Deep Learning-based Segmentation of Computed Tomography Scans Predicts Disease Progression and Mortality in Idiopathic Pulmonary Fibrosis.

Rationale: Despite evidence demonstrating a prognostic role for computed tomography (CT) scans in idiopathic pulmonary fibrosis (IPF), image-based biomarkers are not routinely used in clinical practice or trials. Objectives: To develop automated imaging biomarkers using deep learning-based segmentation of CT scans. Methods: We developed segmentation processes for four anatomical biomarkers, which were applied to a unique cohort of treatment-naive patients with IPF enrolled in the PROFILE (Prospective Observation of Fibrosis in the Lung Clinical Endpoints) study and tested against a further United Kingdom cohort. The relationships among CT biomarkers, lung function, disease progression, and mortality were assessed. Measurements and Main Results: Data from 446 PROFILE patients were analyzed. Median follow-up duration was 39.1 months (interquartile range, 18.1-66.4 mo), with a cumulative incidence of death of 277 (62.1%) over 5 years. Segmentation was successful on 97.8% of all scans, across multiple imaging vendors, at slice thicknesses of 0.5-5 mm. Of four segmentations, lung volume showed the strongest correlation with FVC (r = 0.82; P < 0.001). Lung, vascular, and fibrosis volumes were consistently associated across cohorts with differential 5-year survival, which persisted after adjustment for baseline gender, age, and physiology score. Lower lung volume (hazard ratio [HR], 0.98 [95% confidence interval (CI), 0.96-0.99]; P = 0.001), increased vascular volume (HR, 1.30 [95% CI, 1.12-1.51]; P = 0.001), and increased fibrosis volume (HR, 1.17 [95% CI, 1.12-1.22]; P < 0.001) were associated with reduced 2-year progression-free survival in the pooled PROFILE cohort. Longitudinally, decreasing lung volume (HR, 3.41 [95% CI, 1.36-8.54]; P = 0.009) and increasing fibrosis volume (HR, 2.23 [95% CI, 1.22-4.08]; P = 0.009) were associated with differential survival. Conclusions: Automated models can rapidly segment IPF CT scans, providing prognostic near and long-term information, which could be used in routine clinical practice or as key trial endpoints.

Apparent diffusion coefficient is highly reproducible on preclinical imaging systems: Evidence from a seven-center multivendor study.

PURPOSE: To evaluate between-site agreement of apparent diffusion coefficient (ADC) measurements in preclinical magnetic resonance imaging (MRI) systems. MATERIALS AND METHODS: A miniaturized thermally stable ice-water phantom was devised. ADC (mean and interquartile range) was measured over several days, on 4.7T, 7T, and 9.4T Bruker, Agilent, and Magnex small-animal MRI systems using a common protocol across seven sites. Day-to-day repeatability was expressed as percent variation of mean ADC between acquisitions. Cross-site reproducibility was expressed as 1.96 × standard deviation of percent deviation of ADC values. RESULTS: ADC measurements were equivalent across all seven sites with a cross-site ADC reproducibility of 6.3%. Mean day-to-day repeatability of ADC measurements was 2.3%, and no site was identified as presenting different measurements than others (analysis of variance [ANOVA] P = 0.02, post-hoc test n.s.). Between-slice ADC variability was negligible and similar between sites (P = 0.15). Mean within-region-of-interest ADC variability was 5.5%, with one site presenting a significantly greater variation than the others (P = 0.0013). CONCLUSION: Absolute ADC values in preclinical studies are comparable between sites and equipment, provided standardized protocols are employed.

Seven Pillars of Small Airways Disease in Asthma and COPD: Supporting Opportunities for Novel Therapies.

Identification of pathologic changes in early and mild obstructive lung disease has shown the importance of the small airways and their contribution to symptoms. Indeed, significant small airways dysfunction has been found prior to any overt airway obstruction being detectable by conventional spirometry techniques. However, most therapies for the treatment of obstructive lung disease target the physiological changes and associated symptoms that result from chronic lung disease, rather than directly targeting the specific underlying causes of airflow disruption or the drivers of disease progression. In addition, although spirometry is the current standard for diagnosis and monitoring of response to therapy, the most widely used measure, FEV1 , does not align with the pathologic changes in early or mild disease and may not align with symptoms or exacerbation frequency in the individual patient. Newer functional and imaging techniques allow more effective assessment of small airways dysfunction; however, significant gaps in our understanding remain. Improving our knowledge of the role of small airways dysfunction in early disease in the airways, along with the identification of novel end points to measure subclinical changes in this region (ie, those not captured as symptoms or identified through standard FEV1), may lead to the development of novel therapies that directly combat early airways disease processes with a view to slowing disease progression and reversing damage. This expert opinion paper discusses small airways disease in the context of asthma and COPD and highlights gaps in current knowledge that impede earlier identification of obstructive lung disease and the development and standardization of novel small airways-specific end points for use in clinical trials.

