First- and second-hand exposure to smoke or air pollutants is the primary cause of chronic obstructive pulmonary disease (COPD) pathogenesis, where genetic and age-related factors predispose the subject to the initiation and progression of obstructive lung disease. Briefly, airway inflammation, specifically bronchitis, initiates the lung disease, leading to difficulty in breathing (dyspnea) and coughing as initial symptoms, followed by air trapping and inhibition of the flow of air into the lungs due to damage to the alveoli (emphysema). In addition, mucus obstruction and impaired lung clearance mechanisms lead to recurring acute exacerbations causing progressive decline in lung function, eventually requiring lung transplant and other lifesaving interventions to prevent mortality. It is noteworthy that COPD is much more common in the population than currently diagnosed, as only 16 million adult Americans were reported to be diagnosed with COPD as of 2018, although an additional 14 million American adults were estimated to be suffering from COPD but undiagnosed by the current standard of care (SOC) diagnostic, namely the spirometry-based pulmonary function test (PFT). Thus, the main issue driving the adverse disease outcome and significant mortality for COPD is lack of timely diagnosis in the early stages of the disease. The current treatment regime for COPD emphysema is most effective when implemented early, on COPD onset, where alleviating symptoms and exacerbations with timely intervention(s) can prevent steep lung function decline(s) and disease progression to severe emphysema. Therefore, the key to efficiently combatting COPD relies on early detection. Thus, it is important to detect early regional pulmonary function and structural changes to monitor modest disease progression for implementing timely interventions and effectively eliminating emphysema progression. Currently, COPD diagnosis involves using techniques such as COPD screening questionnaires, PFT, arterial blood gas analysis, and/or lung imaging, but these modalities are limited in their capability for early diagnosis and real-time disease monitoring of regional lung function changes. Hence, promising emerging techniques, such as X-ray phase contrast, photoacoustic tomography, ultrasound computed tomography, electrical impedance tomography, the forced oscillation technique, and the impulse oscillometry system powered by robust artificial intelligence and machine learning analysis capability are emerging as novel solutions for early detection and real time monitoring of COPD progression for timely intervention. We discuss here the scope, risks, and limitations of current SOC and emerging COPD diagnostics, with perspective on novel diagnostics providing real time regional lung function monitoring, and predicting exacerbation and/or disease onset for prognosis-based timely intervention(s) to limit COPD–emphysema progression.
more »
« less
Fractional Dynamics Foster Deep Learning of COPD Stage Prediction
Chronic obstructive pulmonary disease (COPD) is one of the leading causes of death worldwide. Current COPD diagnosis (i.e., spirometry) could be unreliable because the test depends on an adequate effort from the tester and testee. Moreover, the early diagnosis of COPD is challenging. The authors address COPD detection by constructing two novel physiological signals datasets (4432 records from 54 patients in the WestRo COPD dataset and 13824 medical records from 534 patients in the WestRo Porti COPD dataset). The authors demonstrate their complex coupled fractal dynamical characteristics and perform a fractional‐order dynamics deep learning analysis to diagnose COPD. The authors found that the fractional‐order dynamical modeling can extract distinguishing signatures from the physiological signals across patients with all COPD stages—from stage 0 (healthy) to stage 4 (very severe). They use the fractional signatures to develop and train a deep neural network that predicts COPD stages based on the input features (such as thorax breathing effort, respiratory rate, or oxygen saturation). The authors show that the fractional dynamic deep learning model (FDDLM) achieves a COPD prediction accuracy of 98.66% and can serve as a robust alternative to spirometry. The FDDLM also has high accuracy when validated on a dataset with different physiological signals.
