Magnetic Resonance Imaging after Nasopharyngeal Endoscopic Resection and Skull Base Reconstruction.

Background: Postoperative imaging after nasopharyngeal endoscopic resection (NER) and skull base reconstruction is quite challenging due to the complexity of the post-surgical and regional anatomy. Methods: In this retrospective observational study, we included patients treated with NER from 2009 to 2019 and submitted to Magnetic Resonance Imaging (MRI) 6 and 12 months after surgery. A radiologist with 15 years of experience analyzed all MRI scans. Results: A total of 50 patients were considered in this study, 18 of whom were excluded due to imaging unavailability, and 16 of whom were not considered due to major complications and/or persistent disease. Sixteen patients were evaluated to identify the expected findings. Inflammatory changes were observed in 16/64 subsites, and regression of these changes was observed in 8/64 at 1 year. Fibrosis was observed in 5/64 subsites and was unmodified at 1 year. The nasoseptal flap showed homogeneous enhancement at 6 months (100%) and at 1 year. The temporo-parietal fascia flap (TPFF) showed a decrease in the T2- signal intensity of the mucosal layer in 57% of the patients at 1 year and a decrease in enhancement in 43%. Conclusions: Identifying the expected findings after NER and skull base reconstruction has a pivotal role in the identification of complications and recurrence.

CT and MR anatomy of the larynx and hypopharynx.

Imaging of the larynx and hypopharynx is frequently requested to assess the extent of neoplasms beyond the field of view of endoscopic evaluation. The combination of optical and cross-sectional imaging allows tumors to be classified according to AJCC/UICC guidelines. A thorough understanding of laryngeal and hypopharyngeal anatomy is crucial to guide the radiological eye along the possible pathways of the spread of diseases and to guide differential diagnoses. Computed tomography (CT) has been the first cross-sectional imaging technique used to evaluate the larynx and hypopharynx; its spatial resolution combined with volumetric capability and the use of injectable contrast medium made CT the working horse in the assessment of neoplastic and inflammatory diseases. In the last two decades, magnetic resonance (MR) supported CT in the most challenging cases, when the optimal contrast resolution due to the multisequence portfolio is needed to assess the neoplastic involvement of laryngeal cartilages, paraglottic space(s), and extra laryngeal spread. The aim of this paper is to give a comprehensive radiological overview of larynx and hypopharynx complex anatomy, combining in vivo images, anatomical sections, and images of ex vivo specimens.

Pre-TAVI imaging: an Italian survey by the CT PRotocol optimization (CT-PRO) group.

PURPOSE: The purpose of this survey was to evaluate the current state-of-art of pre-TAVI imaging in a large radiological professional community. METHODS: Between December 2022 and January 2023 all members of the Italian Society of Medical and Interventional Radiology (SIRM) were invited by the CT PRotocol Optimization group (CT-PRO group) to complete an online 24-item questionnaire about pre-TAVI imaging. RESULTS: 557 SIRM members participated in the survey. The greatest part of respondents were consultant radiologists employed in public hospitals and 84% claimed to routinely perform pre-TAVI imaging at their institutions. The most widespread acquisition protocol consisted of an ECG-gated CT angiography (CTA) scan of the aortic root and heart followed by a non-ECG-synchronized CTA of the thorax, abdomen, and pelvis. Contrast agent administration was generally tailored on the patient's body weight with a preference for using high concentration contrast media. The reports were commonly written by radiologists with expertise in cardiovascular imaging, and included all the measurements suggested by current guidelines for adequate pre-procedural planning. About 60% of the subjects affirmed that the Heart Team is present at their institutions, however only 7% of the respondents regularly attended the multidisciplinary meetings. CONCLUSIONS: This survey defines the current pre-TAVI imaging practice in a large radiological professional community. Interestingly, despite the majority of radiologists follow the current guidelines regarding acquisition and reporting of pre-TAVI imaging studies, there is still a noteworthy absence from multidisciplinary meetings and from the Heart Team.

Unresectable stage III non-small cell lung cancer: could durvalumab be safe and effective in real-life clinical scenarios? Results of a single-center experience.

Introduction: The standard of care for patients with unresectable stage III non-small cell lung cancer (NSCLC) is chemoradiotherapy (CRT) followed by consolidation durvalumab as shown in the PACIFIC trial. The purpose of this study is to evaluate clinical outcomes and toxicities regarding the use of durvalumab in a real clinical scenario. Methods: A single-center retrospective study was conducted on patients with a diagnosis of unresectable stage III NSCLC who underwent radical CRT followed or not by durvalumab. Tumor response after CRT, pattern of relapse, overall survival (OS) and progression-free survival (PFS), and toxicity profile were investigated. Results: Eighty-five patients met the inclusion criteria. The median age was 67 years (range 45-82 years). Fifty-two patients (61.2%) started sequential therapy with durvalumab. The main reason for excluding patients from the durvalumab treatment was the expression of PD-L1 < 1%. Only two patients presented a grade 4 or 5 pneumonitis. A median follow-up (FU) of 20 months has been reached. Forty-five patients (52.9%) had disease progression, and 21 (24.7%) had a distant progression. The addition of maintenance immunotherapy confirmed a clinical benefit in terms of OS and PFS. Two-year OS and PFS were respectively 69.4% and 54.4% in the durvalumab group and 47.9% and 24.2% in the no-durvalumab group (p = 0.015, p = 0.007). Conclusion: In this real-world study, patients treated with CRT plus durvalumab showed clinical outcomes and toxicities similar to the PACIFIC results. Maintenance immunotherapy after CRT has been shown to be safe and has increased the survival of patients in clinical practice.

Temporomandibular joint involvement in patients with Juvenile Idiopathic Arthritis: comparison of ultrasonography and magnetic resonance imaging in assessing the periarticular space width.

OBJECTIVES: This study aimed to compare the performance of Ultrasonography (US) and Magnetic Resonance Imaging (MRI) in assessing the Lateral Periarticular Space (LPAS) of Temporomandibular Joints (TMJs) in patients with Juvenile Idiopathic Arthritis (JIA). METHODS: The LPAS width was evaluated in two different patient groups. In the JIA group, including 29 children (13 ± 2.8 years) with JIA, the LPAS width was measured with both MRI and US. In the healthy group, including 28 healthy children (12.6 ± 2.5 years), the LPAS width was measured only with US. Comparisons of LPAS width based on patient groups and TMJ contrast enhancement in MRI were evaluated by applying the Mann-Whitney U test. Correlation and agreement between MRI and US measurements in JIA group were tested using Spearman rank correlation and Bland-Altman method. RESULTS: The LPAS width was significantly greater in the JIA group than in the healthy group. In the JIA group, the LPAS width was significantly greater in TMJs with moderate/severe enhancement than those with mild enhancement. A positive significant correlation between MRI and US measurements of LPAS width was found in the JIA group. In the same group, Bland-Altman method showed a good level of agreement between MRI and US measurements. CONCLUSION: Although, US cannot replace MRI in the evaluation of TMJ in patients with JIA, US could be used as a supplementary imaging method to MRI in assessing the TMJ disease.

Software-based quantitative CT analysis to predict the growth trend of persistent nonsolid pulmonary nodules: a retrospective study.

PURPOSE: Persistent nonsolid nodules (NSNs) usually exhibit an indolent course and may remain stable for several years; however, some NSNs grow quickly and require surgical excision. Therefore, identifying quantitative features capable of early discrimination between growing and nongrowing NSNs is becoming a crucial aspect of radiological analysis. The main purpose of this study was to evaluate the performance of an open-source software (ImageJ) to predict the future growth of NSNs detected in a Caucasian (Italian) population. MATERIAL AND METHODS: We retrospectively selected 60 NSNs with an axial diameter of 6-30 mm scanned with the same acquisition-reconstruction parameters and the same computed tomography (CT) scanner. Software-based analysis was performed on thin-section CT images using ImageJ. For each NSNs, several quantitative features were extracted from the baseline CT images. The relationships of NSN growth with quantitative CT features and other categorical variables were analyzed using univariate and multivariable logistic regression analyses. RESULTS: In multivariable analysis, only the skewness and linear mass density (LMD) were significantly associated with NSN growth, and the skewness was the strongest predictor of growth. In receiver operating characteristic curve analyses, the optimal cutoff values of skewness and LMD were 0.90 and 19.16 mg/mm, respectively. The two predictive models that included the skewness, with or without LMD, exhibited an excellent power for predicting NSN growth. CONCLUSION: According to our results, NSNs with a skewness value > 0.90, specifically those with a LMD > 19.16 mg/mm, should require closer follow-up due to their higher growth potential, and higher risk of becoming an active cancer.

M100 ExaData: a data collection campaign on the CINECA's Marconi100 Tier-0 supercomputer.

Supercomputers are the most powerful computing machines available to society. They play a central role in economic, industrial, and societal development. While they are used by scientists, engineers, decision-makers, and data-analyst to computationally solve complex problems, supercomputers and their hosting datacenters are themselves complex power-hungry systems. Improving their efficiency, availability, and resiliency is vital and the subject of many research and engineering efforts. Still, a major roadblock hinders researchers: dearth of reliable data describing the behavior of production supercomputers. In this paper, we present the result of a ten-year-long project to design a monitoring framework (EXAMON) deployed at the Italian supercomputers at CINECA datacenter. We disclose the first holistic dataset of a tier-0 Top10 supercomputer. It includes the management, workload, facility, and infrastructure data of the Marconi100 supercomputer for two and half years of operation. The dataset (published via Zenodo) is the largest ever made public, with a size of 49.9TB before compression. We also provide open-source software modules to simplify access to the data and provide direct usage examples.

COVID-19 Pneumonia with Migratory Pattern in Agammaglobulinemic Patients: A Report of Two Cases and Review of Literature.

X-linked agammaglobulinemia (XLA) is a primary immunodeficiency characterized by marked reduction in serum immunoglobulins and early-onset infections. Coronavirus Disease-2019 (COVID-19) pneumonia in immunocompromised patients presents clinical and radiological peculiarities which have not yet been completely understood. Very few cases of agammaglobulinemic patients with COVID-19 have been reported since the beginning of the pandemic in February 2020. We report two cases of migrant COVID-19 pneumonia in XLA patients.

Multi-center validation of an artificial intelligence system for detection of COVID-19 on chest radiographs in symptomatic patients.

OBJECTIVES: While chest radiograph (CXR) is the first-line imaging investigation in patients with respiratory symptoms, differentiating COVID-19 from other respiratory infections on CXR remains challenging. We developed and validated an AI system for COVID-19 detection on presenting CXR. METHODS: A deep learning model (RadGenX), trained on 168,850 CXRs, was validated on a large international test set of presenting CXRs of symptomatic patients from 9 study sites (US, Italy, and Hong Kong SAR) and 2 public datasets from the US and Europe. Performance was measured by area under the receiver operator characteristic curve (AUC). Bootstrapped simulations were performed to assess performance across a range of potential COVID-19 disease prevalence values (3.33 to 33.3%). Comparison against international radiologists was performed on an independent test set of 852 cases. RESULTS: RadGenX achieved an AUC of 0.89 on 4-fold cross-validation and an AUC of 0.79 (95%CI 0.78-0.80) on an independent test cohort of 5,894 patients. Delong's test showed statistical differences in model performance across patients from different regions (p < 0.01), disease severity (p < 0.001), gender (p < 0.001), and age (p = 0.03). Prevalence simulations showed the negative predictive value increases from 86.1% at 33.3% prevalence, to greater than 98.5% at any prevalence below 4.5%. Compared with radiologists, McNemar's test showed the model has higher sensitivity (p < 0.001) but lower specificity (p < 0.001). CONCLUSION: An AI model that predicts COVID-19 infection on CXR in symptomatic patients was validated on a large international cohort providing valuable context on testing and performance expectations for AI systems that perform COVID-19 prediction on CXR. KEY POINTS: • An AI model developed using CXRs to detect COVID-19 was validated in a large multi-center cohort of 5,894 patients from 9 prospectively recruited sites and 2 public datasets. • Differences in AI model performance were seen across region, disease severity, gender, and age. • Prevalence simulations on the international test set demonstrate the model's NPV is greater than 98.5% at any prevalence below 4.5%.

Residual Lung Abnormalities in Survivors of Severe or Critical COVID-19 at One-Year Follow-Up Computed Tomography: A Narrative Review Comparing the European and East Asian Experiences.

The literature reports that there was a significant difference in the medical impact of the coronavirus disease (COVID-19) pandemic between European and East Asian countries; specifically, the mortality rate of COVID-19 in Europe was significantly higher than that in East Asia. Considering such a difference, our narrative review aimed to compare the prevalence and characteristics of residual lung abnormalities at one-year follow-up computed tomography (CT) after severe or critical COVID-19 in survivors of European and East Asian countries. A literature search was performed to identify articles focusing on the prevalence and characteristics of CT lung abnormalities in survivors of severe or critical COVID-19. Database analysis identified 16 research articles, 9 from Europe and 7 from East Asia (all from China). Our analysis found a higher prevalence of CT lung abnormalities in European than in Chinese studies (82% vs. 52%). While the most prevalent lung abnormalities in Chinese studies were ground-glass opacities (35%), the most prevalent lung abnormalities in European studies were linear (59%) and reticular opacities (55%), followed by bronchiectasis (46%). Although our findings required confirmation, the higher prevalence and severity of lung abnormalities in European than in Chinese survivors of COVID-19 may reflect a greater architectural distortion due to a more severe lung damage.

Home-based pulmonary rehabilitation in patients undergoing (chemo)radiation therapy for unresectable lung cancer: a prospective explorative study.

AIMS: The prevention of pulmonary toxicity is an important goal for patient candidate to radiation therapy for lung cancer. There is a lack of evidence on the role of exercise training for patients with unresectable stage III lung cancer candidated to radical treatment. The aim of this study was to evaluate the feasibility of a home-based pulmonary rehabilitation (PR) program and to identify reliable tools in terms of respiratory function, exercise capacity and quality of life. METHODS: Patients' recruitment lasted from April 2020 till February 2022. The PR program was proposed concomitantly to radiation therapy to the first 20 patients (interventional group, IG), and the other 20 patients were identified as an observational group (OG). All patients were assessed at baseline (T0) and after 8 weeks (T2) with 6 minute walking test (6MWT), modified Borg Scale (mBORG), SF-36 questionnaire (SF-36) and pulmonary function test (PFT); after 4 weeks (T1), only SF-36 was administered. RESULTS: A decrease of 13.8 m in the walked-distance was registered in the OG between T0 and T2 (p = 0.083). Instead, an increase of 56.6 m in the distance walked was recorded in the IG between T0 and T2 (p ≤ 0.001). In the OG, the mBORG scores showed a negative trend. On the contrary, in the IG, these scores showed a slight improvement. In the OG, all the items of SF-36 scores decreased between T0 and T1. In the IG, an increased trend from T0 to T2 was observed for all the items of SF-36. No clinically significant variations were detected from baseline to T2 in both groups regarding PFT. CONCLUSION: The 6MWT, mBORG and SF-36 resulted as useful tools to assess the role of a PR program. A significant gain in functional exercise capacity and a prevention of the physiological impairment of QoL during radio(chemo)therapy was registered.

Breath-holding physiology, radiological severity and adverse outcomes in COVID-19 patients: A prospective validation study.

BACKGROUND AND OBJECTIVE: COVID-19 remains a major cause of respiratory failure, and means to identify future deterioration is needed. We recently developed a prediction score based on breath-holding manoeuvres (desaturation and maximal duration) to predict incident adverse COVID-19 outcomes. Here we prospectively validated our breath-holding prediction score in COVID-19 patients, and assessed associations with radiological scores of pulmonary involvement. METHODS: Hospitalized COVID-19 patients (N = 110, three recruitment centres) performed breath-holds at admission to provide a prediction score (Messineo et al.) based on mean desaturation (20-s breath-holds) and maximal breath-hold duration, plus baseline saturation, body mass index and cardiovascular disease. Odds ratios for incident adverse outcomes (composite of bi-level ventilatory support, ICU admission and death) were described for patients with versus without elevated scores (>0). Regression examined associations with chest x-ray (Brixia score) and computed tomography (CT; 3D-software quantification). Additional comparisons were made with the previously-validated '4C-score'. RESULTS: Elevated prediction score was associated with adverse COVID-19 outcomes (N = 12/110), OR[95%CI] = 4.54[1.17-17.83], p = 0.030 (positive predictive value = 9/48, negative predictive value = 59/62). Results were diminished with removal of mean desaturation from the prediction score (OR = 3.30[0.93-11.72]). The prediction score rose linearly with Brixia score (ß[95%CI] = 0.13[0.02-0.23], p = 0.026, N = 103) and CT-based quantification (ß = 1.02[0.39-1.65], p = 0.002, N = 45). Mean desaturation was also associated with both radiological assessment. Elevated 4C-scores (≥high-risk category) had a weaker association with adverse outcomes (OR = 2.44[0.62-9.56]). CONCLUSION: An elevated breath-holding prediction score is associated with almost five-fold increased adverse COVID-19 outcome risk, and with pulmonary deficits observed in chest imaging. Breath-holding may identify COVID-19 patients at risk of future respiratory failure.

Detection Rate and Variability in Measurement of Mandibular Incisive Canal on Cone-Beam Computed Tomography: A Study of 220 Dentate Hemi-Mandibles from Italy.

The mandibular incisive canal (MIC) is a small bony channel located in the interforaminal region; it represents the anterior continuation of the mandibular canal. Cone-beam computed tomography (CBCT) is the most commonly utilized radiological technique for assessing the MIC. The main purpose of this study was to evaluate the detectability and variability in measurements of the MIC on CBCT. A total of 220 dentate hemi-mandibles were retrospectively selected for this study. For each hemi-mandible, the detectability, diameter, and distance of the MIC from anatomical landmarks (cortical plates and tooth apices) were evaluated in consensus by two observers. The analysis was performed at four different levels (first premolar, canine, lateral incisor, and central incisor) and was repeated after one month. The variability of MIC measurements was expressed as the coefficient of repeatability (CR), obtained from the Bland-Altman analysis. The MIC detection rate reduced from the first premolar to the central incisor (from 82.3% to 0.5%). The CR of MIC measurements (diameter and distances from anatomical landmarks) was ≤0.74 mm. Although the MIC is difficult to detect in a non-negligible percentage of cases, the limited variability in measurements confirms that CBCT is an effective technique for the assessment of the MIC.

Every cloud has a silver lining: COVID-19 chest-CT screening prevents unnecessary cardiac surgery.

BACKGROUND: Unenhanced chest CT can identify incidental findings (IFs) leading to management strategy change. We report our institutional experience with routine chest-CT as preoperative screening tool during the COVID-19 pandemic, focusing on the impact of IFs. METHODS: All patients scheduled for cardiac surgery from May 1st to December 31st 2020, underwent preoperative unenhanced chest-CT according to COVID-19 pandemic institutional protocol. We have analyzed IFs incidence, reported consequent operative changes, and identified IFs clinical determinants. RESULTS: Out of 447, 278 patients were included. IFs rate was 7.2% (20/278): a solid mass (11/20, 55%), lymphoproliferative disease (1/20, 5%), SARS-CoV-2 pneumonia (2/20, 10%), pulmonary artery chronic thromboembolism (1/20, 5%), anomalous vessel anatomy (2/20, 10%), voluminous hiatal hernia (1/20, 5%), mitral annulus calcification (1/20, 5%), and porcelain aorta (1/20, 5%) were reported. Based on IFs, 4 patients (20%-4/278, 1.4%) were not operated, 8 (40%-8/278, 2.9%) underwent a procedure different from the one originally planned one, and 8 (40%-8/278, 2.9%) needed additional preoperative investigations before undergoing the planned surgery. At univariate regression, coronary artery disease, atrial fibrillation, and history of cancer were significantly more often present in patients presenting with significant IFs. History of malignancy was identified as the only independent determinant of significant IFs at chest-CT (OR=4.27 IQR: [1.14-14.58], P=0.0227). CONCLUSIONS: Unenhanced chest-CT as a preoperative screening tool in cardiac surgery led to incidental detection of significant clinical findings, which justified even procedures cancellation. Malignancy history is a determinant for CT incidental findings and could support a tailored screening approach for high-risk patients.

One-year pulmonary impairment after severe COVID-19: a prospective, multicenter follow-up study.

BACKGROUND: Long-term pulmonary sequelae following hospitalization for SARS-CoV-2 pneumonia is largely unclear. The aim of this study was to identify and characterise pulmonary sequelae caused by SARS-CoV-2 pneumonia at 12-month from discharge. METHODS: In this multicentre, prospective, observational study, patients hospitalised for SARS-CoV-2 pneumonia and without prior diagnosis of structural lung diseases were stratified by maximum ventilatory support ("oxygen only", "continuous positive airway pressure (CPAP)" and "invasive mechanical ventilation (IMV)") and followed up at 12 months from discharge. Pulmonary function tests and diffusion capacity for carbon monoxide (DLCO), 6 min walking test, high resolution CT (HRCT) scan, and modified Medical Research Council (mMRC) dyspnea scale were collected. RESULTS: Out of 287 patients hospitalized with SARS-CoV-2 pneumonia and followed up at 1 year, DLCO impairment, mainly of mild entity and improved with respect to the 6-month follow-up, was observed more frequently in the "oxygen only" and "IMV" group (53% and 49% of patients, respectively), compared to 29% in the "CPAP" group. Abnormalities at chest HRCT were found in 46%, 65% and 80% of cases in the "oxygen only", "CPAP" and "IMV" group, respectively. Non-fibrotic interstitial lung abnormalities, in particular reticulations and ground-glass attenuation, were the main finding, while honeycombing was found only in 1% of cases. Older patients and those requiring IMV were at higher risk of developing radiological pulmonary sequelae. Dyspnea evaluated through mMRC scale was reported by 35% of patients with no differences between groups, compared to 29% at 6-month follow-up. CONCLUSION: DLCO alteration and non-fibrotic interstitial lung abnormalities are common after 1 year from hospitalization due to SARS-CoV-2 pneumonia, particularly in older patients requiring higher ventilatory support. Studies with longer follow-ups are needed.

Chest X-ray versus chest computed tomography for outcome prediction in hospitalized patients with COVID-19.

