Diagnostic Performance of MRI for the Detection of Pulmonary Nodules: A Systematic Review and Meta-Analysis

Purpose To perform a meta-analysis of the diagnostic performance of MRI for the detection of pulmonary nodules, with use of CT as the reference standard. Materials and Methods PubMed, Embase, Scopus, and other databases were systematically searched for studies published from January 2000 to March 2023 evaluating the performance of MRI for diagnosis of lung nodules measuring 4 mm or larger, with CT as reference. Studies including micronodules, nodules without size stratification, or those from which data for contingency tables could not be extracted were excluded. Primary outcomes were the per-lesion sensitivity of MRI and the rate of false-positive nodules per patient (FPP). Subgroup analysis by size and meta-regression with other covariates were performed. The study protocol was registered in the International Prospective Register of Systematic Reviews, or PROSPERO (no. CRD42023437509). Results Ten studies met inclusion criteria (1354 patients and 2062 CT-detected nodules). Overall, per-lesion sensitivity of MRI for nodules measuring 4 mm or larger was 87.7% (95% CI: 81.1, 92.2), while the FPP rate was 12.4% (95% CI: 7.0, 21.1). Subgroup analyses demonstrated that MRI sensitivity was 98.5% (95% CI: 90.4, 99.8) for nodules measuring at least 8-10 mm and 80.5% (95% CI: 71.5, 87.1) for nodules less than 8 mm. Conclusion MRI demonstrated a good overall performance for detection of pulmonary nodules measuring 4 mm or larger and almost equal performance to CT for nodules measuring at least 8-10 mm, with a low rate of FPP. Systematic review registry no. CRD42023437509 Keywords: Lung Nodule, Lung Cancer, Lung Cancer Screening, MRI, CT Supplemental material is available for this article. © RSNA, 2024.

Evaluating Acute Pulmonary Changes in Coronavirus Disease 2019: A Comparative Analysis of Computed Tomography, Chest Radiography, Lung Ultrasound, Magnetic Resonance Imaging, and Positron Emission Tomography with Fluorodeoxyglucose Modalities

This review explores imaging's crucial role in acute Coronavirus Disease 2019 (COVID-19) assessment. High Resolution Computer Tomography is especially effective in detection of lung abnormalities. Chest radiography has limited utility in the initial stages of COVID-19 infection. Lung Ultrasound has emerged as a valuable, radiation-free tool in critical care, and Magnetic Resonance Imaging shows promise as a Computed Tomography alternative. Typical and atypical findings of COVID-19 by each of these modalities are discussed with emphasis on their prognostic value. Considerations for pediatric and immunocompromised cases are outlined. A comprehensive diagnostic approach is recommended, as radiological diagnosis remains challenging in the acute phase.

Predicting major adverse cardiovascular events after orthotopic liver transplantation using a supervised machine learning model: A cohort study

BACKGROUND

Liver transplant (LT) patients have become older and sicker. The rate of post-LT major adverse cardiovascular events (MACE) has increased, and this in turn raises 30-d post-LT mortality. Noninvasive cardiac stress testing loses accuracy when applied to pre-LT cirrhotic patients.

AIM

To assess the feasibility and accuracy of a machine learning model used to predict post-LT MACE in a regional cohort.

METHODS

This retrospective cohort study involved 575 LT patients from a Southern Brazilian academic center. We developed a predictive model for post-LT MACE (defined as a composite outcome of stroke, new-onset heart failure, severe arrhythmia, and myocardial infarction) using the extreme gradient boosting (XGBoost) machine learning model. We addressed missing data (below 20%) for relevant variables using the k-nearest neighbor imputation method, calculating the mean from the ten nearest neighbors for each case. The modeling dataset included 83 features, encompassing patient and laboratory data, cirrhosis complications, and pre-LT cardiac assessments. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC). We also employed Shapley additive explanations (SHAP) to interpret feature impacts. The dataset was split into training (75%) and testing (25%) sets. Calibration was evaluated using the Brier score. We followed Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis guidelines for reporting. Scikit-learn and SHAP in Python 3 were used for all analyses. The supplementary material includes code for model development and a user-friendly online MACE prediction calculator.

RESULTS

Of the 537 included patients, 23 (4.46%) developed in-hospital MACE, with a mean age at transplantation of 52.9 years. The majority, 66.1%, were male. The XGBoost model achieved an impressive AUROC of 0.89 during the training stage. This model exhibited accuracy, precision, recall, and F1-score values of 0.84, 0.85, 0.80, and 0.79, respectively. Calibration, as assessed by the Brier score, indicated excellent model calibration with a score of 0.07. Furthermore, SHAP values highlighted the significance of certain variables in predicting postoperative MACE, with negative noninvasive cardiac stress testing, use of nonselective beta-blockers, direct bilirubin levels, blood type O, and dynamic alterations on myocardial perfusion scintigraphy being the most influential factors at the cohort-wide level. These results highlight the predictive capability of our XGBoost model in assessing the risk of post-LT MACE, making it a valuable tool for clinical practice.

CONCLUSION

Our study successfully assessed the feasibility and accuracy of the XGBoost machine learning model in predicting post-LT MACE, using both cardiovascular and hepatic variables. The model demonstrated impressive performance, aligning with literature findings, and exhibited excellent calibration. Notably, our cautious approach to prevent overfitting and data leakage suggests the stability of results when applied to prospective data, reinforcing the model's value as a reliable tool for predicting post-LT MACE in clinical practice.

Evaluating ChatGPT's Accuracy in Providing Screening Mammography Recommendations among Older Women: Artificial Intelligence and Cancer Communication

Abstract Objective: The U.S. Preventive Services Task Force (USPSTF) recommends biennial screening mammography through age 74. Guidelines vary as to whether or not they recommended mammography screening to women aged 75 and older. This study aims to determine the ability of ChatGPT to provide appropriate recommendations for breast cancer screening in patients aged 75 years and older. Methods: 12 questions and 4 clinical vignettes addressing fundamental concepts about breast cancer screening and prevention in patients aged 75 years and older were created and asked to ChatGPT three consecutive times to generate 3 sets of responses. The responses were graded by a multi-disciplinary panel of experts in the intersection of breast cancer screening and aging . The responses were graded as 'appropriate', 'inappropriate', or 'unreliable' based on the reviewer's clinical judgment, content of the response, and whether the content was consistent across the three responses . Appropriateness was determined through a majority consensus. Results: The responses generated by ChatGPT were appropriate for 11/17 questions (64%). Three questions were graded as inappropriate (18%) and 2 questions were graded as unreliable (12%). A consensus was not reached on one question (6%) and was graded as no consensus. Conclusions: While recognizing the limitations of ChatGPT, it has potential to provide accurate health care information and could be utilized by healthcare professionals to assist in providing recommendations for breast cancer screening in patients age 75 years and older. Physician oversight will be necessary, due to the possibility of ChatGPT to provide inappropriate and unreliable responses, and the importance of accuracy in medicine.

The International Association for the Study of Lung Cancer Early Lung Imaging Confederation Open-Source Deep Learning and Quantitative Measurement Initiative

INTRODUCTION

With global adoption of computed tomography (CT) lung cancer screening, there is increasing interest to use artificial intelligence (AI) deep learning methods to improve the clinical management process. To enable AI research using an open-source, cloud-based, globally distributed, screening CT imaging data set and computational environment that are compliant with the most stringent international privacy regulations that also protect the intellectual properties of researchers, the International Association for the Study of Lung Cancer sponsored development of the Early Lung Imaging Confederation (ELIC) resource in 2018. The objective of this report is to describe the updated capabilities of ELIC and illustrate how this resource can be used for clinically relevant AI research.

METHODS

In this second phase of the initiative, metadata and screening CT scans from two time points were collected from 100 screening participants in seven countries. An automated deep learning AI lung segmentation algorithm, automated quantitative emphysema metrics, and a quantitative lung nodule volume measurement algorithm were run on these scans.

RESULTS

A total of 1394 CTs were collected from 697 participants. The LAV950 quantitative emphysema metric was found to be potentially useful in distinguishing lung cancer from benign cases using a combined slice thickness more than or equal to 2.5 mm. Lung nodule volume change measurements had better sensitivity and specificity for classifying malignant from benign lung nodules when applied to solid lung nodules from high-quality CT scans.

CONCLUSIONS

These initial experiments revealed that ELIC can support deep learning AI and quantitative imaging analyses on diverse and globally distributed cloud-based data sets.

MRI with DWI improves detection of liver metastasis and selection of surgical candidates with pancreatic cancer: a systematic review and meta-analysis

OBJECTIVE

To perform a systematic review and meta-analysis to evaluate if magnetic resonance imaging (MRI) with diffusion weighted imaging (DWI) adds value compared to contrast-enhanced computed tomography (CECT) alone in the preoperative evaluation of pancreatic cancer.

METHODS

MEDLINE, EMBASE, and Cochrane databases were searched for relevant published studies through October 2022. Studies met eligibility criteria if they evaluated the per-patient diagnostic performance of MRI with DWI in the preoperative evaluation of newly diagnosed pancreatic cancer compared to CECT. Our primary outcome was the number needed to treat (NNT) to prevent one futile surgery using MRI with DWI, defined as those in which CECT was negative and MRI with DWI was positive for liver metastasis (i.e., surgical intervention in metastatic disease missed by CECT). The secondary outcomes were to determine the diagnostic performance and the NNT of MRI with DWI to change management in pancreatic cancer.

RESULTS

Nine studies met the inclusion criteria with a total of 1121 patients, of whom 172 had liver metastasis (15.3%). The proportion of futile surgeries reduced by MRI with DWI was 6.0% (95% CI, 3.0-11.6%), yielding an NNT of 16.6. The proportion of cases that MRI with DWI changed management was 18.1% (95% CI, 9.9-30.7), corresponding to an NNT of 5.5. The per-patient sensitivity and specificity of MRI were 92.4% (95% CI, 87.4-95.6%) and 97.3% (95% CI, 96.0-98.1).

CONCLUSION

MRI with DWI may prevent futile surgeries in pancreatic cancer by improving the detection of occult liver metastasis on preoperative CECT with an NNT of 16.6.

CLINICAL RELEVANCE STATEMENT

MRI with DWI complements the standard preoperative CECT evaluation for liver metastasis in pancreatic cancer, improving the selection of surgical candidates and preventing unnecessary surgeries.

KEY POINTS

• The NNT of MRI with DWI to prevent potential futile surgeries due to occult liver metastasis on CECT, defined as those in which CECT was negative and MRI with DWI was positive for liver metastasis, in patients with pancreatic cancer was 16.6. • The higher performance of MRI with DWI to detect liver metastasis occult on CECT can be attributed to an increased detection of subcentimeter liver metastasis.

