Added value of an artificial intelligence algorithm in reducing the number of missed incidental acute pulmonary embolism in routine portal venous phase chest CT

OBJECTIVES

The purpose of this study was to evaluate the incremental value of artificial intelligence (AI) compared to the diagnostic accuracy of radiologists alone in detecting incidental acute pulmonary embolism (PE) on routine portal venous contrast-enhanced chest computed tomography (CT).

METHODS

CTs of 3089 consecutive patients referred to the radiology department for a routine contrast-enhanced chest CT between 27-5-2020 and 31-12-2020, were retrospectively analysed by a CE-certified and FDA-approved AI algorithm. The diagnostic performance of the AI was compared to the initial report. To determine the reference standard, discordant findings were independently evaluated by two readers. In case of disagreement, another experienced cardiothoracic radiologist with knowledge of the initial report and the AI output adjudicated.

RESULTS

The prevalence of acute incidental PE in the reference standard was 2.2% (67 of 3089 patients). In 25 cases, AI detected initially unreported PE. This included three cases concerning central/lobar PE. Sensitivity of the AI algorithm was significantly higher than the outcome of the initial report (respectively 95.5% vs. 62.7%, p < 0.001), whereas specificity was very high for both (respectively 99.6% vs 99.9%, p = 0.012). The AI algorithm only showed a slightly higher amount of false-positive findings (11 vs. 2), resulting in a significantly lower PPV (85.3% vs. 95.5%, p = 0.047).

CONCLUSION

The AI algorithm showed high diagnostic accuracy in diagnosing incidental PE, detecting an additional 25 cases of initially unreported PE, accounting for 37.3% of all positive cases.

CLINICAL RELEVANCE STATEMENT

Radiologist support from AI algorithms in daily practice can prevent missed incidental acute PE on routine chest CT, without a high burden of false-positive cases.

KEY POINTS

• Incidental pulmonary embolism is often missed by radiologists in non-diagnostic scans with suboptimal contrast opacification within the pulmonary trunk. • An artificial intelligence algorithm showed higher sensitivity detecting incidental pulmonary embolism on routine portal venous chest CT compared to the initial report. • Implementation of artificial intelligence support in routine daily practice will reduce the number of missed incidental pulmonary embolism.

Retrospective batch analysis to evaluate the diagnostic accuracy of a clinically deployed AI algorithm for the detection of acute pulmonary embolism on CTPA

PURPOSE

To generate and extend the evidence on the clinical validity of an artificial intelligence (AI) algorithm to detect acute pulmonary embolism (PE) on CT pulmonary angiography (CTPA) of patients suspected of PE and to evaluate the possibility of reducing the risk of missed findings in clinical practice with AI-assisted reporting.

METHODS

Consecutive CTPA scan data of 3316 patients referred because of suspected PE between 24-2-2018 and 31-12-2020 were retrospectively analysed by a CE-certified and FDA-approved AI algorithm. The output of the AI was compared with the attending radiologists' report. To define the reference standard, discordant findings were independently evaluated by two readers. In case of disagreement, an experienced cardiothoracic radiologist adjudicated.

RESULTS

According to the reference standard, PE was present in 717 patients (21.6%). PE was missed by the AI in 23 patients, while the attending radiologist missed 60 PE. The AI detected 2 false positives and the attending radiologist 9. The sensitivity for the detection of PE by the AI algorithm was significantly higher compared to the radiology report (96.8% vs. 91.6%, p < 0.001). Specificity of the AI was also significantly higher (99.9% vs. 99.7%, p = 0.035). NPV and PPV of the AI were also significantly higher than the radiology report.

CONCLUSION

The AI algorithm showed a significantly higher diagnostic accuracy for the detection of PE on CTPA compared to the report of the attending radiologist. This finding indicates that missed positive findings could be prevented with the implementation of AI-assisted reporting in daily clinical practice.

CRITICAL RELEVANCE STATEMENT

Missed positive findings on CTPA of patients suspected of pulmonary embolism can be prevented with the implementation of AI-assisted care.

