Lung vascular changes as biomarkers of severity in systemic sclerosis-associated interstitial lung disease.

OBJECTIVES: It has recently become possible to assess lung vascular and parenchymal changes quantitatively in thoracic CT images using automated software tools. We investigated the vessel parameters of patients with SSc, quantified by CT imaging, and correlated them with interstitial lung disease (ILD) features. METHODS: SSc patients undergoing standard of care pulmonary function testing and CT evaluation were retrospectively evaluated. CT images were analysed for ILD patterns and total pulmonary vascular volume (PVV) extents with Imbio lung texture analysis. Vascular analysis (volumes, numbers and densities of vessels, separating arteries and veins) was performed with an in-house developed software. A threshold of 5% ILD extent was chosen to define the presence of ILD, and commonly used cut-offs of lung function were adopted. RESULTS: A total of 79 patients [52 women, 40 ILD, mean age 56.2 (s.d. 14.2) years, total ILD extent 9.5 (10.7)%, PVV/lung volume % 2.8%] were enrolled. Vascular parameters for total and separated PVV significantly correlated with functional parameters and ILD pattern extents. SSc-associated ILD (SSc-ILD) patients presented with an increased number and volume of arterial vessels, in particular those between 2 and 4 mm of diameter, and with a higher density of arteries and veins of <6 mm in diameter. Considering radiological and functional criteria concomitantly, as well as the descriptive trends from the longitudinal evaluations, the normalized PVVs, vessel numbers and densities increased progressively with the increase/worsening of ILD extent and functional impairment. CONCLUSION: In SSc patients CT vessel parameters increase in parallel with ILD extent and functional impairment, and may represent a biomarker of SSc-ILD severity.

Syndrome of Combined Pulmonary Fibrosis and Emphysema: An Official ATS/ERS/JRS/ALAT Research Statement.

Background: The presence of emphysema is relatively common in patients with fibrotic interstitial lung disease. This has been designated combined pulmonary fibrosis and emphysema (CPFE). The lack of consensus over definitions and diagnostic criteria has limited CPFE research. Goals: The objectives of this task force were to review the terminology, definition, characteristics, pathophysiology, and research priorities of CPFE and to explore whether CPFE is a syndrome. Methods: This research statement was developed by a committee including 19 pulmonologists, 5 radiologists, 3 pathologists, 2 methodologists, and 2 patient representatives. The final document was supported by a focused systematic review that identified and summarized all recent publications related to CPFE. Results: This task force identified that patients with CPFE are predominantly male, with a history of smoking, severe dyspnea, relatively preserved airflow rates and lung volumes on spirometry, severely impaired DlCO, exertional hypoxemia, frequent pulmonary hypertension, and a dismal prognosis. The committee proposes to identify CPFE as a syndrome, given the clustering of pulmonary fibrosis and emphysema, shared pathogenetic pathways, unique considerations related to disease progression, increased risk of complications (pulmonary hypertension, lung cancer, and/or mortality), and implications for clinical trial design. There are varying features of interstitial lung disease and emphysema in CPFE. The committee offers a research definition and classification criteria and proposes that studies on CPFE include a comprehensive description of radiologic and, when available, pathological patterns, including some recently described patterns such as smoking-related interstitial fibrosis. Conclusions: This statement delineates the syndrome of CPFE and highlights research priorities.

Development of a clinical decision support system for severity risk prediction and triage of COVID-19 patients at hospital admission: an international multicentre study.

