External validation, radiological evaluation, and development of deep learning automatic lung segmentation in contrast-enhanced chest CT.

OBJECTIVES: There is a need for CT pulmonary angiography (CTPA) lung segmentation models. Clinical translation requires radiological evaluation of model outputs, understanding of limitations, and identification of failure points. This multicentre study aims to develop an accurate CTPA lung segmentation model, with evaluation of outputs in two diverse patient cohorts with pulmonary hypertension (PH) and interstitial lung disease (ILD). METHODS: This retrospective study develops an nnU-Net-based segmentation model using data from two specialist centres (UK and USA). Model was trained (n = 37), tested (n = 12), and clinically evaluated (n = 176) on a diverse 'real-world' cohort of 225 PH patients with volumetric CTPAs. Dice score coefficient (DSC) and normalised surface distance (NSD) were used for testing. Clinical evaluation of outputs was performed by two radiologists who assessed clinical significance of errors. External validation was performed on heterogenous contrast and non-contrast scans from 28 ILD patients. RESULTS: A total of 225 PH and 28 ILD patients with diverse demographic and clinical characteristics were evaluated. Mean accuracy, DSC, and NSD scores were 0.998 (95% CI 0.9976, 0.9989), 0.990 (0.9840, 0.9962), and 0.983 (0.9686, 0.9972) respectively. There were no segmentation failures. On radiological review, 82% and 71% of internal and external cases respectively had no errors. Eighteen percent and 25% respectively had clinically insignificant errors. Peripheral atelectasis and consolidation were common causes for suboptimal segmentation. One external case (0.5%) with patulous oesophagus had a clinically significant error. CONCLUSION: State-of-the-art CTPA lung segmentation model provides accurate outputs with minimal clinical errors on evaluation across two diverse cohorts with PH and ILD. CLINICAL RELEVANCE: Clinical translation of artificial intelligence models requires radiological review and understanding of model limitations. This study develops an externally validated state-of-the-art model with robust radiological review. Intended clinical use is in techniques such as lung volume or parenchymal disease quantification. KEY POINTS: • Accurate, externally validated CT pulmonary angiography (CTPA) lung segmentation model tested in two large heterogeneous clinical cohorts (pulmonary hypertension and interstitial lung disease). • No segmentation failures and robust review of model outputs by radiologists found 1 (0.5%) clinically significant segmentation error. • Intended clinical use of this model is a necessary step in techniques such as lung volume, parenchymal disease quantification, or pulmonary vessel analysis.

Partial anomalous pulmonary venous drainage in patients presenting with suspected pulmonary hypertension: A series of 90 patients from the ASPIRE registry.

BACKGROUND AND OBJECTIVE: There are limited data regarding patients with PAPVD with suspected and diagnosed PH. METHODS: Patients with PAPVD presenting to a large PH referral centre during 2007-2017 were identified from the ASPIRE registry. RESULTS: Ninety patients with PAPVD were identified; this was newly diagnosed at our unit in 71 patients (78%), despite 69% of these having previously undergone CT. Sixty-seven percent had a single right superior and 23% a single left superior anomalous vein. Patients with an SV-ASD had a significantly larger RV area, pulmonary artery and L-R shunt and a higher % predicted DLCO (all P < 0.05). Sixty-five patients were diagnosed with PH (defined as mPAP ≥ 25 mm Hg), which was post-capillary in 24 (37%). No additional causes of PH were identified in 28 patients; 17 of these (26% of those patients with PH) had a PVR > 3 WU. Seven of these patients had isolated PAPVD, five of whom (8% of those patients with PH) had anomalous drainage of a single pulmonary vein. CONCLUSION: Undiagnosed PAPVD with or without ASD may be present in patients with suspected PH; cross-sectional imaging should therefore be specifically assessed whenever this diagnosis is considered. Radiological and physiological markers of L-R shunt are higher in patients with an associated SV-ASD. Although many patients with PAPVD and PH may have other potential causes of PH, a proportion of patients diagnosed with PAH have isolated PAPVD in the absence of other causative conditions.

Mild parenchymal lung disease and/or low diffusion capacity impacts survival and treatment response in patients diagnosed with idiopathic pulmonary arterial hypertension.

