Using AI to Identify Unremarkable Chest Radiographs for Automatic Reporting.

Background Radiology practices have a high volume of unremarkable chest radiographs and artificial intelligence (AI) could possibly improve workflow by providing an automatic report. Purpose To estimate the proportion of unremarkable chest radiographs, where AI can correctly exclude pathology (ie, specificity) without increasing diagnostic errors. Materials and Methods In this retrospective study, consecutive chest radiographs in unique adult patients (≥18 years of age) were obtained January 1-12, 2020, at four Danish hospitals. Exclusion criteria included insufficient radiology reports or AI output error. Two thoracic radiologists, who were blinded to AI output, labeled chest radiographs as "remarkable" or "unremarkable" based on predefined unremarkable findings (reference standard). Radiology reports were classified similarly. A commercial AI tool was adapted to output a chest radiograph "remarkableness" probability, which was used to calculate specificity at different AI sensitivities. Chest radiographs with missed findings by AI and/or the radiology report were graded by one thoracic radiologist as critical, clinically significant, or clinically insignificant. Paired proportions were compared using the McNemar test. Results A total of 1961 patients were included (median age, 72 years [IQR, 58-81 years]; 993 female), with one chest radiograph per patient. The reference standard labeled 1231 of 1961 chest radiographs (62.8%) as remarkable and 730 of 1961 (37.2%) as unremarkable. At 99.9%, 99.0%, and 98.0% sensitivity, the AI had a specificity of 24.5% (179 of 730 radiographs [95% CI: 21, 28]), 47.1% (344 of 730 radiographs [95% CI: 43, 51]), and 52.7% (385 of 730 radiographs [95% CI: 49, 56]), respectively. With the AI fixed to have a similar sensitivity as radiology reports (87.2%), the missed findings of AI and reports had 2.2% (27 of 1231 radiographs) and 1.1% (14 of 1231 radiographs) classified as critical (P = .01), 4.1% (51 of 1231 radiographs) and 3.6% (44 of 1231 radiographs) classified as clinically significant (P = .46), and 6.5% (80 of 1231) and 8.1% (100 of 1231) classified as clinically insignificant (P = .11), respectively. At sensitivities greater than or equal to 95.4%, the AI tool exhibited less than or equal to 1.1% critical misses. Conclusion A commercial AI tool used off-label could correctly exclude pathology in 24.5%-52.7% of all unremarkable chest radiographs at greater than or equal to 98% sensitivity. The AI had equal or lower rates of critical misses than radiology reports at sensitivities greater than or equal to 95.4%. These results should be confirmed in a prospective study. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Yoon and Hwang in this issue.

DCE-CT parameters as new functional imaging biomarkers at baseline and during immune checkpoint inhibitor therapy in patients with lung cancer - a feasibility study.

BACKGROUND: With the development of immune checkpoint inhibitors for the treatment of non-small cell lung cancer, the need for new functional imaging techniques and early response assessments has increased to account for new response patterns and the high cost of treatment. The present study was designed to assess the prognostic impact of dynamic contrast-enhanced computed tomography (DCE-CT) on survival outcomes in non-small cell lung cancer patients treated with immune checkpoint inhibitors. METHODS: Thirty-three patients with inoperable non-small-cell lung cancer treated with immune checkpoint inhibitors were prospectively enrolled for DCE-CT as part of their follow-up. A single target lesion at baseline and subsequent follow-up examinations were enclosed in the DCE-CT. Blood volume deconvolution (BVdecon), blood flow deconvolution (BFdecon), blood flow maximum slope (BFMax slope) and permeability were assessed using overall survival (OS) and progression-free survival (PFS) as endpoints in Kaplan Meier and Cox regression analyses. RESULTS: High baseline Blood Volume (BVdecon) (> 12.97 ml × 100 g-1) was associated with a favorable OS (26.7 vs 7.9 months; p = 0.050) and PFS (14.6 vs 2.5 months; p = 0.050). At early follow-up on day seven a higher relative increase in BFdecon (> 24.50% for OS and > 12.04% for PFS) was associated with an unfavorable OS (8.7 months vs 23.1 months; p < 0.025) and PFS (2.5 vs 13.7 months; p < 0.018). The relative change in BFdecon (categorical) on day seven was a predictor of OS (HR 0.26, CI95: 0.06 to 0.93 p = 0.039) and PFS (HR 0.27, CI95: 0.09 to 0.85 p = 0.026). CONCLUSION: DCE-CT-identified parameters may serve as potential prognostic biomarkers at baseline and during early treatment in patients with NSCLC treated with immune checkpoint inhibitor therapy.