Detection of Liver Lesions in Colorectal Cancer Patients Using 18F-FDG PET/CT Dual-Time-Point Scan Imaging

OBJECTIVE

The aim of this study was to evaluate the diagnostic performance of dual-time-point fluorine-18-fluorodeoxyglucose positron emission computed tomography/computed tomography (18F-FDG PET/CT) compared to conventional early imaging for detecting colorectal liver metastases (CRLM) in colorectal cancer (CRC) patients.

METHODS

One hundred twenty-four consecutive CRC patients underwent dual-time-point imaging scans on a retrospective basis. Histopathological confirmation and/or clinical follow-up were accepted as the gold standard. Standard uptake values (SUV), signal-to-noise ratio (SNR), retention index (RI), tumor-to-normal liver ratio (TNR), and lesion sizes were measured for early and delayed PET scans. The diagnostic performance of early and delayed images was calculated on a per-patient basis and compared using McNemar's test.

RESULTS

Among the 124 patients, 57 (46%) had CRLM, 6 (4.8%) had benign lesions, and 61 (49.2%) had no concerning lesions detected. Smaller CRLM lesions (<5 cm3) showed significantly higher uptake in the delayed scans relative to early imaging (p < 0.001). The SUV and TNR increased significantly in delayed imaging of all metastatic lesions (p < 0.001). The retention index of all CRLM was high (40.8%), especially for small lesions (54.8%). A total of 177 lesions in delayed images and 124 in standard early images were identified. In a per-patient analysis, delayed imaging had significantly higher sensitivity (100% vs. 87.7%) and specificity (91.0% vs. 94.0%) compared to early imaging (p-value = 0.04).

CONCLUSIONS

The detection of liver lesions using dual-time-point PET/CT scan improves the sensitivity and specificity for the detection of colorectal liver metastasis.

A New (Digital) Era in Medical Journalism: Leveraging Social Media and Other Online Tools to Increase Reach and Engagement

Social media is a popular communication and marketing tool in modern society, with the power to reach and engage large audiences. Many members of the medical and radiology communities have embraced social media platforms, particularly X (formerly known as Twitter), as an efficient and economic means for performing patient outreach, disseminating research and educational materials, building networks, and promoting diversity. Editors of medical journals with a clear vision and relevant expertise can leverage social media and other digital tools to advance the journal's mission, further their interests, and directly benefit journal authors and readers. For editors, social media offers a means to increase article visibility and downloads, expand awareness of volunteer opportunities, and use metrics and other feedback to inform future initiatives. Authors benefit from broader dissemination of their work, which aids establishment of a national or international reputation. Readers can receive high-quality high-yield content in a digestible format directly on their devices while actively engaging with journal editors and authors in the online community. The authors highlight the multifaceted benefits of social media engagement and digital tool implementation in the context of medical journalism and summarize the activities of the RadioGraphics Social Media and Digital Innovation Team. By enumerating the social media activities of RadioGraphics and describing the underlying rationale for each activity, the authors present a blueprint for other medical journals considering similar initiatives. ©RSNA, 2023.

Imaging Findings in Aspergillosis: From Head to Toe

Aspergillosis is a mycotic infection induced by airborne fungi that are ubiquitous. Inhalation of Aspergillus conidia results in transmission through the respiratory tract. The clinical presentation is dependent on organism and host specifics, with immunodeficiency, allergies, and preexisting pulmonary disease constituting the most important risk factors. In recent decades, the incidence of fungal infections has increased dramatically, due in part to the increased number of transplants and the pervasive use of chemotherapy and immunosuppressive drugs. The spectrum of clinical manifestations can range from an asymptomatic or mild infection to a swiftly progressive, life-threatening illness. Additionally, invasive infections can migrate to extrapulmonary sites, causing infections in distant organs. Recognition and familiarity with the various radiological findings in the appropriate clinical context are essential for patient management and the prompt initiation of life-saving treatment. We discuss the radiological characteristics of chronic and invasive pulmonary aspergillosis, as well as some of the typically unexpected extrapulmonary manifestations of disseminated disease.

Machine learning for malignant versus benign focal liver lesions on US and CEUS: a meta-analysis

OBJECTIVES

To perform a meta-analysis of the diagnostic performance of learning (ML) algorithms (conventional and deep learning algorithms) for the classification of malignant versus benign focal liver lesions (FLLs) on US and CEUS.

METHODS

Available databases were searched for relevant published studies through September 2022. Studies met eligibility criteria if they evaluate the diagnostic performance of ML for the classification of malignant and benign focal liver lesions on US and CEUS. The pooled per-lesion sensitivities and specificities for each modality with 95% confidence intervals were calculated.

RESULTS

A total of 8 studies on US, 11 on CEUS, and 1 study evaluating both methods met the inclusion criteria with a total of 34,245 FLLs evaluated. The pooled sensitivity and specificity of ML for the malignancy classification of FLLs were 81.7% (95% CI, 77.2-85.4%) and 84.8% (95% CI, 76.0-90.8%) for US, compared to 87.1% (95% CI, 81.8-91.0%) and 87.0% (95% CI, 83.1-90.1%) for CEUS. In the subgroup analysis of studies that evaluated deep learning algorithms, the sensitivity and specificity of CEUS (n = 4) increased to 92.4% (95% CI, 88.5-95.0%) and 88.2% (95% CI, 81.1-92.9%).

CONCLUSIONS

The diagnostic performance of ML algorithms for the malignant classification of FLLs was high for both US and CEUS with overall similar sensitivity and specificity. The similar performance of US may be related to the higher prevalence of DL models in that group.

Factors influencing the outcome of volumetry tools for pulmonary nodule analysis: a systematic review and attempted meta-analysis

Health systems worldwide are implementing lung cancer screening programmes to identify early-stage lung cancer and maximise patient survival. Volumetry is recommended for follow-up of pulmonary nodules and outperforms other measurement methods. However, volumetry is known to be influenced by multiple factors. The objectives of this systematic review (PROSPERO CRD42022370233) are to summarise the current knowledge regarding factors that influence volumetry tools used in the analysis of pulmonary nodules, assess for significant clinical impact, identify gaps in current knowledge and suggest future research. Five databases (Medline, Scopus, Journals@Ovid, Embase and Emcare) were searched on the 21st of September, 2022, and 137 original research studies were included, explicitly testing the potential impact of influencing factors on the outcome of volumetry tools. The summary of these studies is tabulated, and a narrative review is provided. A subset of studies (n = 16) reporting clinical significance were selected, and their results were combined, if appropriate, using meta-analysis. Factors with clinical significance include the segmentation algorithm, quality of the segmentation, slice thickness, the level of inspiration for solid nodules, and the reconstruction algorithm and kernel in subsolid nodules. Although there is a large body of evidence in this field, it is unclear how to apply the results from these studies in clinical practice as most studies do not test for clinical relevance. The meta-analysis did not improve our understanding due to the small number and heterogeneity of studies testing for clinical significance. CRITICAL RELEVANCE STATEMENT: Many studies have investigated the influencing factors of pulmonary nodule volumetry, but only 11% of these questioned their clinical relevance in their management. The heterogeneity among these studies presents a challenge in consolidating results and clinical application of the evidence. KEY POINTS: • Factors influencing the volumetry of pulmonary nodules have been extensively investigated. • Just 11% of studies test clinical significance (wrongly diagnosing growth). • Nodule size interacts with most other influencing factors (especially for smaller nodules). • Heterogeneity among studies makes comparison and consolidation of results challenging. • Future research should focus on clinical applicability, screening, and updated technology.

Natural Language Processing for the Identification of Incidental Lung Nodules in Computed Tomography Reports: A Quality Control Tool

PURPOSE

To evaluate the diagnostic performance of a natural language processing (NLP) model in detecting incidental lung nodules (ILNs) in unstructured chest computed tomography (CT) reports.

METHODS

All unstructured consecutive reports of chest CT scans performed at a tertiary hospital between 2020 and 2021 were retrospectively reviewed (n = 21,542) to train the NLP tool. Internal validation was performed using reference readings by two radiologists of both CT scans and reports, using a different external cohort of 300 chest CT scans. Second, external validation was performed in a cohort of all random unstructured chest CT reports from 57 different hospitals conducted in May 2022. A review by the same thoracic radiologists was used as the gold standard. The sensitivity, specificity, and accuracy were calculated.

RESULTS

Of 21,542 CT reports, 484 mentioned at least one ILN (mean age, 71 ± 17.6 [standard deviation] years; women, 52%) and were included in the training set. In the internal validation (n = 300), the NLP tool detected ILN with a sensitivity of 100.0% (95% CI, 97.6 to 100.0), a specificity of 95.9% (95% CI, 91.3 to 98.5), and an accuracy of 98.0% (95% CI, 95.7 to 99.3). In the external validation (n = 977), the NLP tool yielded a sensitivity of 98.4% (95% CI, 94.5 to 99.8), a specificity of 98.6% (95% CI, 97.5 to 99.3), and an accuracy of 98.6% (95% CI, 97.6 to 99.2). Twelve months after the initial reports, 8 (8.60%) patients had a final diagnosis of lung cancer, among which 2 (2.15%) would have been lost to follow-up without the NLP tool.

CONCLUSION

NLP can be used to identify ILNs in unstructured reports with high accuracy, allowing a timely recall of patients and a potential diagnosis of early-stage lung cancer that might have been lost to follow-up.

Computed tomography aspects of thoracic metastases from osteosarcoma: pictorial essay

Osteosarcoma is the most common primary bone tumor, with a higher incidence in the second decade of life, and it often leads to pulmonary metastases. The most common pattern seen on computed tomography is one of multiple well-defined nodules in the lung parenchyma, often with calcifications. Because of the variety of presentations of pulmonary metastases in osteosarcoma, including atypical forms, knowledge of the computed tomography aspects of these lesions is important for characterizing and evaluating the extent of the disease, as well as for distinguishing metastatic disease from other benign or malignant lung diseases. This essay discusses the main tomographic findings of pulmonary metastases from osteosarcoma.

Medical imaging and multimodal artificial intelligence models for streamlining and enhancing cancer care: opportunities and challenges

INTRODUCTION

Artificial intelligence (AI) has the potential to transform oncologic care. There have been significant developments in AI applications in medical imaging and increasing interest in multimodal models. These are likely to enable improved oncologic care through more precise diagnosis, increasingly in a more personalized and less invasive manner. In this review, we provide an overview of the current state and challenges that clinicians, administrative personnel and policy makers need to be aware of and mitigate for the technology to reach its full potential.

AREAS COVERED

The article provides a brief targeted overview of AI, a high-level review of the current state and future potential AI applications in diagnostic radiology and to a lesser extent digital pathology, focusing on oncologic applications. This is followed by a discussion of emerging approaches, including multimodal models. The article concludes with a discussion of technical, regulatory challenges and infrastructure needs for AI to realize its full potential.

EXPERT OPINION

There is a large volume of promising research, and steadily increasing commercially available tools using AI. For the most advanced and promising precision diagnostic applications of AI to be used clinically, robust and comprehensive quality monitoring systems and informatics platforms will likely be required.

Diagnostic accuracy of pelvis multiparametric MRI against CT virtual hysterosalpingography: a prospective study of tubal patency through female infertility assessment

OBJECTIVE

To evaluate the diagnostic accuracy of MRI-hysterosalpingogram (HSG) with semiquantitative dynamic contrast-enhanced perfusion, against the virtual multislice CT hysterosalpingogram (VHSG) as a reference standard.

METHODS AND MATERIALS

In this prospective study, 26 women (age >18 years) searching for infertility causes and with VHSG physician request. Thereafter, the assessment performance of both techniques was determined by two reader analyses. k statistics were used for the assessment of tubal patency. Receiver operating characteristic (ROC) analysis was used to compare the capability for tubal patency assessment between both exams on a per-patient and per-tube basis. The McNemar test was used to compare the diagnostic accuracy measures.

RESULTS

Tubal patency, uterine morphological, ovarian, and extrauterine abnormalities were evaluated through both exams in all 26 women. There was no significant difference between diagnostic performance measurements between the methods. The ROC curve of VHSG was 0.852 for both per-patient and per-tube analyses, and one and 0.938 for MRI-HSG. Sensitivity and specificity for per-patient and per-tube for VHSG were 95.2 and 97.7, 80 and 87.5%, and for MRI-HSG 100% for both analyses and 100 and 87.5%, respectively.

CONCLUSION

This study demonstrates the feasibility of diagnosing tubal patency through MRI, using a semi-quantitative dynamic contrast-enhanced perfusion sequence, and the satisfactory diagnosing of the uterine morphology, ovarian abnormalities, and ovarian and deep endometriosis.

ADVANCES IN KNOWLEDGE

Multiparametric MRI with a perfusion real-time sequence as a HSG method can be used in the evaluation not only for uterine and ovarian abnormilities but also tubal patency.

Computed tomography findings of pulmonary histoplasmosis: pictorial essay

Endemic systemic mycoses are prevalent in specific geographic areas of the world and are responsible for high rates of morbidity and mortality in the populations of such areas, as well as in immigrants and travelers returning from endemic regions. Pulmonary histoplasmosis is an infection caused by Histoplasma capsulatum, a dimorphic fungus. This infection has a worldwide distribution, being endemic in Brazil. Histoplasmosis can affect the lungs, and its diagnosis and management remain challenging, especially in non-endemic areas. Therefore, recognition of the various radiological manifestations of pulmonary histoplasmosis, together with the clinical and epidemiological history of the patient, is essential to narrowing the differential diagnosis. This essay discusses the main computed tomography findings of pulmonary histoplasmosis.

Applications of Artificial Intelligence in the Radiology Roundtrip: Process Streamlining, Workflow Optimization, and Beyond

There are many impactful applications of artificial intelligence (AI) in the electronic radiology roundtrip and the patient's journey through the healthcare system that go beyond diagnostic applications. These tools have the potential to improve quality and safety, optimize workflow, increase efficiency, and increase patient satisfaction. In this article, we review the role of AI for process improvement and workflow enhancement which includes applications beginning from the time of order entry, scan acquisition, applications supporting the image interpretation task, and applications supporting tasks after image interpretation such as result communication. These non-diagnostic workflow and process optimization tasks are an important part of the arsenal of potential AI tools that can streamline day to day clinical practice and patient care.

Artificial intelligence to predict the need for mechanical ventilation in cases of severe COVID-19

Objective

To determinate the accuracy of computed tomography (CT) imaging assessed by deep neural networks for predicting the need for mechanical ventilation (MV) in patients hospitalized with severe acute respiratory syndrome due to coronavirus disease 2019 (COVID-19).

Materials and Methods

This was a retrospective cohort study carried out at two hospitals in Brazil. We included CT scans from patients who were hospitalized due to severe acute respiratory syndrome and had COVID-19 confirmed by reverse transcription-polymerase chain reaction (RT-PCR). The training set consisted of chest CT examinations from 823 patients with COVID-19, of whom 93 required MV during hospitalization. We developed an artificial intelligence (AI) model based on convolutional neural networks. The performance of the AI model was evaluated by calculating its accuracy, sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve.

Results

For predicting the need for MV, the AI model had a sensitivity of 0.417 and a specificity of 0.860. The corresponding area under the ROC curve for the test set was 0.68.

Conclusion

The high specificity of our AI model makes it able to reliably predict which patients will and will not need invasive ventilation. That makes this approach ideal for identifying high-risk patients and predicting the minimum number of ventilators and critical care beds that will be required.

Introduction to Radiomics and Artificial Intelligence: A Primer for Radiologists

Health informatics and artificial intelligence (AI) are expected to transform the healthcare enterprise and the future practice of radiology. There is an increasing body of literature on radiomics and deep learning/AI applications in medical imaging. There are also a steadily increasing number of FDA cleared AI applications in radiology. It is therefore essential for radiologists to have a basic understanding of these approaches, whether in academia or private practice. In this article, we will provide an overview of the field and familiarize the readers with the fundamental concepts behind these approaches.