KEY POINTS

The AI algorithm showed excellent diagnostic accuracy detecting PE on CTPA. Accuracy of the AI was significantly higher compared to the attending radiologist. Highest diagnostic accuracy can likely be achieved by radiologists supported by AI. Our results indicate that implementation of AI-assisted reporting could reduce the number of missed positive findings.

The effect of maintenance azithromycin on radiological features in patients with bronchiectasis - Analysis from the BAT randomized controlled trial

RATIONALE

Bronchiectasis (abnormal dilatation of bronchi) is usually diagnosed by high resolution computed tomography (HRCT) and radiological severity has been found to correspond with clinical outcome. A beneficial effect of macrolides maintenance treatment in frequent exacerbating bronchiectasis patients has been established in randomized trials. This study was undertaken to prospectively evaluate the effect of long-term azithromycin (AZM) on radiological features in patients with bronchiectasis.

METHODS

The BAT randomized controlled trial (2008-2010) investigated the effect of 1 year of AZM (250 mg OD) in bronchiectasis with frequent exacerbations. Chest (HR)CT-scans at baseline and after one year of study treatment were obtained and scored by two radiologists according to the Brody - and the Bhalla scoring system.

RESULTS

77 (93%) patients conducted the BAT trial were evaluated in this post-hoc analysis. A significant improvement of the radiological features based on the Brody score was found after one year of AZM therapy as compared to placebo (p = 0.024), with a not significant improvement of the Bhalla score (p=0.071). Especially the consolidation (Bhalla) and parenchymal changes (Brody) sub scores significantly improved (both p=0.030), and even a radiological deterioration was seen on the Brody bronchiectasis sub score for the placebo treated patients (mean 14.5 (11.7) vs.15.7 (11.9)).

CONCLUSIONS

The beneficial effect of long-term AZM treatment on radiological features was demonstrated in this randomized controlled trial. (HR)CT's can be used as an objective measure of treatment response in bronchiectasis.

CLINICAL TRIAL REGISTRATION NUMBER

NCT00415350.

Criteria for Early-Phase Diffuse Idiopathic Skeletal Hyperostosis: Development and Validation

Background Diffuse idiopathic skeletal hyperostosis (DISH) is a condition characterized by the formation of new bone along the anterolateral spinal column at four adjacent vertebral bodies. Purpose To propose and validate criteria for the early phase of DISH by using CT data from two large-scale retrospective cohorts, each with 5-year follow-up. Materials and Methods For this retrospective study, CT data at baseline and follow-up in 1367 patients (cohort I) from 2004 to 2011 were evaluated by two observers to define no DISH, early-stage DISH, and definite DISH on the basis of interval development of consecutive complete or incomplete bone bridges. An independent group of 2267 participants from the COPDGene cohort from 2008 to 2016 was used to validate the early DISH criteria (cohort II). The sensitivity and specificity of early DISH criteria were based on findings in the last CT study as the reference standard by using a nested case-control design. κ Values were calculated between seven readers and with a 3-month interval for one reader. Results Cohort I consisted of 100% men, with a mean age of 60.0 years ± 5.6 (standard deviation) and a mean time between baseline and follow-up CT of 5.0 years ± 1.1. Cohort II consisted of 51% men, with a mean age of 59.9 years ± 8.6 and a mean time between baseline and follow-up CT of 5.4 years ± 0.5. In the derivation cohort, 55 patients comprised the early DISH group. Early DISH was defined as the presence of a spinal segment with a complete bone bridge with an adjacent segment of at least a near-complete bone bridge and another adjacent segment with at least the presence of newly formed bone or when three or more adjacent segments were recorded as showing a near-complete bone bridge. In the validation cohort, sensitivity for early DISH (vs no DISH) was 96% (99 of 103 participants; 95% confidence interval [CI]: 90%, 99%). The corresponding specificity was 83% (1695 of 2034 participants; 95% CI: 82%, 85%). The Fleiss κ for interrater reliability was 0.78 (95% CI: 0.77, 0.78), and the κ for intrarater reliability was 0.89 (95% CI: 0.82, 0.96). Conclusion Early diffuse idiopathic skeletal hyperostosis (DISH) criteria had high sensitivity and specificity for predicting the development of DISH. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Block in this issue.