Occupational exposure to respirable crystalline silica and incident idiopathic interstitial pneumonias and pulmonary sarcoidosis: a national prospective follow-up study.

BACKGROUND: Respirable crystalline silica is a well-known cause of silicosis but may also be associated with other types of interstitial lung disease. We examined the associations between occupational exposure to respirable crystalline silica and the risk of idiopathic interstitial pneumonias, pulmonary sarcoidosis and silicosis. METHODS: The total Danish working population was followed 1977-2015. Annual individual exposure to respirable crystalline silica was estimated using a quantitative job exposure matrix. Cases were identified in the Danish National Patient Register. We conducted adjusted analyses of exposure-response relations between cumulative silica exposure and other exposure metrics and idiopathic interstitial pneumonias, pulmonary sarcoidosis and silicosis. RESULTS: Mean cumulative exposure was 125 µg/m3-years among exposed workers. We observed increasing incidence rate ratios with increasing cumulative silica exposure for idiopathic interstitial pneumonias, pulmonary sarcoidosis and silicosis. For idiopathic interstitial pneumonias and pulmonary sarcoidosis, trends per 50 µg/m3-years were 1.03 (95% CI 1.02 to 1.03) and 1.06 (95% CI 1.04 to 1.07), respectively. For silicosis, we observed the well-known exposure-response relation with a trend per 50 µg/m3-years of 1.20 (95% CI 1.17 to 1.23). CONCLUSION: This study suggests that silica inhalation may be related to pulmonary sarcoidosis and idiopathic interstitial pneumonias, though these findings may to some extent be explained by diagnostic misclassification. The observed exposure-response relations for silicosis at lower cumulative exposure levels than previously reported need to be corroborated in analyses that address the limitations of this study.

Risk of hypersensitivity pneumonitis and other interstitial lung diseases following organic dust exposure.

BACKGROUND: Organic dust is associated with hypersensitivity pneumonitis, and associations with other types of interstitial lung disease (ILD) have been suggested. We examined the association between occupational organic dust exposure and hypersensitivity pneumonitis and other ILDs in a cohort study. METHODS: The study population included all residents of Denmark born in 1956 or later with at least 1 year of gainful employment since 1976. Incident cases of hypersensitivity pneumonitis and other ILDs were identified in the Danish National Patient Register 1994-2015. Job exposure matrices were used to assign individual annual levels of exposure to organic dust, endotoxin and wood dust from 1976 to 2015. We analysed exposure-response relations by different exposure metrics using a discrete-time hazard model. RESULTS: For organic dust, we observed increasing risk with increasing cumulative exposure with incidence rate ratios (IRR) per 10 unit-years of 1.19 (95% CI 1.12 to 1.27) for hypersensitivity pneumonitis and 1.04 (95% CI 1.02 to 1.06) for other ILDs. We found increasing risk with increasing cumulative endotoxin exposure for hypersensitivity pneumonitis and other ILDs with IRRs per 5000 endotoxin units/m3-years of 1.55 (95% CI 1.38 to 1.73) and 1.09 (95% CI 1.00 to 1.19), respectively. For both exposures, risk also increased with increasing duration of exposure and recent exposure. No increased risks were observed for wood dust exposure. CONCLUSION: Exposure-response relations were observed between organic dust and endotoxin exposure and hypersensitivity pneumonitis and other ILDs, with lower risk estimates for the latter. The findings indicate that organic dust should be considered a possible cause of any ILD. TRIAL REGISTRATION NUMBER: j.no.: 1-16-02-196-17.

The asbestos-asbestosis exposure-response relationship: a cohort study of the general working population.