Quantitative Assessment of Regional Mucociliary Clearance in Smokers with Mild-to-Moderate Chronic Obstructive Pulmonary Disease and Chronic Bronchitis from Planar Radionuclide Imaging.

Background: Mucociliary clearance (MCC) rate from the lung has been shown to be reduced in chronic obstructive pulmonary disease (COPD). This study investigates the value of regional clearance measurements in assessing MCC in mild-to-moderate disease. Methods: Measurement of lung MCC using planar gamma camera imaging was performed in three groups: (i) healthy nonsmoking controls (NSCs) (n = 9), (ii) smoking controls (SCs) who were current smokers with normal lung function (n = 10), and (iii) current smokers with mild-to-moderate COPD and bronchitis (n = 15). The mean (±standard deviation) forced expiratory volumes at 1 second (FEV1) for the three groups were 109 (± 18), 94 (± 5), and 78 (± 12), respectively. After inhalation of a technetium-99m labeled aerosol, planar imaging was performed over 4 hours and then at 24 hours. Both lung clearance and tracheobronchial clearance (TBC) (normalized to 24 hours clearance) were calculated for inner and outer lung zones. Inner zone clearance was corrected for input from the outer zone. A novel parameter, the bronchial airways clearance index (BACI), which combined clearance data from both zones, was also evaluated. Regional results were compared with whole lung clearance in the same subjects. Results: Corrected inner zone clearance at 3 hours was not reduced compared with NSC in either SCs or COPD. Outer zone clearance was higher in COPD than in the other groups. Corrected inner zone TBC showed significant reductions in SC and COPD compared with NSC. BACI was significantly reduced in COPD compared with NSC and also correlated with FEV1. The mean BACI for SC was also reduced compared with NSC, but the distribution of results was bimodal, with a significant proportion of subjects having values in the NSC range. Conclusions: Regional MCC demonstrated differences between NSCs, SCs, and subjects with mild-to-moderate COPD, which were not apparent with whole lung measurements.

In vivo assessments of mucus dynamics in the rat lung using a Gd-Cy5.5-bilabeled contrast agent for magnetic resonance and optical imaging.

Dysfunctions in mucociliary clearance are associated with the accelerated loss of lung function in several respiratory diseases. Approaches enabling the in vivo visualization of mucus dynamics in rodents at high resolution and sensitivity would be beneficial for experimental lung research. We describe the synthesis and characterization of two bilabeled amino dextran-based probes binding specifically to mucin. Labeling of secreted mucus and of mucin in goblet cells in the lungs of lipopolysaccharide-challenged rats has been demonstrated in vivo with near-infrared fluorescence and MRI and confirmed by histology. The effects of uridine triphosphate were then studied in lipopolysaccharide-challenged rats by simultaneously administering the imaging probe and the compound. The data suggest that uridine triphosphate increased the mucociliary clearance, but at the same time induced a release of mucin from goblet cells, thus not contributing to the overall reduction of mucus in the lung. The approach outlined here enables one to derive information on mucus clearance, as well as secretion. Such a global view on mucus dynamics may prove invaluable when testing new pharmacological agents aimed at improving mucociliary clearance.

Quantitative Assessment of Mucociliary Clearance in Smokers with Mild-to-Moderate Chronic Obstructive Pulmonary Disease and Chronic Bronchitis from Planar Radionuclide Imaging Using the Change in Penetration Index.

Background: Mucociliary clearance (MCC) rate from the lung has been shown to be reduced in chronic obstructive pulmonary disease (COPD). This study compared the use of change in penetration index (PI) with conventional whole lung clearance in assessing MCC in mild-to-moderate disease. Methods: Measurement of lung MCC using planar gamma camera imaging was performed in three groups: (1) healthy nonsmoking controls (n = 9), (2) smoking controls who were current smokers with normal lung function (n = 10), and (3) current smokers with mild-to-moderate COPD and bronchitis (n = 15). The mean (±standard deviation) forced expiratory volume at 1 second (FEV1) for the three groups was 109 (±18), 94 (±5), and 78 (±12), respectively. Following inhalation of a technetium-99m labeled aerosol, planar imaging was performed over 4 hours and then at 24 hours. Total lung clearance and tracheobronchial clearance (TBC; normalized to 24-hour clearance) were calculated. A novel parameter, the normalized change in PI (NOCHIP), was also evaluated. PI is the ratio of counts between outer and inner lung zones normalized to lung volume. Results: More aerosol was deposited in central airways in COPD compared to nonsmoking controls, using 24-hour clearance measurements (p < 0.001). Smoking controls had intermediate values. The optimal endpoint for MCC assessment was chosen to be 3 hours, when intersubject variability was minimal, while preserving a measure of early clearance. There was no statistical difference between the three groups in mean total lung clearance, or TBC, at 3 hours. NOCHIP at 3 hours was reduced significantly, compared to nonsmoking controls, in both smoking controls (p = 0.007) and COPD (p < 0.0001). It also correlated with FEV1 (p = 0.003). A higher proportion of smoking control subjects had NOCHIP values in the nonsmoking control range than in the COPD group. Conclusions: NOCHIP was a more sensitive measure of MCC than whole lung clearance and TBC in mild-to-moderate COPD.