more » « less- NSF-PAR ID:
- 10397819
- Publisher / Repository:
- Wiley Blackwell (John Wiley & Sons)
- Date Published:
- Journal Name:
- Advanced Science
- Volume:
- 10
- Issue:
- 12
- ISSN:
- 2198-3844
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)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 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.more » « less
-
null (Ed.)Respiratory diseases, like Asthma, COPD, have been a significant public health challenge over decades. Portable spirometers are effective in continuous monitoring of respiratory syndromes out-of-clinic. However, existing systems are either costly or provide limited information and require extra hardware. In this paper, we present mmFlow, a low-barrier means to perform at-home spirometry tests using 5G smart devices. mmFlow works like regular spirometers, where a user forcibly exhales onto a device; but instead of relying on special-purpose hardware, mmFlow leverages built-in millimeter-wave technology in general-purpose, ubiquitous mobile devices. mmFlow analyzes the tiny vibrations created by the airflow on the device surface and combines wireless signal processing with deep learning to enable a software-only spirometry solution. From empirical evaluations, we find that, when device distance is fixed, mmFlow can predict the spirometry indicators with performance comparable to inclinic spirometers with <5% prediction errors. Besides, mmFlow generalizes well under different environments and human conditions, making it promising for out-of-clinic daily monitoring.more » « less
-
more » « less
Abstract Background Idiopathic pulmonary fibrosis (IPF) is a progressive, irreversible, and usually fatal lung disease of unknown reasons, generally affecting the elderly population. Early diagnosis of IPF is crucial for triaging patients’ treatment planning into anti‐fibrotic treatment or treatments for other causes of pulmonary fibrosis. However, current IPF diagnosis workflow is complicated and time‐consuming, which involves collaborative efforts from radiologists, pathologists, and clinicians and it is largely subject to inter‐observer variability.
Purpose The purpose of this work is to develop a deep learning‐based automated system that can diagnose subjects with IPF among subjects with interstitial lung disease (ILD) using an axial chest computed tomography (CT) scan. This work can potentially enable timely diagnosis decisions and reduce inter‐observer variability.
Methods Our dataset contains CT scans from 349 IPF patients and 529 non‐IPF ILD patients. We used 80% of the dataset for training and validation purposes and 20% as the holdout test set. We proposed a two‐stage model: at stage one, we built a multi‐scale, domain knowledge‐guided attention model (MSGA) that encouraged the model to focus on specific areas of interest to enhance model explainability, including both high‐ and medium‐resolution attentions; at stage two, we collected the output from MSGA and constructed a random forest (RF) classifier for patient‐level diagnosis, to further boost model accuracy. RF classifier is utilized as a final decision stage since it is interpretable, computationally fast, and can handle correlated variables. Model utility was examined by (1) accuracy, represented by the area under the receiver operating characteristic curve (AUC) with standard deviation (SD), and (2) explainability, illustrated by the visual examination of the estimated attention maps which showed the important areas for model diagnostics.
Results During the training and validation stage, we observe that when we provide no guidance from domain knowledge, the IPF diagnosis model reaches acceptable performance (AUC±SD = 0.93±0.07), but lacks explainability; when including only guided high‐ or medium‐resolution attention, the learned attention maps are not satisfactory; when including both high‐ and medium‐resolution attention, under certain hyperparameter settings, the model reaches the highest AUC among all experiments (AUC±SD = 0.99±0.01) and the estimated attention maps concentrate on the regions of interests for this task. Three best‐performing hyperparameter selections according to MSGA were applied to the holdout test set and reached comparable model performance to that of the validation set.
Conclusions Our results suggest that, for a task with only scan‐level labels available, MSGA+RF can utilize the population‐level domain knowledge to guide the training of the network, which increases both model accuracy and explainability.
-
more » « less
Abstract A de‐identified data repository of electronic medical record data, i2b2 (Informatics for Integrating Biology and the Bedside), including four geographically diverse academic medical centers, was queried to determine the use of diagnostic spirometry testing in African American children and young adults 5‐34 years of age with sickle cell disease (SCD) with or without a documented history of asthma and/or acute chest syndrome (ACS). A total of 2749 patients were identified with SCD, of these 577 had asthma and 409 had ACS. Cross‐referencing the CPT code for diagnostic spirometry showed that for patients identified as having SCD, a history of ACS, and a diagnosis of asthma, only 31% across all four centers had spirometry. Having an asthma diagnosis was associated with ACS. Among SCD patients with asthma, the proportion with ACS for the four centers was 47%, 75%, 38%, and 36% respectively. The bivariate association between asthma and ACS for each Center was significant for each (
P < .001). To summarize, only one third of patients with co‐morbid SCD, ACS, and asthma received the spirometry procedure as recommended in evidence‐based guidelines, suggesting limited testing for changes in pulmonary function. Future studies to determine barriers and facilitators to implementation of pulmonary testing in SCD are warranted.