Feature tracking CMR reveals abnormal strain in preclinical arrhythmogenic right ventricular dysplasia/ cardiomyopathy: a multisoftware feasibility and clinical implementation study

BACKGROUND

Regional right ventricular (RV) dysfunction is the hallmark of Arrhythmogenic Right Ventricular Dysplasia/Cardiomyopathy (ARVD/C), but is currently only qualitatively evaluated in the clinical setting. Feature Tracking Cardiovascular Magnetic Resonance (FT-CMR) is a novel quantitative method that uses cine CMR to calculate strain values. However, most prior FT-CMR studies in ARVD/C have focused on global RV strain using different software methods, complicating implementation of FT-CMR in clinical practice. We aimed to assess the clinical value of global and regional strain using FT-CMR in ARVD/C and to determine differences between commercially available FT-CMR software packages.

METHODS

We analyzed cine CMR images of 110 subjects (39 overt ARVD/C [mutation+/phenotype+], 40 preclinical ARVD/C [mutation+/phenotype-] and 31 control) for global and regional (subtricuspid, anterior, apical) RV strain in the horizontal longitudinal axis using four FT-CMR software methods (Multimodality Tissue Tracking, TomTec, Medis and Circle Cardiovascular Imaging). Intersoftware agreement was assessed using Bland Altman plots.

RESULTS

For global strain, all methods showed reduced strain in overt ARVD/C patients compared to control subjects (p < 0.041), whereas none distinguished preclinical from control subjects (p > 0.275). For regional strain, overt ARVD/C patients showed reduced strain compared to control subjects in all segments which reached statistical significance in the subtricuspid region for all software methods (p < 0.037), in the anterior wall for two methods (p < 0.005) and in the apex for one method (p = 0.012). Preclinical subjects showed abnormal subtricuspid strain compared to control subjects using one of the software methods (p = 0.009). Agreement between software methods for absolute strain values was low (Intraclass Correlation Coefficient = 0.373).

CONCLUSIONS

Despite large intersoftware variability of FT-CMR derived strain values, all four software methods distinguished overt ARVD/C patients from control subjects by both global and subtricuspid strain values. In the subtricuspid region, one software package distinguished preclinical from control subjects, suggesting the potential to identify early ARVD/C prior to overt disease expression.

Small airway imaging phenotypes in biomass- and tobacco smoke-exposed patients with COPD

Chronic obstructive pulmonary disease (COPD) is currently the third most common cause of death, worldwide [1]. Tobacco smoke (TS) is the main risk factor for COPD in developed countries, but biomass smoke (BMS) exposure is the leading cause in developing countries, particularly in women [2].

Biomass and functional small airway diseasehttp://ow.ly/gXu730abpKu

Pulmonary nodule follow-up: be careful with volumetry between contrast enhanced and unenhanced CT

Incident pulmonary nodules are a frequent finding on chest computed tomography (CT) of the lungs requiring follow-up. This case illustrates the importance of taking differences in CT scanning techniques (contrast versus non-contrast enhanced) into account. Comparing nodule size on unenhanced follow-up CT's with initial contrast-enhanced CT may consequently underestimate growth and mask malignant growth rates as demonstrated by our case report.

Accuracy of CT Pulmonary Artery Diameter for Pulmonary Hypertension in End-Stage COPD

Introduction

Pulmonary hypertension (PH) in COPD is associated with a higher mortality and an increased risk on exacerbations compared to COPD patients without PH. The aim was to evaluate the diagnostic value of pulmonary artery (PA) measurements on chest computed tomography (CT) for PH in end-stage COPD.

Methods

COPD patients evaluated for eligibility for lung transplantation between 2004 and 2015 were retrospectively analyzed. Clinical characteristics, chest CTs, spirometry, and right-sided heart catheterizations (RHC) were studied. Diameters of PA and ascending aorta (A) were measured on CT. Diagnostic properties of different cut-offs of PA diameter and PA:A ratio in diagnosing PH were calculated.

Results

Of 92 included COPD patients, 30 (32.6 %) had PH at RHC (meanPAP > 25 mm Hg). PA:A > 1 had a negative predictive value (NPV) of 77.9 % and a positive predictive value (PPV) of 63.1 % with an odds ratio (OR (CI 95 %)) of 5.60 (2.00–15.63). PA diameter ≥30 mm had a NPV of 78 % and PPV of 64 % with an OR (CI 95 %) of 6.95 (2.51–19.24).

Conclusion

A small PA diameter and PA:A make the presence of PH unlikely but cannot exclude its presence in end-stage COPD. A large PA diameter and PA:A maybe used to detect PH early.

Landmark papers in respiratory medicine: Automatic quantification of emphysema and airways disease on computed tomography

Computed tomography (CT) of the lungs is able to visualise the lungs with submillimetre resolution and has become the reference standard for emphysema assessment in vivo. CT, a simple densitometer, can aid automatic localisation and quantification of the extent of both emphysema and airway disease, nowadays with a radiation dose of below 1 millisievert for an average 70 kg adult. In this article we will give a brief overview of the three landmark studies describing the quantification of emphysema and small and large ­airways disease on CT.

Landmark studies on automatic CT quantification of the pathophysiological factors in obstructive pulmonary diseaseshttp://ow.ly/YEKhv

Association of chronic obstructive pulmonary disease and smoking status with bone density and vertebral fractures in male lung cancer screening participants

We studied the vertebral fracture prevalence on low-dose chest computed tomography (CT) in male lung cancer screening participants and the association of fractures and bone density with chronic obstructive pulmonary disease (COPD) and smoking. 1140 male current and former smokers with ≥ 16.5 packyears from the NELSON lung cancer screening trial were included. Age, body mass index, and smoking status were registered. CT scans and pulmonary function tests were obtained on the same day. On CT, vertebral fractures and bone density were measured. The cohort had a mean age of 62.5 years (standard deviation 5.2) old; 531 (46.6%) had quit smoking; and 437 (38.3%) had COPD. Of the group, 100 (8.8%) participants had a vertebral fracture. Fracture prevalence was higher in current compared to former smokers (11.3% versus 5.8%, p = 0.001), but similar in participants with COPD compared to those without (9.6% versus 8.3%, p = 0.430). The multivariable adjusted odds ratio for fracture presence was 1.79 (95% CI: 1.13-2.84) in current smokers and 1.08 (95% CI: 0.69-1.67) in COPD participants. Bone density was lower in current compared to former smokers (103.2 HU versus 108.7 HU, p = 0.006) and in participants with COPD compared to those without [100.7 Hounsfield Units (HU) versus 108.9 HU, p < 0.001]. In multivariate analysis, smoking status and COPD status were independently associated with bone density, corrected for age and body mass index. In conclusion, our study shows that lung cancer screening participants have a substantial vertebral fracture burden. Fractures are more common in current smokers, who also have lower bone density. We could not confirm that COPD is independently associated with vertebral fractures.

Diagnosis of chronic obstructive pulmonary disease in lung cancer screening Computed Tomography scans: independent contribution of emphysema, air trapping and bronchial wall thickening

Background

Beyond lung cancer, screening CT contains additional information on other smoking related diseases (e.g. chronic obstructive pulmonary disease, COPD). Since pulmonary function testing is not regularly incorporated in lung cancer screening, imaging biomarkers for COPD are likely to provide important surrogate measures for disease evaluation. Therefore, this study aims to determine the independent diagnostic value of CT emphysema, CT air trapping and CT bronchial wall thickness for COPD in low-dose screening CT scans.

Methods

Prebronchodilator spirometry and volumetric inspiratory and expiratory chest CT were obtained on the same day in 1140 male lung cancer screening participants. Emphysema, air trapping and bronchial wall thickness were automatically quantified in the CT scans. Logistic regression analysis was performed to derivate a model to diagnose COPD. The model was internally validated using bootstrapping techniques.