There are limited published data defining survival and treatment response in patients with mild lung disease and/or reduced gas transfer who fulfil diagnostic criteria for idiopathic pulmonary arterial hypertension (IPAH).Patients diagnosed with IPAH between 2001 and 2019 were identified in the ASPIRE (Assessing the Spectrum of Pulmonary Hypertension Identified at a Referral Centre) registry. Using prespecified criteria based on computed tomography (CT) imaging and spirometry, patients with a diagnosis of IPAH and no lung disease were termed IPAHno-LD (n=303), and those with minor/mild emphysema or fibrosis were described as IPAHmild-LD (n=190).Survival was significantly better in IPAHno-LD than in IPAHmild-LD (1- and 5-year survival 95% and 70% versus 78% and 22%, respectively; p<0.0001). In the combined group of IPAHno-LD and IPAHmild-LD, independent predictors of higher mortality were increasing age, lower diffusing capacity of the lung for carbon monoxide (D LCO), lower exercise capacity and a diagnosis of IPAHmild-LD (all p<0.05). Exercise capacity and quality of life improved (both p<0.0001) following treatment in patients with IPAHno-LD, but not IPAHmild-LD A proportion of patients with IPAHno-LD had a D LCO <45%; these patients had poorer survival than patients with D LCO ≥45%, although they demonstrated improved exercise capacity following treatment.The presence of even mild parenchymal lung disease in patients who would be classified as IPAH according to current recommendations has a significant adverse effect on outcomes. This phenotype can be identified using lung function testing and clinical CT reports. Patients with IPAH, no lung disease and severely reduced D LCO may represent a further distinct phenotype. These data suggest that randomised controlled trials of targeted therapies in patients with these phenotypes are required.

IodiNe Subtraction mapping in the diagnosis of Pulmonary chronIc thRomboEmbolic disease (INSPIRE): Rationale and methodology of a cross-sectional observational diagnostic study.

Chronic thromboembolic pulmonary hypertension (CTEPH) is a severe but treatable disease that is commonly underdiagnosed. Computed tomography lung subtraction iodine mapping (CT-LSIM) in addition to standard CT pulmonary angiography (CTPA) may improve the evaluation of suspected chronic pulmonary embolism and improve the diagnostic pick up rate. We aim to recruit 100 patients suspected of having CTEPH and perform CT-LSIM scans in addition to the current gold standard test of nuclear medicine test (lung single photon emission computed tomography (SPECT) imaging) as a pilot study which will contribute to and inform the definitive trial. The diagnostic accuracy of CT-LSIM and lung SPECT will be compared. The primary outcome of the full definitive study is non-inferiority of CT-LSIM versus lung SPECT imaging.

Diagnostic and prognostic significance of cardiovascular magnetic resonance native myocardial T1 mapping in patients with pulmonary hypertension.

BACKGROUND: Native T1 may be a sensitive, contrast-free, non-invasive cardiovascular magnetic resonance (CMR) marker of myocardial tissue changes in patients with pulmonary artery hypertension. However, the diagnostic and prognostic value of native T1 mapping in this patient group has not been fully explored. The aim of this work was to determine whether elevation of native T1 in myocardial tissue in pulmonary hypertension: (a) varies according to pulmonary hypertension subtype; (b) has prognostic value and (c) is associated with ventricular function and interaction. METHODS: Data were retrospectively collected from a total of 490 consecutive patients during their clinical 1.5 T CMR assessment at a pulmonary hypertension referral centre in 2015. Three hundred sixty-nine patients had pulmonary hypertension [58 ± 15 years; 66% female], an additional 39 had pulmonary hypertension due to left heart disease [68 ± 13 years; 60% female], 82 patients did not have pulmonary hypertension [55 ± 18; 68% female]. Twenty five healthy subjects were also recruited [58 ±4 years); 51% female]. T1 mapping was performed with a MOdified Look-Locker Inversion Recovery (MOLLI) sequence. T1 prognostic value in patients with pulmonary arterial hypertension was assessed using multivariate Cox proportional hazards regression analysis. RESULTS: Patients with pulmonary artery hypertension had elevated T1 in the right ventricular (RV) insertion point (pulmonary hypertension patients: T1 = 1060 ± 90 ms; No pulmonary hypertension patients: T1 = 1020 ± 80 ms p < 0.001; healthy subjects T1 = 940 ± 50 ms p < 0.001) with no significant difference between the major pulmonary hypertension subtypes. The RV insertion point was the most successful T1 region for discriminating patients with pulmonary hypertension from healthy subjects (area under the curve = 0.863) however it could not accurately discriminate between patients with and without pulmonary hypertension (area under the curve = 0.654). T1 metrics did not contribute to prediction of overall mortality (septal: p = 0.552; RV insertion point: p = 0.688; left ventricular free wall: p = 0.258). Systolic interventricular septal angle was a significant predictor of T1 in patients with pulmonary hypertension (p < 0.001). CONCLUSIONS: Elevated myocardial native T1 was found to a similar extent in pulmonary hypertension patient subgroups and is independently associated with increased interventricular septal angle. Native T1 mapping may not be of additive value in the diagnostic or prognostic evaluation of patients with pulmonary artery hypertension.

