LDCT image biomarkers that matter most for the deep learning classification of indeterminate pulmonary nodules.

BACKGROUND: Continued improvement in deep learning methodologies has increased the rate at which deep neural networks are being evaluated for medical applications, including diagnosis of lung cancer. However, there has been limited exploration of the underlying radiological characteristics that the network relies on to identify lung cancer in computed tomography (CT) images. OBJECTIVE: In this study, we used a combination of image masking and saliency activation maps to systematically explore the contributions of both parenchymal and tumor regions in a CT image to the classification of indeterminate lung nodules. METHODS: We selected individuals from the National Lung Screening Trial (NLST) with solid pulmonary nodules 4-20 mm in diameter. Segmentation masks were used to generate three distinct datasets; 1) an Original Dataset containing the complete low-dose CT scans from the NLST, 2) a Parenchyma-Only Dataset in which the tumor regions were covered by a mask, and 3) a Tumor-Only Dataset in which only the tumor regions were included. RESULTS: The Original Dataset significantly outperformed the Parenchyma-Only Dataset and the Tumor-Only Dataset with an AUC of 80.80 ± 3.77% compared to 76.39 ± 3.16% and 78.11 ± 4.32%, respectively. Gradient-weighted class activation mapping (Grad-CAM) of the Original Dataset showed increased attention was being given to the nodule and the tumor-parenchyma boundary when nodules were classified as malignant. This pattern of attention remained unchanged in the case of the Parenchyma-Only Dataset. Nodule size and first-order statistical features of the nodules were significantly different with the average malignant and benign nodule maximum 3d diameter being 23 mm and 12 mm, respectively. CONCLUSION: We conclude that network performance is linked to textural features of nodules such as kurtosis, entropy and intensity, as well as morphological features such as sphericity and diameter. Furthermore, textural features are more positively associated with malignancy than morphological features.

Association of Pulmonary Function With Late-Life Cardiac Function and Heart Failure Risk: The ARIC Study.

Background Pulmonary and cardiac functions decline with age, but the associations of pulmonary dysfunction with cardiac function and heart failure (HF) risk in late life is not known. We aimed to determine the associations of percent predicted forced vital capacity (ppFVC) and the ratio of forced expired volume in 1 second (FEV1) to forced vital capacity (FVC; FEV1/FVC) with cardiac function and incident HF with preserved or reduced ejection fraction in late life. Methods and Results Among 3854 HF-free participants in the ARIC (Atherosclerosis Risk in Communities) cohort study who underwent echocardiography and spirometry at the fifth study visit (2011-2013), associations of FEV1/FVC and ppFVC with echocardiographic measures, cardiac biomarkers, and risk of HF, HF with preserved ejection fraction, and HF with reduced ejection fraction were assessed. Multivariable linear and Cox regression models adjusted for demographics, body mass index, coronary disease, atrial fibrillation, hypertension, and diabetes. Mean age was 75±5 years, 40% were men, 19% were Black, and 61% were ever smokers. Mean FEV1/FVC was 72±8%, and ppFVC was 98±17%. In adjusted analyses, lower FEV1/FVC and ppFVC were associated with higher NT-proBNP (N-terminal pro-B-type natriuretic peptide; both P<0.001) and pulmonary artery pressure (P<0.004). Lower ppFVC was also associated with higher left ventricular mass, left ventricular filling pressure, and high-sensitivity C-reactive protein (all P<0.01). Lower FEV1/FVC was associated with a trend toward higher risk of incident HF with preserved ejection fraction (hazard ratio [HR] per 10-point decrease, 1.31; 95% CI, 0.98-1.74; P=0.07) and HF with reduced ejection fraction (HR per 10-point decrease, 1.24; 95% CI, 0.91-1.70; P=0.18), but these associations did not reach statistical significance. Lower ppFVC was associated with incident HF with preserved ejection fraction (HR per 10-unit decrease, 1.21; 95% CI, 1.04-1.41; P=0.013) but not with HF with reduced ejection fraction (HR per 10-unit decrease, 0.90; 95% CI, 0.76-1.07; P=0.24). Conclusions Subclinical reductions in FEV1/FVC and ppFVC differentially associate with cardiac function and HF risk in late life.

