skip to main content

Title: Improving clinical disease subtyping and future events prediction through a chest CT-based deep learning approach
Purpose: To develop and evaluate a deep learning (DL) approach to extract rich information from high-resolution computed tomography (HRCT) of patients with chronic obstructive pulmonary disease (COPD). Methods: We develop a DL-based model to learn a compact representation of a subject, which is predictive of COPD physiologic severity and other outcomes. Our DL model learned: (a) to extract informative regional image features from HRCT; (b) to adaptively weight these features and form an aggregate patient representation; and finally, (c) to predict several COPD outcomes. The adaptive weights correspond to the regional lung contribution to the disease. We evaluate the model on 10 300 participants from the COPDGene cohort. Results: Our model was strongly predictive of spirometric obstruction ( r2 = 0.67) and grouped 65.4% of subjects correctly and 89.1% within one stage of their GOLD severity stage. Our model achieved an accuracy of 41.7% and 52.8% in stratifying the population-based on centrilobular (5-grade) and paraseptal (3-grade) emphysema severity score, respectively. For predicting future exacerbation, combining subjects' representations from our model with their past exacerbation histories achieved an accuracy of 80.8% (area under the ROC curve of 0.73). For all-cause mortality, in Cox regression analysis, we outperformed the BODE index improving the concordance more » metric (ours: 0.61 vs BODE: 0.56). Conclusions: Our model independently predicted spirometric obstruction, emphysema severity, exacerbation risk, and mortality from CT imaging alone. This method has potential applicability in both research and clinical practice. « less
Award ID(s):
Publication Date:
Journal Name:
Medical physics
Page Range or eLocation-ID:
Sponsoring Org:
National Science Foundation
More Like this
  1. Probabilistic topic models, have been widely deployed for various applications such as learning disease or tissue subtypes. Yet, learning the parameters of such models is usually an ill-posed problem and may result in losing valuable information about disease severity. A common approach is to add a discriminative loss term to the generative model’s loss in order to learn a representation that is also predictive of disease severity. However, finding a balance between these two losses is not straightforward. We propose an alternative way in this paper. We develop a framework which allows for incorporating external covariates into the generative model’s approximate posterior. These covariates can have more discriminative power for disease severity compared to the representation that we extract from the posterior distribution. For instance, they can be features extracted from a neural network which predicts disease severity from CT images. Effectively, we enforce the generative model’s approximate posterior to reside in the subspace of these discriminative covariates. We illustrate our method’s application on a large-scale lung CT study of Chronic Obstructive Pulmonary Disease (COPD), a highly heterogeneous disease. We aim at identifying tissue subtypes by using a variant of topic model as a generative model. We quantitatively evaluate themore »predictive performance of the inferred subtypes and demonstrate that our method outperforms or performs on par with some reasonable baselines. We also show that some of the discovered subtypes are correlated with genetic measurements, suggesting that the identified subtypes may characterize the disease’s underlying etiology.« less
  2. Summary In this article, we develop a graphical modeling framework for the inference of networks across multiple sample groups and data types. In medical studies, this setting arises whenever a set of subjects, which may be heterogeneous due to differing disease stage or subtype, is profiled across multiple platforms, such as metabolomics, proteomics, or transcriptomics data. Our proposed Bayesian hierarchical model first links the network structures within each platform using a Markov random field prior to relate edge selection across sample groups, and then links the network similarity parameters across platforms. This enables joint estimation in a flexible manner, as we make no assumptions on the directionality of influence across the data types or the extent of network similarity across the sample groups and platforms. In addition, our model formulation allows the number of variables and number of subjects to differ across the data types, and only requires that we have data for the same set of groups. We illustrate the proposed approach through both simulation studies and an application to gene expression levels and metabolite abundances on subjects with varying severity levels of chronic obstructive pulmonary disease. Bayesian inference; Chronic obstructive pulmonary disease (COPD); Data integration; Gaussian graphical model;more »Markov random field prior; Spike and slab prior.« less
  3. Background The novel coronavirus SARS-CoV-2 and its associated disease, COVID-19, have caused worldwide disruption, leading countries to take drastic measures to address the progression of the disease. As SARS-CoV-2 continues to spread, hospitals are struggling to allocate resources to patients who are most at risk. In this context, it has become important to develop models that can accurately predict the severity of infection of hospitalized patients to help guide triage, planning, and resource allocation. Objective The aim of this study was to develop accurate models to predict the mortality of hospitalized patients with COVID-19 using basic demographics and easily obtainable laboratory data. Methods We performed a retrospective study of 375 hospitalized patients with COVID-19 in Wuhan, China. The patients were randomly split into derivation and validation cohorts. Regularized logistic regression and support vector machine classifiers were trained on the derivation cohort, and accuracy metrics (F1 scores) were computed on the validation cohort. Two types of models were developed: the first type used laboratory findings from the entire length of the patient’s hospital stay, and the second type used laboratory findings that were obtained no later than 12 hours after admission. The models were further