CT derived left atrial size identifies left heart disease in suspected pulmonary hypertension: Derivation and validation of predictive thresholds.

BACKGROUND: Patients with pulmonary hypertension due to left heart disease (PH-LHD) have overlapping clinical features with pulmonary arterial hypertension making diagnosis reliant on right heart catheterization (RHC). This study aimed to investigate computed tomography pulmonary angiography (CTPA) derived cardiopulmonary structural metrics, in comparison to magnetic resonance imaging (MRI) for the diagnosis of left heart disease in patients with suspected pulmonary hypertension. METHODS: Patients with suspected pulmonary hypertension who underwent CTPA, MRI and RHC were identified. Measurements of the cardiac chambers and vessels were recorded from CTPA and MRI. The diagnostic thresholds of individual measurements to detect elevated pulmonary arterial wedge pressure (PAWP) were identified in a derivation cohort (n = 235). Individual CT and MRI derived metrics were tested in validation cohort (n = 211). RESULTS: 446 patients, of which 88 had left heart disease. Left atrial area was a strong predictor of elevated PAWP>15 mm Hg and PAWP>18 mm Hg, area under curve (AUC) 0.854, and AUC 0.873 respectively. Similar accuracy was also identified for MRI derived LA volume, AUC 0.852 and AUC 0.878 for PAWP > 15 and 18 mm Hg, respectively. Left atrial area of 26.8 cm2 and 30.0 cm2 were optimal specific thresholds for identification of PAWP > 15 and 18 mm Hg, had sensitivity of 60%/53% and specificity 89%/94%, respectively in a validation cohort. CONCLUSIONS: CTPA and MRI derived left atrial size identifies left heart disease in suspected pulmonary hypertension with high specificity. The proposed diagnostic thresholds for elevated left atrial area on routine CTPA may be a useful to indicate the diagnosis of left heart disease in suspected pulmonary hypertension.

Current and emerging imaging techniques in the diagnosis and assessment of pulmonary hypertension.

INTRODUCTION: Pulmonary hypertension (PH) is a challenging condition to diagnose and treat. Over the last two decades, there have been significant advances in therapeutic approaches and imaging technologies. Current guidelines emphasize the importance of cardiac catheterization; however, the increasing availability of non-invasive imaging has the potential to improve diagnostic rates, whilst providing additional information on patient phenotypes. Areas covered: This review discusses the role of imaging in the diagnosis, prognostic assessment and follow-up of patients with PH. Imaging methods, ranging from established investigations (chest radiography, echocardiography, nuclear medicine and computerized tomography (CT)), to emerging modalities (dual energy CT, magnetic resonance imaging (MRI), optical coherence tomography and positron emission tomography (PET)) are reviewed. The value and limitations of the clinical utility of these imaging modalities and their potential clinical application are reviewed. Expert commentary: Imaging plays a key role in the diagnosis and classification of pulmonary hypertension. It also provides valuable prognostic information and emerging evidence supports a role for serial assessments. The authors anticipate an increasing role for imaging in the pulmonary hypertension clinic. This will reduce the need for invasive investigations, whilst providing valuable insights that will improve our understanding of disease facilitate a more targeted approach to treatment.

Rapid lung volumetry using ultrafast dynamic magnetic resonance imaging during forced vital capacity maneuver: correlation with spirometry.

INTRODUCTION: Dynamic magnetic resonance imaging (MRI) has the potential for rapid noninvasive evaluation of changes in lung volume. The aim of this study was to perform rapid lung volumetry using ultrafast dynamic MRI to capture a forced vital capacity (FVC) maneuver. MATERIALS AND METHODS: Nine healthy volunteers underwent 2-dimensional spoiled gradient echo imaging in coronal and sagittal planes during FVC maneuvers. An elliptical model of the axial cross section of the lungs was used to generate rapid volume-time curves. Spirometric indices were correlated with MR volumetry findings. RESULTS: Total lung volume calculated from static MRI correlated well with the dynamic MR scans (r = 0.83; P < 0.01). Spirometric indices (first second of forced expiration and FVC) calculated from our MR volumetry technique correlated well with conventional spirometry (P < 0.01). CONCLUSION: The technique provides a means of sampling lung volume change during the rapid subsecond movements that take place during a FVC maneuver.