolerance τ=0.07 , the IJF and SPECT-V correlations across all 15 test cases ranged between 0.12 and 0.90. For the optimal MCVC tolerance τ=0.03 , the MCVC and SPECT-V correlations across all 15 test cases ranged between -0.06 and 0.84.ConclusionThe reported correlations indicate that robust methods generate ventilation images that are spatially consistent with SPECT-V, with the transformation-based IJF method yielding higher correlations than those previously reported in the literature. For both methods, overall correlations were found to marginally vary for τ∈[0.03,0.15] , indicating that the clinical utility of both methods is robust to both uncertainty tolerance and DIR solution.

Project description:OBJECTIVES:To develop a proof-of-concept "interpretable" deep learning prototype that justifies aspects of its predictions from a pre-trained hepatic lesion classifier. METHODS:A convolutional neural network (CNN) was engineered and trained to classify six hepatic tumor entities using 494 lesions on multi-phasic MRI, described in Part 1. A subset of each lesion class was labeled with up to four key imaging features per lesion. A post hoc algorithm inferred the presence of these features in a test set of 60 lesions by analyzing activation patterns of the pre-trained CNN model. Feature maps were generated that highlight regions in the original image that correspond to particular features. Additionally, relevance scores were assigned to each identified feature, denoting the relative contribution of a feature to the predicted lesion classification. RESULTS:The interpretable deep learning system achieved 76.5% positive predictive value and 82.9% sensitivity in identifying the correct radiological features present in each test lesion. The model misclassified 12% of lesions. Incorrect features were found more often in misclassified lesions than correctly identified lesions (60.4% vs. 85.6%). Feature maps were consistent with original image voxels contributing to each imaging feature. Feature relevance scores tended to reflect the most prominent imaging criteria for each class. CONCLUSIONS:This interpretable deep learning system demonstrates proof of principle for illuminating portions of a pre-trained deep neural network's decision-making, by analyzing inner layers and automatically describing features contributing to predictions. KEY POINTS:• An interpretable deep learning system prototype can explain aspects of its decision-making by identifying relevant imaging features and showing where these features are found on an image, facilitating clinical translation. • By providing feedback on the importance of various radiological features in performing differential diagnosis, interpretable deep learning systems have the potential to interface with standardized reporting systems such as LI-RADS, validating ancillary features and improving clinical practicality. • An interpretable deep learning system could potentially add quantitative data to radiologic reports and serve radiologists with evidence-based decision support.

Project description:Motivation:A majority of known genetic variants associated with human-inherited diseases lie in non-coding regions that lack adequate interpretation, making it indispensable to systematically discover functional sites at the whole genome level and precisely decipher their implications in a comprehensive manner. Although computational approaches have been complementing high-throughput biological experiments towards the annotation of the human genome, it still remains a big challenge to accurately annotate regulatory elements in the context of a specific cell type via automatic learning of the DNA sequence code from large-scale sequencing data. Indeed, the development of an accurate and interpretable model to learn the DNA sequence signature and further enable the identification of causative genetic variants has become essential in both genomic and genetic studies. Results:We proposed Deopen, a hybrid framework mainly based on a deep convolutional neural network, to automatically learn the regulatory code of DNA sequences and predict chromatin accessibility. In a series of comparison with existing methods, we show the superior performance of our model in not only the classification of accessible regions against background sequences sampled at random, but also the regression of DNase-seq signals. Besides, we further visualize the convolutional kernels and show the match of identified sequence signatures and known motifs. We finally demonstrate the sensitivity of our model in finding causative noncoding variants in the analysis of a breast cancer dataset. We expect to see wide applications of Deopen with either public or in-house chromatin accessibility data in the annotation of the human genome and the identification of non-coding variants associated with diseases. Availability and implementation:Deopen is freely available at https://github.com/kimmo1019/Deopen. Contact:ruijiang@tsinghua.edu.cn. Supplementary information:Supplementary data are available at Bioinformatics online.

Project description:PurposeTo describe and evaluate a new segmentation method using deep convolutional neural network (CNN), 3D fully connected conditional random field (CRF), and 3D simplex deformable modeling to improve the efficiency and accuracy of knee joint tissue segmentation.MethodsA segmentation pipeline was built by combining a semantic segmentation CNN, 3D fully connected CRF, and 3D simplex deformable modeling. A convolutional encoder-decoder network was designed as the core of the segmentation method to perform high resolution pixel-wise multi-class tissue classification for 12 different joint structures. The 3D fully connected CRF was applied to regularize contextual relationship among voxels within the same tissue class and between different classes. The 3D simplex deformable modeling refined the output from 3D CRF to preserve the overall shape and maintain a desirable smooth surface for joint structures. The method was evaluated on 3D fast spin-echo (3D-FSE) MR image data sets. Quantitative morphological metrics were used to evaluate the accuracy and robustness of the method in comparison to the ground truth data.ResultsThe proposed segmentation method provided good performance for segmenting all knee joint structures. There were 4 tissue types with high mean Dice coefficient above 0.9 including the femur, tibia, muscle, and other non-specified tissues. There were 7 tissue types with mean Dice coefficient between 0.8 and 0.9 including the femoral cartilage, tibial cartilage, patella, patellar cartilage, meniscus, quadriceps and patellar tendon, and infrapatellar fat pad. There was 1 tissue type with mean Dice coefficient between 0.7 and 0.8 for joint effusion and Baker's cyst. Most musculoskeletal tissues had a mean value of average symmetric surface distance below 1 mm.ConclusionThe combined CNN, 3D fully connected CRF, and 3D deformable modeling approach was well-suited for performing rapid and accurate comprehensive tissue segmentation of the knee joint. The deep learning-based segmentation method has promising potential applications in musculoskeletal imaging.

Dataset Information

Technical Note: Deriving ventilation imaging from 4DCT by deep convolutional neural network.

Publications

Technical Note: Deriving ventilation imaging from 4DCT by deep convolutional neural network.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure