Implementation of artificial intelligence in thoracic imaging-a what, how, and why guide from the European Society of Thoracic Imaging (ESTI).

This statement from the European Society of Thoracic imaging (ESTI) explains and summarises the essentials for understanding and implementing Artificial intelligence (AI) in clinical practice in thoracic radiology departments. This document discusses the current AI scientific evidence in thoracic imaging, its potential clinical utility, implementation and costs, training requirements and validation, its' effect on the training of new radiologists, post-implementation issues, and medico-legal and ethical issues. All these issues have to be addressed and overcome, for AI to become implemented clinically in thoracic radiology. KEY POINTS: • Assessing the datasets used for training and validation of the AI system is essential. • A departmental strategy and business plan which includes continuing quality assurance of AI system and a sustainable financial plan is important for successful implementation. • Awareness of the negative effect on training of new radiologists is vital.

Strategies to safely rule out pulmonary embolism in COVID-19 outpatients: a multicenter retrospective study.

OBJECTIVES: The objective was to define a safe strategy to exclude pulmonary embolism (PE) in COVID-19 outpatients, without performing CT pulmonary angiogram (CTPA). METHODS: COVID-19 outpatients from 15 university hospitals who underwent a CTPA were retrospectively evaluated. D-Dimers, variables of the revised Geneva and Wells scores, as well as laboratory findings and clinical characteristics related to COVID-19 pneumonia, were collected. CTPA reports were reviewed for the presence of PE and the extent of COVID-19 disease. PE rule-out strategies were based solely on D-Dimer tests using different thresholds, the revised Geneva and Wells scores, and a COVID-19 PE prediction model built on our dataset were compared. The area under the receiver operating characteristics curve (AUC), failure rate, and efficiency were calculated. RESULTS: In total, 1369 patients were included of whom 124 were PE positive (9.1%). Failure rate and efficiency of D-Dimer > 500 µg/l were 0.9% (95%CI, 0.2-4.8%) and 10.1% (8.5-11.9%), respectively, increasing to 1.0% (0.2-5.3%) and 16.4% (14.4-18.7%), respectively, for an age-adjusted D-Dimer level. D-dimer > 1000 µg/l led to an unacceptable failure rate to 8.1% (4.4-14.5%). The best performances of the revised Geneva and Wells scores were obtained using the age-adjusted D-Dimer level. They had the same failure rate of 1.0% (0.2-5.3%) for efficiency of 16.8% (14.7-19.1%), and 16.9% (14.8-19.2%) respectively. The developed COVID-19 PE prediction model had an AUC of 0.609 (0.594-0.623) with an efficiency of 20.5% (18.4-22.8%) when its failure was set to 0.8%. CONCLUSIONS: The strategy to safely exclude PE in COVID-19 outpatients should not differ from that used in non-COVID-19 patients. The added value of the COVID-19 PE prediction model is minor. KEY POINTS: • D-dimer level remains the most important predictor of pulmonary embolism in COVID-19 patients. • The AUCs of the revised Geneva and Wells scores using an age-adjusted D-dimer threshold were 0.587 (95%CI, 0.572 to 0.603) and 0.588 (95%CI, 0.572 to 0.603). • The AUC of COVID-19-specific strategy to rule out pulmonary embolism ranged from 0.513 (95%CI: 0.503 to 0.522) to 0.609 (95%CI: 0.594 to 0.623).

Pulmonary hypertension due to left heart disease: diagnostic value of pulmonary artery distensibility.