Results

Each of the three CT biomarkers independently contributed diagnostic value for COPD, additional to age, body mass index, smoking history and smoking status. The diagnostic model that included all three CT biomarkers had a sensitivity and specificity of 73.2% and 88.%, respectively. The positive and negative predictive value were 80.2% and 84.2%, respectively. Of all participants, 82.8% was assigned the correct status. The C-statistic was 0.87, and the Net Reclassification Index compared to a model without any CT biomarkers was 44.4%. However, the added value of the expiratory CT data was limited, with an increase in Net Reclassification Index of 4.5% compared to a model with only inspiratory CT data.

Conclusion

Quantitatively assessed CT emphysema, air trapping and bronchial wall thickness each contain independent diagnostic information for COPD, and these imaging biomarkers might prove useful in the absence of lung function testing and may influence lung cancer screening strategy. Inspiratory CT biomarkers alone may be sufficient to identify patients with COPD in lung cancer screening setting.

Low-dose CT measurements of airway dimensions and emphysema associated with airflow limitation in heavy smokers: a cross sectional study

Background

Increased airway wall thickness (AWT) and parenchymal lung destruction both contribute to airflow limitation. Advances in computed tomography (CT) post-processing imaging allow to quantify these features. The aim of this Dutch population study is to assess the relationships between AWT, lung function, emphysema and respiratory symptoms.

Methods

AWT and emphysema were assessed by low-dose CT in 500 male heavy smokers, randomly selected from a lung cancer screening population. AWT was measured in each lung lobe in cross-sectionally reformatted images with an automated imaging program at locations with an internal diameter of 3.5 mm, and validated in smaller cohorts of patients. The 15^th percentile method (Perc15) was used to assess the severity of emphysema. Information about respiratory symptoms and smoking behavior was collected by questionnaires and lung function by spirometry.

Results

Median AWT in airways with an internal diameter of 3.5 mm (AWT_3.5) was 0.57 (0.44 - 0.74) mm. Median AWT in subjects without symptoms was 0.52 (0.41-0.66) and in those with dyspnea and/or wheezing 0.65 (0.52-0.81) mm (p<0.001). In the multivariate analysis only AWT_3.5 and emphysema independently explained 31.1%and 9.5%of the variance in FEV₁%predicted, respectively, after adjustment for smoking behavior.

Conclusions

Post processing standardization of airway wall measurements provides a reliable and useful method to assess airway wall thickness. Increased airway wall thickness contributes more to airflow limitation than emphysema in a smoking male population even after adjustment for smoking behavior.

Genome-wide study identifies two loci associated with lung function decline in mild to moderate COPD

Accelerated lung function decline is a key COPD phenotype; however, its genetic control remains largely unknown. We performed a genome-wide association study using the Illumina Human660W-Quad v.1_A BeadChip. Generalized estimation equations were used to assess genetic contributions to lung function decline over a 5-year period in 4,048 European American Lung Health Study participants with largely mild COPD. Genotype imputation was performed using reference HapMap II data. To validate regions meeting genome-wide significance, replication of top SNPs was attempted in independent cohorts. Three genes (TMEM26, ANK3 and FOXA1) within the regions of interest were selected for tissue expression studies using immunohistochemistry. Two intergenic SNPs (rs10761570, rs7911302) on chromosome 10 and one SNP on chromosome 14 (rs177852) met genome-wide significance after Bonferroni. Further support for the chromosome 10 region was obtained by imputation, the most significantly associated imputed SNPs (rs10761571, rs7896712) being flanked by observed markers rs10761570 and rs7911302. Results were not replicated in four general population cohorts or a smaller cohort of subjects with moderate to severe COPD; however, we show novel expression of genes near regions of significantly associated SNPS, including TMEM26 and FOXA1 in airway epithelium and lung parenchyma, and ANK3 in alveolar macrophages. Levels of expression were associated with lung function and COPD status. We identified two novel regions associated with lung function decline in mild COPD. Genes within these regions were expressed in relevant lung cells and their expression related to airflow limitation suggesting they may represent novel candidate genes for COPD susceptibility.