BACKGROUND: The outbreak of coronavirus disease 2019 (COVID-19) has globally strained medical resources and caused significant mortality. OBJECTIVE: To develop and validate a machine-learning model based on clinical features for severity risk assessment and triage for COVID-19 patients at hospital admission. METHOD: 725 patients were used to train and validate the model. This included a retrospective cohort from Wuhan, China of 299 hospitalised COVID-19 patients from 23 December 2019 to 13 February 2020, and five cohorts with 426 patients from eight centres in China, Italy and Belgium from 20 February 2020 to 21 March 2020. The main outcome was the onset of severe or critical illness during hospitalisation. Model performances were quantified using the area under the receiver operating characteristic curve (AUC) and metrics derived from the confusion matrix. RESULTS: In the retrospective cohort, the median age was 50 years and 137 (45.8%) were male. In the five test cohorts, the median age was 62 years and 236 (55.4%) were male. The model was prospectively validated on five cohorts yielding AUCs ranging from 0.84 to 0.93, with accuracies ranging from 74.4% to 87.5%, sensitivities ranging from 75.0% to 96.9%, and specificities ranging from 55.0% to 88.0%, most of which performed better than the pneumonia severity index. The cut-off values of the low-, medium- and high-risk probabilities were 0.21 and 0.80. The online calculators can be found at www.covid19risk.ai. CONCLUSION: The machine-learning model, nomogram and online calculator might be useful to access the onset of severe and critical illness among COVID-19 patients and triage at hospital admission.

Quantitative analysis of pulmonary vasculature in systemic sclerosis at spirometry-gated chest CT.

OBJECTIVE: To prospectively investigate whether differences in pulmonary vasculature exist in systemic sclerosis (SSc) and how they are distributed in patients with different pulmonary function. METHODS: Seventy-four patients with SSc undergoing chest CT scan for interstitial lung disease (ILD) screening or follow-up were prospectively enrolled. A thorough clinical, laboratory and functional evaluation was performed the same day. Chest CT was spirometry gated at total lung capacity and images were analysed by two automated software programs to quantify emphysema, ILD patterns (ground-glass, reticular, honeycombing), and pulmonary vascular volume (PVV). Patients were divided in restricted (FVC% <80, DLco%<80), isolated DLco% reduction (iDLco- FVC%≥80, DLco%<80) and normals (FVC%≥80, DLco%≥80). Spearman ρ, Mann-Whitney tests and logistic regressions were used to assess for correlations, differences among groups and relationships between continuous variables. RESULTS: Absolute and lung volume normalised PVV (PVV/LV) correlated inversely with functional parameters and positively with all ILD patterns (ρ=0.75 with ground glass, ρ=0.68 with reticular). PVV/LV was the only predictor of DLco at multivariate analysis (p=0.007). Meanwhile, the reticular pattern prevailed in peripheral regions and lower lung thirds, PVV/LV prevailed in central regions and middle lung thirds. iDLco group had a significantly higher PVV/LV (2.2%) than normal (1.6%), but lower than restricted ones (3.8%). CONCLUSIONS: Chest CT in SSc detects a progressive increase in PVV/LV as DLco decreases. Redistribution of perfusion to less affected lung regions rather than angiogenesis nearby fibrotic lung may explain the results. Further studies to ascertain whether the increase in PVV/LV reflects a real increase in blood volume are needed.

Pleuroparenchymal fibroelastosis in rheumatic autoimmune diseases: a systematic literature review.

OBJECTIVES: Pleuroparenchymal fibroelastosis (PPFE) is characterized by predominantly upper lobe pleural and subjacent parenchymal fibrosis; PPFE features were described in patients with rheumatic autoimmune diseases (RAID). A systematic literature review was performed to investigate the prevalence, prognosis and potential association of PPFE with previous immunosuppression in RAID. METHODS: EMBASE, Web of Science and PubMed databases were questioned from inception to 1 September 2019. Articles published in English and addressing PPFE in patients with RAID were selected. RESULTS: Twenty out of 794 papers were selected with a total of 76 cases of RAID-PPFE patients (20 SSc, 9 RA, 6 IIM6 primary SS, 5 overlap syndromes, 3 ANCA-associated vasculitides, 2 granulomatosis with polyangiitis, 1 microscopic polyangiitis, 1 UCTD, 1 SLE, 1 GCA and 21 patients with non-specified RAID). Dyspnoea was the most frequently reported symptom (37/48 patients, 77%). Patients frequently presented with a restrictive pattern and decline in diffusing lung capacity for carbon monoxide. During the follow-up, 7/12 patients had progression at imaging, 22/39 presented a generic clinical worsening, 19/38 had a functional deterioration and 15/43 remained stable. CONCLUSION: The present systematic literature review confirms that PPFE features are present in RAID. Rheumatologists should be aware of this new radiological pattern that holds a bad prognosis.

Validation of a method to assess emphysema severity by spirometry in the COPDGene study.