OptiJ: Open-source optical projection tomography of large organ samples.

The three-dimensional imaging of mesoscopic samples with Optical Projection Tomography (OPT) has become a powerful tool for biomedical phenotyping studies. OPT uses visible light to visualize the 3D morphology of large transparent samples. To enable a wider application of OPT, we present OptiJ, a low-cost, fully open-source OPT system capable of imaging large transparent specimens up to 13 mm tall and 8 mm deep with 50 µm resolution. OptiJ is based on off-the-shelf, easy-to-assemble optical components and an ImageJ plugin library for OPT data reconstruction. The software includes novel correction routines for uneven illumination and sample jitter in addition to CPU/GPU accelerated reconstruction for large datasets. We demonstrate the use of OptiJ to image and reconstruct cleared lung lobes from adult mice. We provide a detailed set of instructions to set up and use the OptiJ framework. Our hardware and software design are modular and easy to implement, allowing for further open microscopy developments for imaging large organ samples.

Allergen-induced lung inflammation in actively sensitized mice assessed with MR imaging.

PURPOSE: To demonstrate the feasibility of using proton magnetic resonance (MR) imaging to noninvasively detect extravascular and luminal fluid in a murine model of allergen-induced airway inflammation. MATERIALS AND METHODS: The Basel Veterinary Authority approved this experiment. Actively sensitized female Balb/c mice received ovalbumin or saline and underwent MR imaging (a) once 24 hours after the fourth administration of ovalbumin or saline (n = 25) or (b) several times between and after ovalbumin or saline administrations (n = 22) to determine the volume of fluid signal induced by an allergen. Images were acquired in spontaneously breathing animals, without cardiac or respiratory gating. Signal detected with a gradient-echo sequence was compared with bronchoalveolar lavage (BAL) fluid parameters and with perivascular and peribronchial edema and mucus observed at histologic analysis. RESULTS: Up to 24 hours after the fourth administration of ovalbumin, intense and continuous fluid signals (volume, 40-50 microL) were detected in proximal lung regions. At 72 hours after the fourth administration of ovalbumin, remaining signals (21.1 microL +/- 3.8) had a discontinuous texture. The number of eosinophils in the BAL fluid at 24 and 72 hours and their activation were higher in mice that received ovalbumin than in those that received saline. Histologic analysis revealed edema and secreted mucus in the early phase, whereas only mucus was encountered in the late phase. CONCLUSION: These findings suggest that the main component of the early response was plasma leakage (edema), while the main component of the late response was secreted mucus. With the technique validated, the basis for pharmacologic studies in this murine model of lung inflammation with use of MR imaging as a noninvasive readout was provided.

Modeling Gadoxetate Liver Uptake and Efflux Using Dynamic Contrast-Enhanced Magnetic Resonance Imaging Enables Preclinical Quantification of Transporter Drug-Drug Interactions.

OBJECTIVES: The aim of this study was to model the in vivo transporter-mediated uptake and efflux of the hepatobiliary contrast agent gadoxetate in the liver. The efficacy of the proposed technique was assessed for its ability to provide quantitative insights into drug-drug interactions (DDIs), using rifampicin as inhibitor. MATERIALS AND METHODS: Three groups of C57 mice were scanned twice with a dynamic gadoxetate-enhanced magnetic resonance imaging protocol, using a 3-dimensional spoiled gradient-echo sequence for approximately 72 minutes. Before the second magnetic resonance imaging session, 2 of the groups received a rifampicin dose of 20 (n = 7) or 40 (n = 7) mg/kg, respectively. Data from regions of interest in the liver were analyzed using 2 simplifications of a 2-compartment uptake and efflux model to provide estimates for the gadoxetate uptake rate (ki) into the hepatocytes and its efflux rate (kef) into the bile. Both models were assessed for goodness-of-fit in the group without rifampicin (n = 9), and the appropriate model was selected for assessing the ability to monitor DDIs in vivo. RESULTS: Seven of 9 mice from the group without rifampicin were assessed for model implementation and reproducibility. A simple 3 parameter model (ki, kef, and extracellular space, vecs) adequately described the observed liver concentration time series with mean ki = 0.47 ± 0.11 min and mean kef = 0.039 ± 0.016 min. Visually, the area under the liver concentration time profile was reduced for the groups receiving rifampicin. Furthermore, tracer kinetic modeling demonstrated a significant dose-dependent decrease in the uptake (5.9- and 17.3-fold decrease for 20 mg/kg and 40 mg/kg, respectively) and efflux rates (2.2- and 7.9-fold decrease) compared with the first scan for each group. CONCLUSIONS: This study presents the first in vivo implementation of a 2-compartment uptake and efflux model to monitor DDIs at the transporter-protein level, using the clinically relevant organic anion transporting polypeptide inhibitor rifampicin. The technique has the potential to be a novel alternative to other methods, allowing real-time changes in transporter DDIs to be measured directly in vivo.