Search

Lung cancer is by far the leading cause of cancer death in the US. Recent studies have demonstrated the effectiveness of screening using low dose CT (LDCT) in reducing lung cancer related mortality. While lung nodules are detected with a high rate of sensitivity, this exam has a low specificity rate and it is still difficult to separate benign and malignant lesions. The ISBI 2018 Lung Nodule Malignancy Prediction Challenge, developed by a team from the Quantitative Imaging Network of the National Cancer Institute, was focused on the prediction of lung nodule malignancy from two sequential LDCT screening exams using automated (non-manual) algorithms. We curated a cohort of 100 subjects who participated in the National Lung Screening Trial and had established pathological diagnoses. Data from 30 subjects were randomly selected for training and the remaining was used for testing. Participants were evaluated based on the area under the receiver operating characteristic curve (AUC) of nodule-wise malignancy scores generated by their algorithms on the test set. The challenge had 17 participants, with 11 teams submitting reports with method description, mandated by the challenge rules. Participants used quantitative methods, resulting in a reporting test AUC ranging from 0.698 to 0.913. The top five contestants used deep learning approaches, reporting an AUC between 0.87 - 0.91. The team's predictor did not achieve significant differences from each other nor from a volume change estimate (p =.05 with Bonferroni-Holm's correction).

Purpose Most state-of-the-art automated medical image analysis methods for volumetric data rely on adaptations of two-dimensional (2D) and three-dimensional (3D) convolutional neural networks (CNNs). In this paper, we develop a novel unified CNN-based model that combines the benefits of 2D and 3D networks for analyzing volumetric medical images. Methods In our proposed framework, multiscale contextual information is first extracted from 2D slices inside a volume of interest (VOI). This is followed by dilated 1D convolutions across slices to aggregate in-plane features in a slice-wise manner and encode the information in the entire volume. Moreover, we formalize a curriculum learning strategy for a two-stage system (i.e., a system that consists of screening and false positive reduction), where the training samples are presented to the network in a meaningful order to further improve the performance. Results We evaluated the proposed approach by developing a computer-aided detection (CADe) system for lung nodules. Our results on 888 CT exams demonstrate that the proposed approach can effectively analyze volumetric data by achieving a sensitivity of > 0.99 in the screening stage and a sensitivity of > 0.96 at eight false positives per case in the false positive reduction stage. Conclusion Our experimental results show that the proposed method provides competitive results compared to state-of-the-art 3D frameworks. In addition, we illustrate the benefits of curriculum learning strategies in two-stage systems that are of common use in medical imaging applications.

Most of the Deep Neural Networks (DNNs) based CT image denoising literature shows that DNNs outperform traditional iterative methods in terms of metrics such as the RMSE, the PSNR and the SSIM. In many instances, using the same metrics, the DNN results from low-dose inputs are also shown to be comparable to their high-dose counterparts. However, these metrics do not reveal if the DNN results preserve the visibility of subtle lesions or if they alter the CT image properties such as the noise texture. Accordingly, in this work, we seek to examine the image quality of the DNN results from a holistic viewpoint for low-dose CT image denoising. First, we build a library of advanced DNN denoising architectures. This library is comprised of denoising architectures such as the DnCNN, U-Net, Red-Net, GAN, etc. Next, each network is modeled, as well as trained, such that it yields its best performance in terms of the PSNR and SSIM. As such, data inputs (e.g. training patch-size, reconstruction kernel) and numeric-optimizer inputs (e.g. minibatch size, learning rate, loss function) are accordingly tuned. Finally, outputs from thus trained networks are further subjected to a series of CT bench testing metrics such as the contrast-dependent MTF, the NPS and the HU accuracy. These metrics are employed to perform a more nuanced study of the resolution of the DNN outputs' low-contrast features, their noise textures, and their CT number accuracy to better understand the impact each DNN algorithm has on these underlying attributes of image quality., Comment: 13 pages, 9 figures, SPIE proceeding

While several datasets containing CT images of lung nodules exist, they do not contain definitive diagnoses and often rely on radiologists' visual assessment for malignancy rating. This is in spite of the fact that lung cancer is one of the top three most frequently misdiagnosed diseases based on visual assessment. In this paper, we propose a dataset of difficult-to-diagnose lung nodules based on data from the National Lung Screening Trial (NLST), which we refer to as NLSTx. In NLSTx, each malignant nodule has a definitive ground truth label from biopsy. Herein, we also propose a novel deep convolutional neural network (CNN) / recurrent neural network framework that allows for use of pre-trained 2-D convolutional feature extractors, similar to those developed in the ImageNet challenge. Our results show that the proposed framework achieves comparable performance to an equivalent 3-D CNN while requiring half the number of parameters.