BACKGROUND: Standard spirometry cannot identify the predominant mechanism underlying airflow obstruction in COPD, namely emphysema or airway disease. We aimed at validating a previously developed methodology to detect emphysema by mathematical analysis of the maximal expiratory flow-volume (MEFV) curve in standard spirometry. METHODS: From the COPDGene population we selected those 5930 subjects with MEFV curve and inspiratory-expiratory CT obtained on the same day. The MEFV curve descending limb was fit real-time using forced vital capacity (FVC), peak expiratory flow, and forced expiratory flows at 25, 50 and 75% of FVC to derive an emphysema severity index (ESI), expressed as a continuous positive numeric parameter ranging from 0 to 10. According to inspiratory CT percent lung attenuation area below - 950 HU we defined three emphysema severity subgroups (%LAA-950insp < 6, 6-14, ≥14). By co-registration of inspiratory-expiratory CT we quantified persistent (%pLDA) and functional (%fLDA) low-density areas as CT metrics of emphysema and airway disease, respectively. RESULTS: ESI differentiated CT emphysema severity subgroups increasing in parallel with GOLD stages (p < .001), but with high variability within each stage. ESI had significantly higher correlations (p < .001) with emphysema than with airway disease CT metrics, explaining 67% of %pLDA variability. Conversely, standard spirometric variables (FEV1, FEV1/FVC) had significantly lower correlations than ESI with emphysema CT metrics and did not differentiate between emphysema and airways CT metrics. CONCLUSIONS: ESI adds to standard spirometry the power to discriminate whether emphysema is the predominant mechanism of airway obstruction. ESI methodology has been validated in the large multiethnic population of smokers of the COPDGene study and therefore it could be applied for clinical and research purposes in the general population of smokers, using a readily available online website.

The role of diffusion-weighted and dynamic contrast enhancement perfusion-weighted imaging in the evaluation of salivary glands neoplasms.

OBJECTIVES: To evaluate the association of magnetic resonance diffusion-weighted imaging (DwI) and dynamic contrast-enhanced perfusion-weighted imaging (DCE-PwI) with a temporal resolution of 5 s, wash-in < 120 s, and wash-out ratio > 30% in the evaluation of salivary glands neoplasms. METHODS: DwI and DCE-PwI of 92 salivary glands neoplasms were assessed. The apparent diffusion coefficient (ADC) was calculated by drawing three regions of interest with an average area of 0.30-0.40 cm2 on three contiguous axial sections. The time/intensity curve was generated from DCE-PwI images by drawing a region of interest that included at least 50% of the largest lesion section. Vessels, calcifications, and necrotic/haemorrhagic or cystic areas within solid components were excluded. The association of ADC ≥ 1.4 × 10-3 mm2/s with type A curves (progressive wash-in) and ADC 0.9-1.4 × 10-3 mm2/s with type C curves (rapid wash-in/slow wash-out) were tested as parameters of benignity and malignancy, respectively. Type B curve (rapid wash-in/rapid wash-out) was not used as a reference parameter. RESULTS: ADC ≥ 1.4 × 10-3 mm2/s and type A curves were observed only in benign neoplasms. ADC of 0.9-1.4 × 10-3 mm2/s and type C curves association showed specificity of 94.9% and positive predictive value of 81.8% for epithelial malignancies. The association of ADC < 0.9 × 10-3 mm2/s with type B and C curves showed diagnostic accuracy of 94.6% and 100% for Warthin tumour and lymphoma, respectively. CONCLUSIONS: ADC ≥ 1.4 × 10-3 mm2/s and type A curves association was indicative of benignity. Lymphomas exhibited ADC < 0.7 × 10-3 mm2/s and type C curves. The association of ADC < 0.9 × 10-3 mm2/s and type B and C curves had accuracy 94.6% and 88.5% for Warthin tumour and epithelial malignancies, respectively.

Spirometric assessment of emphysema presence and severity as measured by quantitative CT and CT-based radiomics in COPD.

BACKGROUND: The mechanisms underlying airflow obstruction in COPD cannot be distinguished by standard spirometry. We ascertain whether mathematical modeling of airway biomechanical properties, as assessed from spirometry, could provide estimates of emphysema presence and severity, as quantified by computed tomography (CT) metrics and CT-based radiomics. METHODS: We quantified presence and severity of emphysema by standard CT metrics (VIDA) and co-registration analysis (ImbioLDA) of inspiratory-expiratory CT in 194 COPD patients who underwent pulmonary function testing. According to percentages of low attenuation area below - 950 Hounsfield Units (%LAA-950insp) patients were classified as having no emphysema (NE) with %LAA-950insp < 6, moderate emphysema (ME) with %LAA-950insp ≥ 6 and < 14, and severe emphysema (SE) with %LAA-950insp ≥ 14. We also obtained stratified clusters of emphysema CT features by an automated unsupervised radiomics approach (CALIPER). An emphysema severity index (ESI), derived from mathematical modeling of the maximum expiratory flow-volume curve descending limb, was compared with pulmonary function data and the three CT classifications of emphysema presence and severity as derived from CT metrics and radiomics. RESULTS: ESI mean values and pulmonary function data differed significantly in the subgroups with different emphysema degree classified by VIDA, ImbioLDA and CALIPER (p < 0.001 by ANOVA). ESI differentiated NE from ME/SE CT-classified patients (sensitivity 0.80, specificity 0.85, AUC 0.86) and SE from ME CT-classified patients (sensitivity 0.82, specificity 0.87, AUC 0.88). CONCLUSIONS: Presence and severity of emphysema in patients with COPD, as quantified by CT metrics and radiomics can be estimated by mathematical modeling of airway function as derived from standard spirometry.

The Relevance of Family History Taking in the Detection and Management of Birt-Hogg-Dubé Syndrome.

BACKGROUND: Birt-Hogg-Dubé syndrome (BHDS) is a rare autosomal-dominant inherited disorder characterized by inactivation of the gene Folliculin (FLCN), pulmonary cysts with recurrent spontaneous pneumothorax, dermatological lesions, and an increased risk of developing renal malignancies. OBJECTIVES: We aimed to investigate the real prevalence of BHDS and its prevalence among patients with a familial history of pneumothorax. METHODS: From July 2014 to December 2016, we consecutively studied all patients with spontaneous pneumothorax and a positive family history for the same condition referring to our Institution. The suspicious cases underwent genetic analysis of the BHDS-causative gene FLCN. FLCN-positive cases were further evaluated with routine blood tests, chest radiography, chest CT, abdominal MRI, and dermatological evaluation. RESULTS: Among 114 patients admitted with spontaneous pneumothorax, 7 patients had a family history of pneumothorax, and 6/7 (85.7%) patients had positive genetic test for FLCN as well as 7/13 family members. Pulmonary cysts were found in all patients with a FLCN-positive genetic test. Most patients (10/13, 76.9%) had tiny pulmonary cysts less than 1 cm in diameter. The vast majority of cysts were intraparenchymal (12/13, 92.3%) and located in lower lobes. Dermatological lesions were found in 7/13 (54%) patients, renal cysts in 4/13 (31%) patients, and renal cancer in 1 (1/13, 7.7%) patient. CONCLUSIONS: Although BHDS is considered a rare disease, BHDS underlies spontaneous pneumothorax more often than usually believed, especially whenever a family history of pneumothorax is present. Diagnosis of BHDS is essential to start monitoring patients for the risk of developing renal malignancies.

Emphysematous and Nonemphysematous Gas Trapping in Chronic Obstructive Pulmonary Disease: Quantitative CT Findings and Pulmonary Function.

Purpose To identify a prevalent computed tomography (CT) subtype in patients with chronic obstructive pulmonary disease (COPD) by separating emphysematous from nonemphysematous contributions to total gas trapping and to attempt to predict and grade the emphysematous gas trapping by using clinical and functional data. Materials and Methods Two-hundred and two consecutive eligible patients (159 men and 43 women; mean age, 70 years [age range, 41-85 years]) were prospectively studied. Pulmonary function and CT data were acquired by pulmonologists and radiologists. Noncontrast agent-enhanced thoracic CT scans were acquired at full inspiration and expiration, and were quantitatively analyzed by using two software programs. CT parameters were set as follows: 120 kVp; 200 mAs; rotation time, 0.5 second; pitch, 1.1; section thickness, 0.75 mm; and reconstruction kernels, b31f and b70f. Gas trapping obtained by difference of inspiratory and expiratory CT density thresholds (percentage area with CT attenuation values less than -950 HU at inspiration and percentage area with CT attenuation values less than -856 HU at expiration) was compared with that obtained by coregistration analysis. A logistic regression model on the basis of anthropometric and functional data was cross-validated and trained to classify patients with COPD according to the relative contribution of emphysema to total gas trapping, as assessed at CT. Results Gas trapping obtained by difference of inspiratory and expiratory CT density thresholds was highly correlated (r = 0.99) with that obtained by coregistration analysis. Four groups of patients were distinguished according to the prevalent CT subtype: prevalent emphysematous gas trapping, prevalent functional gas trapping, mixed severe, and mixed mild. The predictive model included predicted forced expiratory volume in 1 second/vital capacity, percentage of predicted forced expiratory volume in 1 second, percentage of diffusing capacity for carbon monoxide, and body mass index as emphysema regressors at CT, with 81% overall accuracy in classifying patients according to its extent. Conclusion The relative contribution of emphysematous and nonemphysematous gas trapping obtained by coregistration of inspiratory and expiratory CT scanning can be determined accurately by difference of CT inspiratory and expiratory density thresholds. CT extent of emphysema can be predicted with accuracy suitable for clinical purposes by pulmonary function data and body mass index. © RSNA, 2018 Online supplemental material is available for this article.

Radiation dose in non-dental cone beam CT applications: a systematic review.

BACKGROUND: Radiation-induced health risks are broadly questioned in the literature. As cone beam computed tomography (CBCT) is increasingly used in non-dental examinations, its effective dose needs to be known. This study aimed to review the published evidence on effective dose of non-dental CBCT for diagnostic use by focusing on dosimetry system used to estimate dose. MATERIALS AND METHODS: A systematic review of the literature was performed on 12 November 2017. All the literature up to this date was included. The PubMed and web of science databases were searched. Studies were screened for inclusion based on defined inclusion and exclusion criteria according to the preferred reporting items for systematic reviews. RESULTS: Fifteen studies met the inclusion criteria and were included in our review. Thirteen and two of them examined one and two anatomical areas, respectively. The anatomical areas were: ear (6), paranasal sinuses (4), ankle (3), wrist (2), knee (1), and cervical spine (1). Effective dose was estimated by different methods: (i) RANDO phantom associated with thermoluminescent dosimeters (6), metal oxide semiconductor field-effect transistor dosimeters (3), and optically stimulated luminescent dosimeters (1). (ii) Scanner outputs, namely computed tomography dose index (1) and dose area product (2). (iii) Monte Carlo simulations (2). CONCLUSION: CBCT of extremities, cervical spine, ears and paranasal sinuses was found to be a low-dose volumetric imaging technique. Effective doses varied significantly because of different exposure settings of CBCT-units and different dosimetry systems used to estimate dose.

Normal spectrum of pulmonary parametric response map to differentiate lung collapsibility: distribution of densitometric classifications in healthy adult volunteers.

OBJECTIVES: Pulmonary parametric response map (PRM) was proposed for quantitative densitometric phenotypization of chronic obstructive pulmonary disease. However, little is known about this technique in healthy subjects. The purpose of this study was to describe the normal spectrum of densitometric classification of pulmonary PRM in a group of healthy adults. METHODS: 15 healthy volunteers underwent spirometrically monitored chest CT at total lung capacity (TLC) and functional residual capacity (FRC). The paired CT scans were analyzed by PRM for voxel-by-voxel characterization of lung parenchyma according to 4 densitometric classifications: normal lung (TLC ≥ -950 HU, FRC ≥ -856 HU); expiratory low attenuation area (LAA) (TLC ≥ -950 HU, FRC < -856 HU); dual LAA (TLC<-950 HU, FRC < -856 HU); uncharacterized (TLC < -950 HU, FRC ≥ -856 HU). RESULTS: PRM spectrum was 78 % ± 10 % normal lung, 20 % ± 8 % expiratory LAA, and 1 % ± 1 % dual LAA. PRM was similar between genders, there was moderate correlation between dual LAA and spirometrically assessed TLC (R = 0.531; p = 0.042), and between expiratory LAA and VolExp/Insp ratio (R = -0.572; p = 0.026). CONCLUSIONS: PRM reflects the predominance of normal lung parenchyma in a group of healthy volunteers. However, PRM also confirms the presence of physiological expiratory LAA seemingly related to air trapping and a minimal amount of dual LAA likely reflecting emphysema. KEY POINTS: • Co-registration of inspiratory and expiratory computed tomography allows dual-phase densitometry. • Dual-phase co-registered densitometry reflects heterogeneous regional changes in lung function. • Quantification of lung in healthy subjects is needed to set reference values. • Expiratory low attenuation areas <30 % could be considered within normal range.