OBJECTIVES: To evaluate how pulmonary artery (PA) distensibility performs in detecting pulmonary hypertension due to left heart disease (PH-LHD) in comparison with parameters from ungated computed tomography (CT) and echocardiography. METHODS: One hundred patients (79 men, mean age = 63 ± 17 years) with either severe heart failure with reduced ejection fraction (HFrEF), aortic stenosis, or primary mitral regurgitation prospectively underwent right heart catheterization, ungated CT, ECG-gated CT, and echocardiography. During the ECG-gated CT, the right PA distensibility was calculated. In ungated CT, dPA, dPA/AA, the ratio of dPA to the diameter of the vertebra, segmental PA diameter, segmental PA-to-bronchus ratio, and the main PA volume were measured; the egg-and-banana sign was recorded. During echocardiography, the tricuspid regurgitation (TR) gradient was measured. The areas under the ROC curves (AUC) of these signs were computed and compared with DeLong test. Correlation between PA distensibility and PA pressure (PAP) was investigated through Pearson's coefficient. RESULTS: PA distensibility was lower in patients with PH than in those without PH (11.4 vs. 21.2%, p < 0.001) and correlated negatively with mean PAP (r = - 0.72, p < 0.001). Age, PA size, and mean PAP were independent predictors of PA distensibility. PA distensibility < 18% detected PH-LHD with 96% sensitivity and 73% specificity; its AUC was 0.92, larger than that of any other sign at ungated CT and TR gradient (AUC ranging from 0.54 to 0.83, DeLong: p ranging from 0.020 to < 0.001). CONCLUSION: PA distensibility on an ECG-gated CT can detect PH-LHD better than the parameters reflecting PA dilatation in ungated CT or TR gradient in the echocardiography of patients with severe HFrEF, aortic stenosis, or mitral regurgitation. KEY POINTS: • In left heart disease, pulmonary artery distensibility is lower in patients with PH than in those without pulmonary hypertension (11.4 vs. 21.2%, p < 0.001). • In left heart disease, pulmonary artery distensibility detects pulmonary hypertension with an area under the receiver operating curve of 0.92. • In left heart disease, the area under the receiver operating curve of pulmonary artery distensibility for detecting pulmonary hypertension is larger than that of all other signs at ungated CT (p from 0.019 to < 0.001) and tricuspid regurgitation gradient at echocardiography (p = 0.020).

An unusual case of cardiac tamponade: Bronchogenic cyst infection due to Salmonella bredeney.

We present an unusual case of cardiac tamponade in a 17-year-old girl immunocompetent patient due to Salmonella enterica ssp. bredeney following infection of a bronchogenic cyst. The patient was admitted to hospital with pleuritic chest pain, dyspnoea and fever. Pulmonary angio-CT showed a bronchogenic cyst compressing the left atrium. The echocardiography showed diffuse pericardial effusion with right ventricular collapse consistent with cardiac tamponade. Pericardiocentesis was performed and microbiological cultures of the pericardial fluid became positive for Salmonella species confirmed later as bredeney subspecies by PCR. Empirical antibiotherapy was started with intravenous (IV) ceftriaxone. Bronchogenic cyst infection was suspected and confirmed by 18FDG PET CT. The patient was successfully treated by complete resection of the cyst and continuation of IV ceftriaxone followed by oral amoxicillin/clavulanate for a total duration of 6 weeks. She then completely recovered and didn't present any relapse after 6 months of follow up.

Pulmonary hypertension due to left heart disease: diagnostic and prognostic value of CT in chronic systolic heart failure.

OBJECTIVES: To evaluate the ability of chest computed tomography (CT) to predict pulmonary hypertension (PH) and outcome in chronic heart failure with reduced ejection fraction (HFrEF). METHODS: We reviewed 119 consecutive patients with HFrEF by CT, transthoracic echocardiography (TTE) and right heart catheterization (RHC). CT-derived pulmonary artery (PA) diameter and PA to ascending aorta diameter ratio (PA:A ratio), left atrial, right atrial, right ventricular (RV) and left ventricular volumes were correlated with RHC mean pulmonary arterial pressure (mPAP) . Diagnostic accuracy to predict PH and ability to predict primary composite endpoint of all-cause mortality and HF events were evaluated. RESULTS: RV volume was significantly higher in 81 patients with PH compared to 38 patients without PH (133 ml/m2 vs. 79 ml/m2, p < 0.001) and was moderately correlated with mPAP (r=0.55, p < 0.001). Also, RV volume had higher ability to predict PH (area under the curve: 0.88) than PA diameter (0.79), PA:A ratio (0.76) by CT and tricuspid regurgitation gradient (0.83) and RV basal diameter by TTE (0.84, all p < 0.001). During the follow-up period (median: 3.4 years), 51 patients (43%) had HF events or died. After correction for important clinical, TTE and RHC parameters, RV volume (adjusted hazard ratio [HR]: 1.71, 95% CI 1.31-2.23, p < 0.001) and PA diameter (HR: 1.61, 95% CI 1.18-2.22, p = 0.003) were independent predictors of the primary endpoint. CONCLUSION: In patients with HFrEF, measurement of RV volume and PA diameter on ungated CT are non-invasive markers of PH and may help to predict the patient outcome. KEY POINTS: • Right ventricular (RV) volume measured by chest CT has good ability to identify pulmonary hypertension (PH) in patients with chronic heart failure (HF) and reduced ejection fraction (HFrEF). • The accuracy of pulmonary artery (PA) diameter and PA to ascending aorta diameter ratio (PA:A ratio) to predict PH was similar to previous studies, however, with lower cut-offs (28.1 mm and 0.92, respectively). • Chest CT-derived PA diameter and RV volume independently predict all-cause mortality and HF events and improve outcome prediction in patients with advanced HFrEF.