Our work aims at developing a machine learning-based image conversion algorithm to adjust quantum noise, sharpness, scattering, and other characteristics of radiographic images acquired with a given imaging system as if they had been acquired with a different acquisition system. Purely physics-based methods which have previously been developed for image conversion rely on the measurement of the physical properties of the acquisition devices, which limit the range of their applicability. In this study, we focused on the conversion of mammographic images from a source acquisition system into a target system using a conditional Generative Adversarial Network (cGAN). This network penalizes any possible structural differences between network-generated and target images. The optimization process was enhanced by designing new reconstruction loss terms which emphasized the quality of high frequency image contents. We trained our cGAN model on a dataset of paired synthetic mammograms and slanted edge phantom images. We coupled one independent slanted edge phantom image with each anthropomorphic breast image and presented the pair as a combined input into the network. To improve network performance at high frequencies, we incorporated an edge-based loss function into the reconstruction loss. Qualitative results demonstrated the feasibility of our method to adjust the sharpness of mammograms acquired with a source system to appear as if the they were acquired with a different target system. Our method was validated by comparing the presampled modulation transfer function (MTF) of the network-generated edge image and the MTF of the source and target mammography acquisition systems at different spatial frequencies. This image conversion technique may help training of machine learning algorithms so that their applicability generalizes to a larger set of medical image acquisition devices. Our work may also facilitate performance assessment of computer-aided detection systems.

SCoTS captures a sparse representation of shapes in an input image through a linear span of previously delineated shapes in a training repository. The model updates shape prior over level set iterations and captures variabilities in shapes by a sparse combination of the training data. The level set evolution is therefore driven by a data term as well as a term capturing valid prior shapes. During evolution, the shape prior influence is adjusted based on shape reconstruction, with the assigned weight determined from the degree of sparsity of the representation. For the problem of lung nodule segmentation in X-ray CT, SCoTS offers a unified framework, capable of segmenting nodules of all types. Experimental validations are demonstrated on 542 3-D lung nodule images from the LIDC-IDRI database. Despite its generality, SCoTS is competitive with domain specific state of the art methods for lung nodule segmentation.

Purpose Multiview two-dimensional (2D) convolutional neural networks (CNNs) and three-dimensional (3D) CNNs have been successfully used for analyzing volumetric data in many state-of-the-art medical imaging applications. We propose an alternative modular framework that analyzes volumetric data with an approach that is analogous to radiologists' interpretation, and apply the framework to reduce false positives that are generated in computer-aided detection (CADe) systems for pulmonary nodules in thoracic computed tomography (CT) scans. Methods In our approach, a deep network consisting of 2D CNNs first processes slices individually. The features extracted in this stage are then passed to a recurrent neural network (RNN), thereby modeling consecutive slices as a sequence of temporal data and capturing the contextual information across all three dimensions in the volume of interest. Outputs of the RNN layer are weighed before the final fully connected layer, enabling the network to scale the importance of different slices within a volume of interest in an end-to-end training framework. Results We validated the proposed architecture on the false positive reduction track of the lung nodule analysis (LUNA) challenge for pulmonary nodule detection in chest CT scans, and obtained competitive results compared to 3D CNNs. Our results show that the proposed approach can encode the 3D information in volumetric data effectively by achieving a sensitivity >0.8 with just 1/8 false positives per scan. Conclusions Our experimental results demonstrate the effectiveness of temporal analysis of volumetric images for the application of false positive reduction in chest CT scans and show that state-of-the-art 2D architectures from the literature can be directly applied to analyzing volumetric medical data. As newer and better 2D architectures are being developed at a much faster rate compared to 3D architectures, our approach makes it easy to obtain state-of-the-art performance on volumetric data using new 2D architectures.