EMPHYSEMA QUANTIFICATION ON SIMULATED X-RAYS THROUGH DEEP LEARNING TECHNIQUES.

Emphysema quantification techniques rely on the use of CT scans, but they are rarely used in the diagnosis and management of patients with COPD; X-ray films are the preferred method to do this. However, this diagnosis method is very controversial, as there are not established guidelines to define the disease, sensitivity is low, and quantification cannot be done. We developed a quantification method based on a CNN, capable of predicting the emphysema percentage of a patient based on an X-ray image. We used real CT scans to simulate X-ray films and to calculate emphysema percentage using the LAA%. The model developed was able to calculate emphysema percentage with an LAA% mean error of 3.96, and it obtained an AUC accuracy of 90.73% for an emphysema definition of ≥10%, with a mean sensitivity of 85.68%, significantly improving X-ray-based emphysema diagnosis.

Statistical Framework for the Definition of Emphysema in CT Scans: Beyond Density Mask.

Lung parenchyma destruction (emphysema) is a major factor in the description of Chronic Obstructive Pulmonary Disease (COPD) and its prognosis. It is defined as an abnormal enlargement of air spaces distal to the terminal bronchioles and the destruction of alveolar walls. In CT imaging, the presence of emphysema is observed by a local decrease of the lung density and the diagnose is usually set as more than 5% of the lung below -950 HU, the so-called emphysema density mask. There is still debate, however, about the definition of this percentage and many researchers set it depending on the population under study. Additionally, the -950 HU threshold may vary depending on factors as the slice thickness or the respiratory phase of the acquisition. In this paper we propose (1) a statistical framework that provides an automatic definition of the density threshold based on the statistical characterization of air and lung parenchyma; (2) the definition of a statistical test for emphysema detection that accounts for the CT noise characteristics. Results show that this novel statistical framework improves the quantification of emphysema against a visual reference and improves the association of emphysema with the pulmonary function tests.

On the Relevance of the Loss Function in the Agatston Score Regression from Non-ECG Gated CT Scans.

In this work, we evaluate the relevance of the choice of loss function in the regression of the Agatston score from 3D heart volumes obtained from non-contrast non-ECG gated chest computed tomography scans. The Agatston score is a well-established metric of cardiovascular disease, where an index of coronary artery disease (CAD) is computed by segmenting the calcifications of the arteries and multiplying each calcification by a factor related to their intensity and their volume, creating a final aggregated index. Recent work has automated such task with deep learning techniques, even skipping the segmentation step and performing a direct regression of the Agatston score. We study the effect of the choice of the loss function in such methodologies. We use a large database of 6983 CT scans to which the Agatston score has been manually computed. The dataset is split into a training set and a validation set of n = 1000. We train a deep learning regression network using such data with different loss functions while keeping the structure of the network and training parameters constant. Pearson correlation coefficient ranges from 0.902 to 0.938 depending on the loss function. Correct risk group assignment measurements range between 59.5% and 81.7%. There is a trade-off between the accuracy of the Pearson correlation coefficient and the risk group measurement, which leads to optimize for one or the other.

Accurate Measurement of Airway Morphology on Chest CT Images.

In recent years, the ability to accurately measuring and analyzing the morphology of small pulmonary structures on chest CT images, such as airways, is becoming of great interest in the scientific community. As an example, in COPD the smaller conducting airways are the primary site of increased resistance in COPD, while small changes in airway segments can identify early stages of bronchiectasis. To date, different methods have been proposed to measure airway wall thickness and airway lumen, but traditional algorithms are often limited due to resolution and artifacts in the CT image. In this work, we propose a Convolutional Neural Regressor (CNR) to perform cross-sectional measurements of airways, considering wall thickness and airway lumen at once. To train the networks, we developed a generative synthetic model of airways that we refined using a Simulated and Unsupervised Generative Adversarial Network (SimGAN). We evaluated the proposed method by first computing the relative error on a dataset of synthetic images refined with SimGAN, in comparison with other methods. Then, due to the high complexity to create an in-vivo ground-truth, we performed a validation on an airway phantom constructed to have airways of different sizes. Finally, we carried out an indirect validation analyzing the correlation between the percentage of the predicted forced expiratory volume in one second (FEV1%) and the value of the Pi10 parameter. As shown by the results, the proposed approach paves the way for the use of CNNs to precisely and accurately measure small lung airways with high accuracy.

Diffeomorphic Lung Registration Using Deep CNNs and Reinforced Learning.

Image registration is a well-known problem in the field of medical imaging. In this paper, we focus on the registration of chest inspiratory and expiratory computed tomography (CT) scans from the same patient. Our method recovers the diffeomorphic elastic displacement vector field (DVF) by jointly regressing the direct and the inverse transformation. Our architecture is based on the RegNet network but we implement a reinforced learning strategy that can accommodate a large training dataset. Our results show that our method performs with a lower estimation error for the same number of epochs than the RegNet approach.

Multi-structure Segmentation from Partially Labeled Datasets. Application to Body Composition Measurements on CT Scans.

Labeled data is the current bottleneck of medical image research. Substantial efforts are made to generate segmentation masks to characterize a given organ. The community ends up with multiple label maps of individual structures in different cases, not suitable for current multi-organ segmentation frameworks. Our objective is to leverage segmentations from multiple organs in different cases to generate a robust multi-organ deep learning segmentation network. We propose a modified cost-function that takes into account only the voxels labeled in the image, ignoring unlabeled structures. We evaluate the proposed methodology in the context of pectoralis muscle and subcutaneous fat segmentation on chest CT scans. Six different structures are segmented from an axial slice centered on the transversal aorta. We compare the performance of a network trained on 3,000 images where only one structure has been annotated (PUNet) against six UNets (one per structure) and a multi-class UNet trained on 500 completely annotated images, showing equivalence between the three methods (Dice coefficients of 0.909, 0.906 and 0.909 respectively). We further propose a modification of the architecture by adding convolutions to the skip connections (CUNet). When trained with partially labeled images, it outperforms statistically significantly the other three methods (Dice 0.916, p< 0.0001). We, therefore, show that (a) when keeping the number of organ annotation constant, training with partially labeled images is equivalent to training with wholly labeled data and (b) adding convolutions in the skip connections improves performance.

3D Pulmonary Artery Segmentation from CTA Scans Using Deep Learning with Realistic Data Augmentation.

The characterization of the vasculature in the mediastinum, more specifically the pulmonary artery, is of vital importance for the evaluation of several pulmonary vascular diseases. Thus, the goal of this study is to automatically segment the pulmonary artery (PA) from computed tomography angiography images, which opens up the opportunity for more complex analysis of the evolution of the PA geometry in health and disease and can be used in complex fluid mechanics models or individualized medicine. For that purpose, a new 3D convolutional neural network architecture is proposed, which is trained on images coming from different patient cohorts. The network makes use a strong data augmentation paradigm based on realistic deformations generated by applying principal component analysis to the deformation fields obtained from the affine registration of several datasets. The network is validated on 91 datasets by comparing the automatic segmentations with semi-automatically delineated ground truths in terms of mean Dice and Jaccard coefficients and mean distance between surfaces, which yields values of 0.89, 0.80 and 1.25 mm, respectively. Finally, a comparison against a Unet architecture is also included.

A Bayesian Nonparametric Model for Disease Subtyping: Application to Emphysema Phenotypes.

We introduce a novel Bayesian nonparametric model that uses the concept of disease trajectories for disease subtype identification. Although our model is general, we demonstrate that by treating fractions of tissue patterns derived from medical images as compositional data, our model can be applied to study distinct progression trends between population subgroups. Specifically, we apply our algorithm to quantitative emphysema measurements obtained from chest CT scans in the COPDGene Study and show several distinct progression patterns. As emphysema is one of the major components of chronic obstructive pulmonary disease (COPD), the third leading cause of death in the United States [1], an improved definition of emphysema and COPD subtypes is of great interest. We investigate several models with our algorithm, and show that one with age , pack years (a measure of cigarette exposure), and smoking status as predictors gives the best compromise between estimated predictive performance and model complexity. This model identified nine subtypes which showed significant associations to seven single nucleotide polymorphisms (SNPs) known to associate with COPD. Additionally, this model gives better predictive accuracy than multiple, multivariate ordinary least squares regression as demonstrated in a five-fold cross validation analysis. We view our subtyping algorithm as a contribution that can be applied to bridge the gap between CT-level assessment of tissue composition to population-level analysis of compositional trends that vary between disease subtypes.