validated on a multicenter external cohortmore »of 542 patients. Results Of the 375 patients with COVID-19, 174 (46.4%) died of the infection. The study cohort was composed of 224/375 men (59.7%) and 151/375 women (40.3%), with a mean age of 58.83 years (SD 16.46). The models developed using data from throughout the patients’ length of stay demonstrated accuracies as high as 97%, whereas the models with admission laboratory variables possessed accuracies of up to 93%. The latter models predicted patient outcomes an average of 11.5 days in advance. Key variables such as lactate dehydrogenase, high-sensitivity C-reactive protein, and percentage of lymphocytes in the blood were indicated by the models. In line with previous studies, age was also found to be an important variable in predicting mortality. In particular, the mean age of patients who survived COVID-19 infection (50.23 years, SD 15.02) was significantly lower than the mean age of patients who died of the infection (68.75 years, SD 11.83; P<.001). Conclusions Machine learning models can be successfully employed to accurately predict outcomes of patients with COVID-19. Our models achieved high accuracies and could predict outcomes more than one week in advance; this promising result suggests that these models can be highly useful for resource allocation in hospitals.« less
  4. Background Heart failure is a leading cause of mortality and morbidity worldwide. Acute heart failure, broadly defined as rapid onset of new or worsening signs and symptoms of heart failure, often requires hospitalization and admission to the intensive care unit (ICU). This acute condition is highly heterogeneous and less well-understood as compared to chronic heart failure. The ICU, through detailed and continuously monitored patient data, provides an opportunity to retrospectively analyze decompensation and heart failure to evaluate physiological states and patient outcomes. Objective The goal of this study is to examine the prevalence of cardiovascular risk factors among those admitted to ICUs and to evaluate combinations of clinical features that are predictive of decompensation events, such as the onset of acute heart failure, using machine learning techniques. To accomplish this objective, we leveraged tele-ICU data from over 200 hospitals across the United States. Methods We evaluated the feasibility of predicting decompensation soon after ICU admission for 26,534 patients admitted without a history of heart failure with specific heart failure risk factors (ie, coronary artery disease, hypertension, and myocardial infarction) and 96,350 patients admitted without risk factors using remotely monitored laboratory, vital signs, and discrete physiological measurements. Multivariate logistic regression andmore »random forest models were applied to predict decompensation and highlight important features from combinations of model inputs from dissimilar data. Results The most prevalent risk factor in our data set was hypertension, although most patients diagnosed with heart failure were admitted to the ICU without a risk factor. The highest heart failure prediction accuracy was 0.951, and the highest area under the receiver operating characteristic curve was 0.9503 with random forest and combined vital signs, laboratory values, and discrete physiological measurements. Random forest feature importance also highlighted combinations of several discrete physiological features and laboratory measures as most indicative of decompensation. Timeline analysis of aggregate vital signs revealed a point of diminishing returns where additional vital signs data did not continue to improve results. Conclusions Heart failure risk factors are common in tele-ICU data, although most patients that are diagnosed with heart failure later in an ICU stay presented without risk factors making a prediction of decompensation critical. Decompensation was predicted with reasonable accuracy using tele-ICU data, and optimal data extraction for time series vital signs data was identified near a 200-minute window size. Overall, results suggest combinations of laboratory measurements and vital signs are viable for early and continuous prediction of patient decompensation.« less
  5. Abstract Radiogenomics uses machine-learning (ML) to directly connect the morphologic and physiological appearance of tumors on clinical imaging with underlying genomic features. Despite extensive growth in the area of radiogenomics across many cancers, and its potential role in advancing clinical decision making, no published studies have directly addressed uncertainty in these model predictions. We developed a radiogenomics ML model to quantify uncertainty using transductive Gaussian Processes (GP) and a unique dataset of 95 image-localized biopsies with spatially matched MRI from 25 untreated Glioblastoma (GBM) patients. The model generated predictions for regional EGFR amplification status (a common and important target in GBM) to resolve the intratumoral genetic heterogeneity across each individual tumor—a key factor for future personalized therapeutic paradigms. The model used probability distributions for each sample prediction to quantify uncertainty, and used transductive learning to reduce the overall uncertainty. We compared predictive accuracy and uncertainty of the transductive learning GP model against a standard GP model using leave-one-patient-out cross validation. Additionally, we used a separate dataset containing 24 image-localized biopsies from 7 high-grade glioma patients to validate the model. Predictive uncertainty informed the likelihood of achieving an accurate sample prediction. When stratifying predictions based on uncertainty, we observed substantially highermore »performance in the group cohort (75% accuracy, n = 95) and amongst sample predictions with the lowest uncertainty (83% accuracy, n = 72) compared to predictions with higher uncertainty (48% accuracy, n = 23), due largely to data interpolation (rather than extrapolation). On the separate validation set, our model achieved 78% accuracy amongst the sample predictions with lowest uncertainty. We present a novel approach to quantify radiogenomics uncertainty to enhance model performance and clinical interpretability. This should help integrate more reliable radiogenomics models for improved medical decision-making.« less