Imaging of Large Airways Disorders.

OBJECTIVE: Recent technical advances, including the routine use of CT thin sections and techniques such as 2D minimum-intensity-projection and 3D volume images, have increased our ability to detect large airways diseases. Furthermore, dedicated CT protocols allow the evaluation of dynamic airway dysfunction. CONCLUSION: With diseases of the large airways more commonly seen in daily practice, it is important that radiologists be familiar with the appearances, differential diagnosis, and clinical implications of these entities.

A single-centre experience of 200 consecutive unselected patients in percutaneous EVAR.

PURPOSE: This study evaluated the feasibility and safety of percutaneous endovascular aneurysm repair (P-EVAR) in an "unselected" population of patients using Prostar XL closure device, comparing our results with the literature data reported for open EVAR (O-EVAR) and P-EVAR in selected populations. MATERIALS AND METHODS: We prospectively enrolled 200 patients with abdominal aortic aneurysms to be treated with the P-EVAR procedure, without any selection criteria. Four hundred percutaneous femoral accesses using the Prostar XL closure device were performed. The early follow-up protocol consisted of a Doppler ultrasound (US) examination. Later evaluations were performed with US and computed tomography. RESULTS: Technical success was achieved in all percutaneous accesses (100 %). Ten early complications related to access site were recorded (10/400; 2.5 %): four pseudoaneurysms and six cases of lower limb ischaemia. Five of ten complications occurred in patients presenting calcification of the common femoral arteries, whereas 4/10 were in patients with "complex" groin anatomy. Eight of ten complications occurred at the access site of the main body of the prosthesis using a sheath size >20 Fr. CONCLUSIONS: Percutaneous endovascular aneurysm repair in "unselected" patients is safe and efficient, with a very low risk of access-related complications, comparable to P-EVAR in selected populations and to the best O-EVAR series.

Spectral Photon-Counting Computed Tomography: Technical Principles and Applications in the Assessment of Cardiovascular Diseases.

Spectral Photon-Counting Computed Tomography (SPCCT) represents a groundbreaking advancement in X-ray imaging technology. The core innovation of SPCCT lies in its photon-counting detectors, which can count the exact number of incoming x-ray photons and individually measure their energy. The first part of this review summarizes the key elements of SPCCT technology, such as energy binning, energy weighting, and material decomposition. Its energy-discriminating ability represents the key to the increase in the contrast between different tissues, the elimination of the electronic noise, and the correction of beam-hardening artifacts. Material decomposition provides valuable insights into specific elements' composition, concentration, and distribution. The capability of SPCCT to operate in three or more energy regimes allows for the differentiation of several contrast agents, facilitating quantitative assessments of elements with specific energy thresholds within the diagnostic energy range. The second part of this review provides a brief overview of the applications of SPCCT in the assessment of various cardiovascular disease processes. SPCCT can support the study of myocardial blood perfusion and enable enhanced tissue characterization and the identification of contrast agents, in a manner that was previously unattainable.

Machine Learning to Identify Patients at Risk of Developing New-Onset Atrial Fibrillation after Coronary Artery Bypass.