The purpose of this study was to compare the prognostic value of chest X-ray (CXR) and chest computed tomography (CT) in a group of hospitalized patients with COVID-19. For this study, we retrospectively selected a cohort of 106 hospitalized patients with COVID-19 who underwent both CXR and chest CT at admission. For each patient, the pulmonary involvement was ranked by applying the Brixia score for CXR and the percentage of well-aerated lung (WAL) for CT. The Brixia score was assigned at admission (A-Brixia score) and during hospitalization. During hospitalization, only the highest score (H-Brixia score) was considered. At admission, the percentage of WAL (A-CT%WAL) was quantified using a dedicated software. On logistic regression analyses, H-Brixia score was the most effective radiological marker for predicting in-hospital mortality and invasive mechanical ventilation. Additionally, A-CT%WAL did not provide substantial advantages in the risk stratification of hospitalized patients with COVID-19 compared to A-Brixia score.

Early signs of pneumoconiosis in a dental technician in Italy: a case report.

BACKGROUND: Dental technicians are at high risk of pneumoconiosis, usually driven by inhalation of mixed dusts, including metals. An etiological diagnosis is not easy to be performed, particularly in advanced stages. CASE PRESENTATION: We describe the case of an early pneumoconiosis occurring in a 47-year-old dental technician who developed respiratory symptoms shortly after beginning work. She described the work environment as dusty and lacking relevant primary prevention tools. A chest CT showed multiple peripheral pseudonodular lesions in both lower lobes; bronchoalveolar lavage and bronchial aspirate evidenced numerous macrophages with reflective metal bodies included into the cytoplasm, that at scanning electron microscopy coupled to Energy Dispersive X-Ray Analysis resulted Zirconium and Aluminum, whereas Tungsten (W) was localized outside cells. End of shift urinary concentrations of W were substantially raised as compared to pre-shift (1.1 vs. 0.2 µg/L). CONCLUSIONS: We concluded for diagnosis of early work-related pneumoconiosis due to abnormal occupational exposure to metals. The case demonstrates the need also for dental professionals to comply with industrial hygiene standards and to be monitored by occupational health physicians.

Anterior and Central Skull Base Tumors: Key Points for the Radiologist to Analyze Prior to Endoscopic Approaches.

Initially developed as a minimally invasive technique to approach inflammatory conditions, transnasal endoscopic surgery has progressively expanded its anatomic targets and clinical indications. Consequently, numerous surgical approaches to the anterior and central skull base were developed, referred to as extended endonasal approaches (EEA). The intrinsic advantage of EEA is the exploitation of a natural corridor provided by sinonasal airspaces, with no need for skin incision and osteotomy and limited soft tissue damage. In this context, imaging plays essential role, demonstrating the relevant anatomic relationships of the lesion, the proper surgical corridor, the anatomic variants that may increase the surgical risk.

Early prediction of in-hospital death of COVID-19 patients: a machine-learning model based on age, blood analyses, and chest x-ray score.

An early-warning model to predict in-hospital mortality on admission of COVID-19 patients at an emergency department (ED) was developed and validated using a machine-learning model. In total, 2782 patients were enrolled between March 2020 and December 2020, including 2106 patients (first wave) and 676 patients (second wave) in the COVID-19 outbreak in Italy. The first-wave patients were divided into two groups with 1474 patients used to train the model, and 632 to validate it. The 676 patients in the second wave were used to test the model. Age, 17 blood analytes, and Brescia chest X-ray score were the variables processed using a random forests classification algorithm to build and validate the model. Receiver operating characteristic (ROC) analysis was used to assess the model performances. A web-based death-risk calculator was implemented and integrated within the Laboratory Information System of the hospital. The final score was constructed by age (the most powerful predictor), blood analytes (the strongest predictors were lactate dehydrogenase, D-dimer, neutrophil/lymphocyte ratio, C-reactive protein, lymphocyte %, ferritin std, and monocyte %), and Brescia chest X-ray score (https://bdbiomed.shinyapps.io/covid19score/). The areas under the ROC curve obtained for the three groups (training, validating, and testing) were 0.98, 0.83, and 0.78, respectively. The model predicts in-hospital mortality on the basis of data that can be obtained in a short time, directly at the ED on admission. It functions as a web-based calculator, providing a risk score which is easy to interpret. It can be used in the triage process to support the decision on patient allocation.