An individualized radiation dose escalation trial in non-small cell lung cancer based on FDG-PET imaging.

AIM: The aim of the study was to assess the feasibility of an individualized 18F fluorodeoxyglucose positron emission tomography (FDG-PET)-guided dose escalation boost in non-small cell lung cancer (NSCLC) patients and to assess its impact on local tumor control and toxicity. PATIENTS AND METHODS: A total of 13 patients with stage II-III NSCLC were enrolled to receive a dose of 62.5 Gy in 25 fractions to the CT-based planning target volume (PTV; primary turmor and affected lymph nodes). The fraction dose was increased within the individual PET-based PTV (PTVPET) using intensity modulated radiotherapy (IMRT) with a simultaneous integrated boost (SIB) until the predefined organ-at-risk (OAR) threshold was reached. Tumor response was assessed during follow-up by means of repeat FDG-PET/computed tomography. Acute and late toxicity were recorded and classified according to the CTCAE criteria (Version 4.0). Local progression-free survival was determined using the Kaplan-Meier method. RESULTS: The average dose to PTVPET reached 89.17 Gy for peripheral and 75 Gy for central tumors. After a median follow-up period of 29 months, seven patients were still alive, while six had died (four due to distant progression, two due to grade 5 toxicity). Local progression was seen in two patients in association with further recurrences. One and 2-year local progression free survival rates were 76.9% and 52.8%, respectively. Three cases of acute grade 3 esophagitis were seen. Two patients with central tumors developed late toxicity and died due to severe hemoptysis. CONCLUSION: These results suggest that a non-uniform and individualized dose escalation based on FDG-PET in IMRT delivery is feasible. The doses reached were higher in patients with peripheral compared to central tumors. This strategy enables good local control to be achieved at acceptable toxicity rates. However, dose escalation in centrally located tumors with direct invasion of mediastinal organs must be performed with great caution in order to avoid severe late toxicity.

Severe COPD exacerbation: CT features.

UNLABELLED: Abstract Objective: To describe CT features associated with severe exacerbations of Chronic Obstructive Pulmonary Disease (COPD). MATERIALS AND METHODS: In this prospective ethical-committee-approved study, 44 COPD patients (34 men, 10 women, age range 49-83 years) who provided written informed consent were included at the time of hospital admission for severe exacerbation. Pulmonary function tests (PFT) and chest CT scans were performed at admission and after resolution of the episode following a minimum of 4 weeks free of any acute symptom. For each CT scan, two radiologists independently scored 15 features in each lobe and side. CT features and PFT results were compared for exacerbation and control through Mac-Nemar tests and paired t-tests, respectively. RESULTS: Forced expiratory volume in 1 second and vital capacity improved significantly after exacerbation (p = 0.023 and 0.012, respectively). Bronchial wall thickening and lymphadenopathy were graded significantly higher at exacerbation than at control by both readers (p ranging from < 0.001 to 0.028). Other CT features were not observed during exacerbation, or were so only by one reader (p ranging from < 0.001 to 0.928). CONCLUSION: Only lymphadenopathy and bronchial wall thickening are CT features associated with severe COPD exacerbation, respectively in 25% and 50% of patients. Our findings do not advocate a role for CT in the routine work-up of patients with severe COPD exacerbation.

Unusual vanishing interstitial lymphatic "pearls" in a patient presenting with extensive interstitial and mediastinal MDCT features of acute cardiogenic failure related to bradycardia and mitral regurgitation.

Thoracic multidetector computed tomography-MDCT-was simultaneously performed during emergency abdominal CT in a patient presenting with abdominal pain and acute cardiogenic edema related to sick sinus syndrome and mitral prolapse with regurgitation. A constellation of severe but completely reversible interstitial and mediastinal features was found comprising pleural effusions, diffuse alveolar ground glass, thickening of the bronchial walls and septal lines, hazy infiltration of the mediastinal fat, and enlarged lymphatic nodes. Multiple atypical hypodense nodular "pearls" were also found. These oval shape or fusiform pearls were distributed along the thickened septal lines and disappeared completely after treatment. The hypothesis of transient lymphatic ectasia or lakes is proposed for these never previously described abnormalities.

Radiomics and Delta-Radiomics Signatures to Predict Response and Survival in Patients with Non-Small-Cell Lung Cancer Treated with Immune Checkpoint Inhibitors.

The aim of our study was to determine the potential role of CT-based radiomics in predicting treatment response and survival in patients with advanced NSCLC treated with immune checkpoint inhibitors. We retrospectively included 188 patients with NSCLC treated with PD-1/PD-L1 inhibitors from two independent centers. Radiomics analysis was performed on pre-treatment contrast-enhanced CT. A delta-radiomics analysis was also conducted on a subset of 160 patients who underwent a follow-up contrast-enhanced CT after 2 to 4 treatment cycles. Linear and random forest (RF) models were tested to predict response at 6 months and overall survival. Models based on clinical parameters only and combined clinical and radiomics models were also tested and compared to the radiomics and delta-radiomics models. The RF delta-radiomics model showed the best performance for response prediction with an AUC of 0.8 (95% CI: 0.65-0.95) on the external test dataset. The Cox regression delta-radiomics model was the most accurate at predicting survival with a concordance index of 0.68 (95% CI: 0.56-0.80) (p = 0.02). The baseline CT radiomics signatures did not show any significant results for treatment response prediction or survival. In conclusion, our results demonstrated the ability of a CT-based delta-radiomics signature to identify early on patients with NSCLC who were more likely to benefit from immunotherapy.

CT of blunt diaphragmatic rupture.

The diagnosis of blunt diaphragmatic rupture (BDR) is difficult and often missed, leaving many patients with this traumatic injury at risk for life-threatening complications. The potential diagnostic pitfalls are numerous and include anatomic variants and congenital and acquired abnormalities. Chest radiography, despite its known limitations, may still be helpful in the early assessment of severe thoracoabdominal trauma and for detecting initially overlooked BDR or late complications of BDR. However, since the development of helical and multidetector scanners, computed tomography (CT) has become the reference standard; thus, knowledge of the CT signs suggestive of BDR is important for recognition of this injury pattern. A large number of CT signs of BDR have been described elsewhere, many of them individually, but the use of various appellations for the same sign can make previously published reports confusing. The systematic description and classification of CT signs provided in this article may help clarify matters and provide clues for diagnosing BDR. The authors describe 19 distinct CT signs grouped in three categories: direct signs of rupture, indirect signs that are consequences of rupture, and signs that are of uncertain origin. Since no single CT sign can be considered a marker leading to a correct diagnosis in every case of BDR, accurate diagnosis depends on the analysis of all signs present.

Platypnea-orthodeoxia syndrome in a postoperative patient: a case report.

BACKGROUND: We report a case of platypnea-orthodeoxia syndrome observed in a complex clinical situation associating a bilateral pleural effusion, lobar pulmonary embolism, and a partial anomalous pulmonary venous return. CASE PRESENTATION: A 57-year-old Caucasian woman developed acute dyspnea in the postoperative course of an elective gynecological surgery for advanced stage ovarian cancer. Preoperative evaluation had failed to reveal any respiratory or cardiac problem. After evidence of a low arterial oxygen saturation, blood gas analysis from the central venous line correctly inserted in the right internal jugular vein revealed a higher oxygen saturation than in the arterial compartment. A thoracic computed tomography showed bilateral pleural effusion, lobar pulmonary embolism, and a drainage of a left pulmonary vein into the left innominate vein. This unique combination resulted in an uncommon cause of platypnea-orthodeoxia syndrome. CONCLUSION: Often associated with right-to-left shunting, platypnea-orthodeoxia syndrome may be observed in complex clinical conditions with several factors influencing the ventilation/perfusion ratio. The paradoxical finding of a higher oxygen saturation in a central venous line than in an arterial line should prompt the clinician to look at the possibility of partial anomalous pulmonary venous return. No specific treatment is required in asymptomatic adults, except for an echocardiographic follow-up to detect the onset of pulmonary hypertension.

Tumor response to EGFR/BRAF/MEK co-inhibition in a patient with EGFR mutated lung adenocarcinoma developing a BRAFV600 mutation as an acquired resistance mechanism.

EGFR-mutant adenocarcinomas represent 12% of unselected lung adenocarcinomas and 44% of never smoker adenocarcinomas in the Caucasian population. Activating mutations like exon19 deletion or exon21 Leu858Arg point mutations are predictive of tumor response to EGFR tyrosine kinase inhibitors. Unfortunately, acquired resistance inevitably occurs by the development of novel EGFR mutations, mutations in other genes, gene amplification, gene fusion or tumor transformation. The management of tumors presenting multiple targetable mutations is unclear. We present the case of a patient developing a BRAFV600 mutation as acquired resistance mechanism to osimertinib, who responded favorably to the combination of dabrafenib, trametinib and osimertinib.