BACKGROUND: This study aims to get an effective machine learning (ML) prediction model of new-onset postoperative atrial fibrillation (POAF) following coronary artery bypass grafting (CABG) and to highlight the most relevant clinical factors. METHODS: Four ML algorithms were employed to analyze 394 patients undergoing CABG, and their performances were compared: Multivariate Adaptive Regression Spline, Neural Network, Random Forest, and Support Vector Machine. Each algorithm was applied to the training data set to choose the most important features and to build a predictive model. The better performance for each model was obtained by a hyperparameters search, and the Receiver Operating Characteristic Area Under the Curve metric was selected to choose the best model. The best instances of each model were fed with the test data set, and some metrics were generated to assess the performance of the models on the unseen data set. A traditional logistic regression was also performed to be compared with the machine learning models. RESULTS: Random Forest model showed the best performance, and the top five predictive features included age, preoperative creatinine values, time of aortic cross-clamping, body surface area, and Logistic Euro-Score. CONCLUSIONS: The use of ML for clinical predictions requires an accurate evaluation of the models and their hyperparameters. Random Forest outperformed all other models in the clinical prediction of POAF following CABG.

Towards the Definition of Radiomic Features and Clinical Indices to Enhance the Diagnosis of Clinically Significant Cancers in PI-RADS 4 and 5 Lesions.

Prostate cancer (PC) is the most frequently diagnosed cancer among adult men, and its incidence is increasing worldwide [...].

Ultrashort Echo-Time Magnetic Resonance Imaging Sequence in the Assessment of Systemic Sclerosis-Interstitial Lung Disease.

PURPOSE: To test respiratory-triggered ultrashort echo-time (UTE) Spiral VIBE-MRI sequence in systemic sclerosis-interstitial lung disease assessment compared with computed tomography (CT). MATERIAL AND METHODS: Fifty four SSc patients underwent chest CT and UTE (1.5 T). Two radiologists, independently and in consensus, verified ILD presence/absence and performed a semiquantitative analysis (sQA) of ILD, ground-glass opacities (GGO), reticulations and honeycombing (HC) extents on both scans. A CT software quantitative texture analysis (QA) was also performed. For ILD detection, intra-/inter-reader agreements were computed with Cohen K coefficient. UTE sensitivity and specificity were assessed. For extent assessments, intra-/inter-reader agreements and UTE performance against CT were computed by Lin's concordance coefficient (CCC). RESULTS: Three UTE were discarded for low quality, 51 subjects were included in the study. Of them, 42 QA segmentations were accepted. ILD was diagnosed in 39/51 CT. UTE intra-/inter-reader K in ILD diagnosis were 0.56 and 0.26. UTE showed 92.8% sensitivity and 75.0% specificity. ILD, GGO, and reticulation extents were 14.8%, 7.7%, and 7.1% on CT sQA and 13.0%, 11.2%, and 1.6% on CT QA. HC was <1% and not further considered. UTE intra-/inter-reader CCC were 0.92 and 0.89 for ILD extent and 0.84 and 0.79 for GGO extent. UTE RET extent intra-/inter-reader CCC were 0.22 and 0.18. UTE ILD and GGO extents CCC against CT sQA and QA were ≥0.93 and ≥0.88, respectively. RET extent CCC were 0.35 and 0.22 against sQA and QA, respectively. CONCLUSION: UTE Spiral VIBE-MRI sequence is reliable in assessing ILD and GGO extents in systemic sclerosis-interstitial lung disease patients.

External validation of a radiomic signature to predict p16 (HPV) status from standard CT images of anal cancer patients.

The paper deals with the evaluation of the performance of an existing and previously validated CT based radiomic signature, developed in oropharyngeal cancer to predict human papillomavirus (HPV) status, in the context of anal cancer. For the validation in anal cancer, a dataset of 59 patients coming from two different centers was collected. The primary endpoint was HPV status according to p16 immunohistochemistry. Predefined statistical tests were performed to evaluate the performance of the model. The AUC obtained here in anal cancer is 0.68 [95% CI (0.32-1.00)] with F1 score of 0.78. This signature is TRIPOD level 4 (57%) with an RQS of 61%. This study provides proof of concept that this radiomic signature has the potential to identify a clinically relevant molecular phenotype (i.e., the HPV-ness) across multiple cancers and demonstrates potential for this radiomic signature as a CT imaging biomarker of p16 status.

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.