Artificial Intelligence for Cardiothoracic Imaging: Overview of Current and Emerging Applications

Artificial intelligence algorithms can learn by assimilating information from large datasets in order to decipher complex associations, identify previously undiscovered pathophysiological states, and construct prediction models. There has been tremendous interest and increased incorporation of artificial intelligence into various industries, including healthcare. As a result, there has been an exponential rise in the number of research articles and industry participants producing models intended for a variety of applications in medical imaging, which can be challenging to navigate for radiologists. In thoracic imaging, multiple applications are being evaluated for chest radiography and computed tomography and include applications for lung nodule evaluation and cancer imaging, quantifying diffuse lung disorders, and cardiac imaging, to name a few. This review aims to provide an overview of current clinical AI models, focusing on the most common clinical applications of AI in cardiothoracic imaging.

Diagnostic accuracy of magnetic resonance elastography and point-shear wave elastography for significant hepatic fibrosis screening: Systematic review and meta-analysis

The hepatic diseases are extremely common in clinical practice. The correct classification of liver fibrosis is extremely important, as it influences therapy and predicts disease outcomes. The purpose of this study is to compare the diagnostic performance of point-shear wave elastography (pSWE) and magnetic resonance elastography (MRE) in the hepatic fibrosis diagnostic. A meta-analysis was carried out based on articles published until October 2020. The articles are available at following databases: MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, Scientific Electronic Library Online, LILACS, Scopus, and CINAHL. Diagnostic performances were analyzed per METAVIR F2, using 3.5kPa as target fibrosis. Assessment of the methodological quality of the incorporated papers by the QUADAS-2 tool for pSWE and MRE. A total 2,153 studies articles were evaluated and 44 studies, comprising 6,081 patients with individual data, were included in the meta-analysis: 28 studies for pSWE and 16 studies for MRE. The pooled sensitivity and specificity were 0.86 (95%CI 0.80-0.90) and 0.88 (95%CI 0.85-0.91), respectively, for pSWE, compared with 0.94 (95%CI 0.89-0.97) and 0.95 (95%CI 0.89-0.98) respectively, for MRE. The pooled SROC curve for pSWE shows in the area under the curve (AUC) of 0.93 (95%CI 0.90-0.95), whereas the AUC for MRE was 0.98 (95%CI 0.96-0.99). The diagnostic odds ratio for pSWE and MRE were 41 (95%CI 24-72) and 293 (95%CI 86-1000), respectively. There was statistically significant heterogeneity for pSWE sensitivity (I² = 85.26, P<0.001) and specificity (I² = 89.46, P<0.001). The heterogeneity for MRE also was significant for sensitivity (I² = 73.28, P<0.001) and specificity (I² = 87.24, P<0.001). Therefore, both pSWE and MRE are suitable modalities for assessing liver fibrosis. In addition, MRE is a more accurate imaging technique than pSWE and can be used as alternative to invasive biopsy.

Fluoroscopia por tomografia computadorizada - biópsia percutânea guiada de nódulos pulmonares ≤ 10 mm: análise retrospectiva de procedimentos realizados no período de pandemia de COVID-19

Resumo Objetivo: Avaliar o desempenho diagnóstico da biópsia pulmonar percutânea transtorácica (BPPT) guiada por fluoroscopia associada a tomografia computadorizada (FTC) em nódulos pulmonares ≤ 10 mm no período de pandemia de COVID-19. Materiais e Métodos: No período de 1º de janeiro de 2020 a 30 de abril de 2022, 359 BPPTs guiadas por FTC foram realizadas em um centro terciário de radiologia intervencionista. As lesões pulmonares mediam entre 2 mm e 108 mm. Dessas 359 BPPTs, 27 (7,5%) foram realizadas com agulha 18G em nódulos de 2 mm a 10 mm. Resultados: Das 27 BPPTs realizadas nos nódulos ≤ 10 mm, quatro lesões tinham dimensões menores que 5 mm e 23 lesões mediam entre 5 e 10 mm. Sensibilidade e acurácia diagnóstica das BPPTs guiadas por FTC foram de 100% e 92,3%, respectivamente. A dose média de radiação ionizante para os pacientes durante o procedimento de BPPT guiada por FTC foi de 581,33 mGy*cm, variando de 303 a 1129 mGy*cm. A média de tempo dos procedimentos de biópsia foi de 6,6 minutos, variando de 2 a 12 minutos. Nas 27 BPPTs, nenhuma complicação maior foi descrita. Conclusão: A BBPT guiada por FTC resultou em alto rendimento diagnóstico e baixas taxas de complicações.

CT imaging findings in lung transplant recipients with COVID-19

OBJECTIVES

Our goal was to compare the chest computed tomography (CT) imaging findings of COVID-19 in lung transplant recipients (LTR) and a group of non-transplanted controls (NTC).

METHODS

This retrospective study included 51 consecutive LTR hospitalized with COVID-19 from two centers. A total of 75 NTC were included for comparison. Images were classified regarding the standardized RSNA category, main pattern of lung attenuation, and longitudinal and axial distribution. Quantitative CT (QCT) analysis was performed to evaluate percentage of high attenuation areas (%HAA, threshold -250 to -700 HU). CT scoring was used to measure severity of parenchymal abnormalities.

RESULTS

The imaging findings of COVID-19 in LTR were significantly different from controls regarding the RSNA classification and pattern of lung attenuation. LTR had a significantly higher proportion of patients with an indeterminate pattern on CT (0.31 vs. 0.11, p = 0.014). The most frequent pattern of attenuation in LTR was predominantly consolidation (0.39 vs. 0.22, p = 0.144) followed by a mixed pattern of ground-glass opacities (GGO) and consolidation (0.37 vs. 0.20, adjusted p = 0.102). On the other hand, the most common pattern in NTC was GGO predominant (0.58 vs. 0.24 of LTR, p = 0.001). LTR had significantly more severe parenchymal disease measured by CT score and %HAA by QCT (0.372 ± 0.08 vs. 0.148 ± 0.06, p < 0.001).

CONCLUSION

The most frequent finding of COVID-19 in LTR is a predominant pattern of consolidation. Compared to NTC, LTR more frequently demonstrated an indeterminate pattern according to the RSNA classification and more extensive lung abnormalities on QCT and semi-quantitative scoring.

KEY POINTS

• The most common CT finding of COVID-19 in LTR is a predominant pattern of consolidation followed by a mixed pattern of GGO and consolidation, while controls more often have a predominant pattern of GGO. • LTR more often presents with an indeterminate pattern of COVID-19 by RSNA classification than controls; therefore, molecular testing for COVID-19 is essential for LTR presenting with lower airway infection independently of imaging findings. • LTR had more extensive disease by semi-quantitative CT score and increased percentage areas of high attenuation on QCT.

Computed tomography fluoroscopy-guided percutaneous biopsy of pulmonary nodules ≤ 10 mm: retrospective analysis of procedures performed during the COVID-19 pandemic

Objective

To evaluate the diagnostic performance of computed tomography (CT) fluoroscopy-guided percutaneous transthoracic needle biopsy (PTNB) in pulmonary nodules ≤ 10 mm during the coronavirus disease 2019 pandemic.

Materials and Methods

Between January 1, 2020 and April 30, 2022, a total of 359 CT fluoroscopy-guided PTNBs were performed at an interventional radiology center. Lung lesions measured between 2 mm and 108 mm. Of the 359 PTNBs, 27 (7.5%) were performed with an 18G core needle on nodules ≤ 10 mm in diameter.

Results

Among the 27 biopsies performed on nodules ≤ 10 mm, the lesions measured < 5 mm in four and 5-10 mm in 23. The sensitivity and overall diagnostic accuracy of PTNB were 100% and 92.3%, respectively. The mean dose of ionizing radiation during PTNB was 581.33 mGy*cm (range, 303-1,129 mGy*cm), and the mean biopsy procedure time was 6.6 min (range, 2-12 min). There were no major postprocedural complications.

Conclusion

CT fluoroscopy-guided PTNB appears to provide a high diagnostic yield with low complication rates.

Narrowing the Differential Diagnosis of Cystic Lesions in Smokers with Expiratory CT Acquisition Using the Cyst-Airway Communication Hypothesis

The aim of this study was to assess percentage respiratory changes (δ) in the size of pulmonary cysts of different smoking-related etiologies. Retrospectively, we measured the cystic lesions due to histopathological-confirmed honeycombing from interstitial pulmonary fibrosis, pulmonary Langerhans cell histiocytosis (PLCH), and paraseptal emphysema, using paired inspiratory and expiratory CT scans. In a sample of 72 patients and 216 lesions, the mean diameter of PLCH and honeycombing decreased during expiration (PLCH, δ = 60.9%; p = 0.001; honeycombing, δ = 47.5%; p = 0.014). Conversely, paraseptal emphysema did not show any changes (δ = 5.2%; p = 0.34). In summary, our results demonstrated that cysts in smokers with PLCH and honeycombing fibrosis get smaller during expiratory CT scans, whereas the size of cystic-like lesions due to paraseptal emphysema and bullae tend to remain constant during respiratory cycles. These results support the hypothesis of cyst-airway communication in some cystic diseases, which could assist in the differential diagnosis in smoking-related lung diseases.

The diagnostic value of magnetic resonance imaging compared to computed tomography in the evaluation of fat-containing thoracic lesions

Intrathoracic fat-containing lesions may arise in the mediastinum, lungs, pleura, or chest wall. While CT can be helpful in the detection and diagnosis of these lesions, it can only do so if the lesions contain macroscopic fat. Furthermore, because CT cannot demonstrate microscopic or intravoxel fat, it can fail to identify and diagnose microscopic fat-containing lesions. MRI, employing spectral and chemical shift fat suppression techniques, can identify both macroscopic and microscopic fat, with resultant enhanced capability to diagnose these intrathoracic lesions non-invasively and without ionizing radiation. This paper aims to review the CT and MRI findings of fat-containing lesions of the chest and describes the fat-suppression techniques utilized in their assessment.

Chest Imaging in Systemic Endemic Mycoses

Endemic fungal infections are responsible for high rates of morbidity and mortality in certain regions of the world. The diagnosis and management remain a challenge, and the reason could be explained by the lack of disease awareness, variability of symptoms, and insidious and often overlooked clinical presentation. Imaging findings are nonspecific and frequently misinterpreted as other more common infectious or malignant diseases. Patient demographics and clinical and travel history are important clues that may lead to a proper diagnosis. The purpose of this paper is to review the presentation and differential diagnosis of endemic mycoses based on the most common chest imaging findings.