Pulmonary tumor thrombotic microangiopathy in a patient with a metastatic urothelial carcinoma.

A 78-year-old woman was admitted for acute dyspnoea. One year before, she had been treated with cisplatin and gemcitabine for a high grade urothelial carcinoma. Immunotherapy was discussed 9 months later due the progression of bone metastases but could not be administered before this episode of respiratory distress. There was a major discrepancy between the findings of a limited pulmonary embolism at thoracic tomodensitometry and the severity of a recently developed pulmonary hypertension at echocardiography. The patient presented cardiac arrest on day 6 and post-mortem findings were consistent with diffuse pulmonary tumor thrombotic microangiopathy, a rare complication of urothelial carcinoma.

Real-life prevalence of progressive fibrosing interstitial lung diseases.

The concept of progressive fibrosing interstitial lung disease (PF-ILD) has recently emerged. However, real-life proportion of PF-ILDs outside IPF is still hard to evaluate. Therefore, we sought to estimate the proportion of PF-ILD in our ILD cohort. We also determined the proportion of ILD subtypes within PF-ILD and investigated factors associated with PF-ILDs. Finally, we quantified interobserver agreement between radiologists for the assessment of fibrosis. We reviewed the files of ILD patients discussed in multidisciplinary discussion between January 1st 2017 and December 31st 2019. Clinical data, pulmonary function tests (PFTs) and high-resolution computed tomography (HRCTs) were centrally reviewed. Fibrosis was defined as the presence of traction bronchiectasis, reticulations with/out honeycombing. Progression was defined as a relative forced vital capacity (FVC) decline of ≥ 10% in ≤ 24 months or 5% < FVC decline < 10% and progression of fibrosis on HRCT in ≤ 24 months. 464 consecutive ILD patients were included. 105 had a diagnosis of IPF (23%). Most frequent non-IPF ILD were connective tissue disease (CTD)-associated ILD (22%), hypersensitivity pneumonitis (13%), unclassifiable ILD (10%) and sarcoidosis (8%). Features of fibrosis were common (82% of CTD-ILD, 81% of HP, 95% of uILD). After review of HRCTs and PFTs, 68 patients (19% of non-IPF ILD) had a PF-ILD according to our criteria. Interobserver agreement for fibrosis between radiologists was excellent (Cohen's kappa 0.86). The main diagnosis among PF-ILD were CTD-ILD (36%), HP (22%) and uILD (20%). PF-ILD patients were significantly older than non-F-ILD (P = 0.0005). PF-ILDs represent about 20% of ILDs outside IPF. This provides an estimation of the proportion of patients who might benefit from antifibrotics. Interobserver agreement between radiologists for the diagnosis of fibrotic ILD is excellent.

Integrative respiratory follow-up of severe COVID-19 reveals common functional and lung imaging sequelae.

BACKGROUND: COVID-19 pandemic resulted in an unprecedented number of hospitalizations in general wards and intensive care units (ICU). Severe and critical COVID-19 patients suffer from extensive pneumonia; therefore, long-term respiratory sequelae may be expected. RESEARCH QUESTION: We conducted a cohort study to determine respiratory sequelae in patients with severe and critical COVID-19. We aimed at evaluating the proportion of patients with persisting respiratory symptoms and/or abnormalities in pulmonary function tests (PFT) or in lung imaging. STUDY DESIGN: and methods: This is a single center cohort study including COVID-19 survivors who underwent a three-month follow-up with clinical evaluation, PFT and lung high-resolution computed tomography (HRCT). All clinical, functional, and radiological data were centrally reviewed. Multiple linear regression analysis was performed to identify factors associated with residual lesions on HRCT. RESULTS: Full clinical evaluation, PFT and lung HRCT were available for central review in 126, 122 and 107 patients, respectively. At follow-up, 25% of patients complained from dyspnea and 35% from fatigue, lung diffusion capacity (DLCO) was decreased in 45%, 17% had HRCT abnormalities affecting more than 5% of their lung parenchyma while signs of fibrosis were found in 21%. In multiple linear regression model, number of days in ICU were related to the extent of persisting lesions on HRCT, while intubation was associated with signs of fibrosis at follow-up (P = 0.0005, Fisher's exact test). In contrast, the severity of lung imaging or PFT changes were not predictive of fatigue and dyspnea. INTERPRETATION: Although most hospitalized COVID-19 patients recover, a substantial proportion complains from persisting dyspnea and fatigue. Impairment of DLCO and signs suggestive of fibrosis are common but are not strictly related to long-lasting symptoms.