Multiple Instance Learning (MIL) is proposed for Computer Aided Diagnosis (CADx) without predefined Regions Of Interest (ROIs) from lung cancer screening thoracic CT scans. The method was used to classify nodules as malignant or benign on 225 malignant and 210 benign samples from the publicly available Lung Image database consortium Image Collection (LIDC-IDRI). Subsequent to feature selection based on the Gray Level Co-occurrence Matrix (GLCM), 5-fold cross-validation was carried out where training was performed on 4 folds with testing on the 5th fold in a round robin fashion. The classification was performed with Support Vector Machines for Multiple-Instance Learning (MI-SVM) classifier. The proposed method has been compared to the Single Instance Learning (SIL) paradigm based on ground truth regions provided by radiologists and to other state of the art methodologies and was proven to outperform them with resulting average Receiver Operating Characteristic Area Under Curve (AUC), Specificity, Sensitivity and Accuracy of: 0.9767, 0.9524, 0.9111 and 0.9310 respectively.

The conventional graph cuts technique has been widely used for image segmentation due to its ability to find the global minimum and its ease of implementation. However, it is an intensity-based technique and as a result is limited to segmentation applications where there is significant contrast between the object and the background. We modified the conventional graph cuts method by adding shape prior and motion information. Active shape models (ASM) with signed distance functions were used to capture the shape prior information, preventing unwanted surrounding tissue from becoming part of the segmented object. The optical flow method was used to estimate the local motion and to extend 3D segmentation to 4D by warping a prior shape model through time. The method has been applied to segmentation of whole lung boundary and whole liver boundary from respiratory gated CT data. 4D lung boundary segmentation in five patients, and 4D liver boundary segmentation in five patients were performed and in each case, results were compared with the results from expert-delineated ground truth. 4D segmentation for five phases of CT data took approximately ten minutes on a PC workstation with AMD Phenom II and 32GB of memory. An important by-product is quantitative whole organ volumes from respiratory gated CT from end-inspiration to end-expiration which can be determined with high accuracy.

Bag of visual words model has recently attracted much attention from computer vision society because of its notable success in analysing images and exploring their content. This study improves this model by utilising the adjacency information between words. To explore this information, a binary tree structure is constructed from the visual words in order to model the is ? a relationships in the vocabulary. Informative nodes of this tree are extracted by using the X 2 criterion and are used to capture the adjacency information of visual words. This approach is a simple and computationally effective way for modelling the spatial relations of visual words, which improves the image classification performance. The authors evaluated our method for visual classification of three known datasets: 15 natural scenes, Caltech-101 and Graz-01.

Lung tumor segmentation is important for therapy in the radiation treatment of patients with thoracic malignancies. In this paper, we describe a 4D image segmentation method based on graph-cuts optimization, shape prior and optical flow. Due to small size, the location, and low contrast between the tumor and the surrounding tissue, tumor segmentation in 3D+t is challenging. We performed 4D lung tumor segmentation in 5 patients, and in each case compared the results with the expert-delineated lung nodules. In each case, 4D image segmentation took approximately ten minutes on a PC with AMD Phenom II and 32GB of memory for segmenting tumor in five phases of lung CT data.

In this paper, a novel method of embedding shape information into level set image segmentation is proposed. Our method is based on inferring shape variations by a sparse linear combination of instances in the shape repository. Given a sufficient number of training shapes with variations, a new shape can be approximated by a linear span of training shapes associated with those variations. At each step of curve evolution the curve is moved to minimize Chan-Vese energy functional as well as toward the best approximation based on a linear combination of training samples. Although the method is general, in this paper it has been applied to the problem of segmentation of corpus callosum from 2D sagittal MR images.

Bag of visual words (BOW) model is an effective way to represent images in order to classify and detect their contents. However, this type of representation suffers from the fact that, it does not contain any spatial information. In this paper we propose a novel image representation which adds two types of spatial information. The first type which is the spatial locations of the words in the image is added using the spatial pyramid matching approach. The second type is the spatial relation between words. To explore this information a binary tree structure which models the is-a relationships in the vocabulary is constructed from the visual words. This approach is a simple and computationally effective way for modeling the spatial relations of the visual words which shows improvement on the visual classification performance. We evaluated our method on visual classification of two known data sets, namely 15 natural scenes and Caltech-101.