Variability of pulmonary nodule volumetry on coronary CT angiograms

This study aims to investigate the variability of pulmonary nodule (PN) volumetry on multiphase coronary CT angiograms (CCTA). Two radiologists reviewed 5973 CCTA scans in this cross-sectional study to detect incidental solid noncalcified PNs measuring between 5 and 8 mm. Each radiologist measured the nodules' diameters and volume, in systole and diastole, using 2 commercially available software packages to analyze PNs. Bland-Altman analysis was applied between different observers, software packages, and cardiac phases. Bland-Altman subanalysis for the systolic and diastolic datasets were also performed. A total of 195 PNs were detected within the inclusion criteria and measured in systole and diastole. Bland-Altman analysis was used to test the variability of volumetry between cardiac phases ([-47.0%; 52.3%]), software packages ([-50.2%; 68.2%]), and observers ([-14.5%; 27.8%]). The inter-observer variability of the systolic and diastolic subsets was [-13.6%; 31.4%] and [-13.9%; 19.7%], respectively. Using diastolic volume measurements, the variability of PN volumetry on CCTA scans is similar to the reported variability of volumetry on low-dose CT scans. Therefore, growth estimation of PNs on CCTA scans could be feasible.

The Normal Lung Index From Quantitative Computed Tomography for the Evaluation of Obstructive and Restrictive Lung Disease

PURPOSE

Our objective was to evaluate whether the normal lung index (NLI) from quantitative computed tomography (QCT) analysis can be used to predict mortality as well as pulmonary function tests (PFTs) in patients with chronic obstructive pulmonary disease (COPD) and interstitial lung disease (ILD).

MATERIALS AND METHODS

Normal subjects (n=20) and patients with COPD (n=172) and ILD (n=114) who underwent PFTs and chest CT were enrolled retrospectively in this study. QCT measures included the NLI, defined as the ratio of the lung with attenuation between -950 and -700 Hounsfield units (HU) over the total lung volume (-1024 to -250 HU, mL), high-attenuation area (-700 to -250 HU, %), emphysema index (>6% of pixels < -950 HU), skewness, kurtosis, and mean lung attenuation. Coefficients of correlation between QCT measurements and PFT results in all subjects were calculated. Univariate and multivariate survival analyses were performed to assess mortality prediction by disease.

RESULTS

The Pearson correlation analysis showed that the NLI correlated moderately with the forced expiratory volume in 1 second in subjects with COPD (r=0.490, P<0.001) and the forced vital capacity in subjects with ILD (r=0.452, P<0.001). Multivariate analysis revealed that the NLI of <70% was a significant independent predictor of mortality in subjects with COPD (hazard ratio=3.14, P=0.034) and ILD (hazard ratio=2.72, P=0.005).

CONCLUSION

QCT analysis, specifically the NLI, can also be used to predict mortality in individuals with COPD and ILD.

Personalised Lung Cancer Screening (PLuS) study to assess the importance of coexisting chronic conditions to clinical practice and policy: protocol for a multicentre observational study

INTRODUCTION

Lung cancer is the leading cause of cancer death in the USA and worldwide, and lung cancer screening (LCS) with low-dose CT (LDCT) has the potential to improve lung cancer outcomes. A critical question is whether the ratio of potential benefits to harms found in prior LCS trials applies to an older and potentially sicker population. The Personalised Lung Cancer Screening (PLuS) study will help close this knowledge gap by leveraging real-world data to fully characterise LCS recipients. The principal goal of the PLuS study is to characterise the comorbidity burden of individuals undergoing LCS and quantify the benefits and harms of LCS to enable informed decision-making.

METHODS AND ANALYSIS

PLuS is a multicentre observational study designed to assemble an LCS cohort from the electronic health records of ~40 000 individuals undergoing annual LCS with LDCT from 2016 to 2022. Data will be integrated into a unified repository to (1) examine the burden of multimorbidity by race/ethnicity, socioeconomic status and age; (2) quantify potential benefits and harms; and (3) use the observational data with validated simulation models in the Cancer Intervention and Surveillance Modeling Network (CISNET) to provide LCS outcomes in the real-world US population. We will fit a multivariable logistic regression model to estimate the adjusted ORs of comorbidity, functional limitations and impaired pulmonary function adjusted for relevant covariates. We will also estimate the cumulative risk of LCS outcomes using discrete-time survival models. To our knowledge, this is the first study to combine observational data and simulation models to estimate the long-term impact of LCS with LDCT.

ETHICS AND DISSEMINATION

The study was approved by the Kaiser Permanente Southern California Institutional Review Board and VA Portland Health Care System. The results will be disseminated through publications and presentations at national and international conferences. Safety considerations include protection of patient confidentiality.

Incidental chest findings on coronary CT angiography: a pictorial essay and management proposal

Many health systems have been using coronary CT angiography (CCTA) as a first-line examination for ischaemic heart disease patients in various countries. The rising number of CCTA examinations has led to a significant increase in the number of reported incidental extracardiac findings, mainly in the chest. Pulmonary nodules are the most common incidental findings on CCTA scans, as there is a substantial overlap of risk factors between the population seeking to exclude ischaemic heart disease and those at risk of developing lung cancer (i.e., advanced age and smoking habits). However, most incidental findings are clinically insignificant and actively pursuing them could be cost-prohibitive and submit the patient to unnecessary and potentially harmful examinations. Furthermore, there is little consensus regarding when to report or actively exclude these findings and how to manage them, that is, when to trigger an alert or to immediately refer the patient to a pulmonologist, a thoracic surgeon or a multidisciplinary team. This pictorial essay discusses the current literature on this topic and is illustrated with a review of CCTA scans. We also propose a checklist organised by organ and system, recommending actions to raise awareness of pulmonologists, thoracic surgeons, cardiologists and radiologists regarding the most significant and actionable incidental findings on CCTA scans.