OBJECTIVES: The association between asbestos exposure and asbestosis in high-exposed industrial cohorts is well-known, but there is a lack of knowledge about the exposure-response relationship for asbestosis in a general working population setting. We examined the exposure-response relationship between occupational asbestos exposure and asbestosis in asbestos-exposed workers of the Danish general working population. METHODS: We followed all asbestos-exposed workers from 1979 to 2015 and identified incident cases of asbestosis using the Danish National Patient Register. Individual asbestos exposure was estimated with a quantitative job exposure matrix (SYN-JEM) from 1976 onwards and back-extrapolated to age 16 for those exposed in 1976. Exposure-response relations for cumulative exposure and other exposure metrics were analyzed using a discrete time hazard model and adjusted for potential confounders. RESULTS: The range of cumulative exposure in the population was 0.001 to 18 fibers per milliliter-year (f/ml-year). We found increasing incidence rate ratios (IRR) of asbestosis with increasing cumulative asbestos exposure with a fully adjusted IRR per 1 f/ml-years of 1.18 [95% confidence interval (CI) 1.15- -1.22]. The IRR was 1.94 (95% CI 1.53-2.47) in the highest compared to the lowest exposure tertile. We similarly observed increasing risk with increasing cumulative exposure in the inception population. CONCLUSIONS: This study found exposure-response relations between cumulative asbestos exposure and incident asbestosis in the Danish general working population with mainly low-level exposed occupations, but there is some uncertainty regarding the exposure levels.

Commercially Available Chest Radiograph AI Tools for Detecting Airspace Disease, Pneumothorax, and Pleural Effusion.

Background Commercially available artificial intelligence (AI) tools can assist radiologists in interpreting chest radiographs, but their real-life diagnostic accuracy remains unclear. Purpose To evaluate the diagnostic accuracy of four commercially available AI tools for detection of airspace disease, pneumothorax, and pleural effusion on chest radiographs. Materials and Methods This retrospective study included consecutive adult patients who underwent chest radiography at one of four Danish hospitals in January 2020. Two thoracic radiologists (or three, in cases of disagreement) who had access to all previous and future imaging labeled chest radiographs independently for the reference standard. Area under the receiver operating characteristic curve, sensitivity, and specificity were calculated. Sensitivity and specificity were additionally stratified according to the severity of findings, number of findings on chest radiographs, and radiographic projection. The χ2 and McNemar tests were used for comparisons. Results The data set comprised 2040 patients (median age, 72 years [IQR, 58-81 years]; 1033 female), of whom 669 (32.8%) had target findings. The AI tools demonstrated areas under the receiver operating characteristic curve ranging 0.83-0.88 for airspace disease, 0.89-0.97 for pneumothorax, and 0.94-0.97 for pleural effusion. Sensitivities ranged 72%-91% for airspace disease, 63%-90% for pneumothorax, and 62%-95% for pleural effusion. Negative predictive values ranged 92%-100% for all target findings. In airspace disease, pneumothorax, and pleural effusion, specificity was high for chest radiographs with normal or single findings (range, 85%-96%, 99%-100%, and 95%-100%, respectively) and markedly lower for chest radiographs with four or more findings (range, 27%-69%, 96%-99%, 65%-92%, respectively) (P < .001). AI sensitivity was lower for vague airspace disease (range, 33%-61%) and small pneumothorax or pleural effusion (range, 9%-94%) compared with larger findings (range, 81%-100%; P value range, > .99 to < .001). Conclusion Current-generation AI tools showed moderate to high sensitivity for detecting airspace disease, pneumothorax, and pleural effusion on chest radiographs. However, they produced more false-positive findings than radiology reports, and their performance decreased for smaller-sized target findings and when multiple findings were present. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Yanagawa and Tomiyama in this issue.

Late adverse events to iodinated contrast media in patients treated with IL-2: a safety report from the Danish Renal Carcinoma Group (DaRenCa) study - 1.

BACKGROUND: A higher incidence of late adverse events (LAEs) to iodinated contrast media in interleukin-2 (IL-2)-treated patients has been reported. PURPOSE: To assess the incidence of LAEs after administration of iodinated contrast media in patients with metastatic renal cell carcinoma (mRCC) treated with IL-2. MATERIAL AND METHODS: Patients were randomized to treatment with IL-2 and interferon-α with/without bevacizumab in the Danish Renal Carcinoma Group study - 1. Patients underwent a computed tomography (CT) scan at baseline, at one month, at three months, and every third month until RECIST 1.1 defined progression. LAEs due to iodinated contrast media were systematically registered according to the Common Terminology Criteria for Adverse Events classification. RESULTS: In total, 89 patients were included and underwent a total of 507 contrast-enhanced CT scans. An overall incidence of 46 (9.1%) LAEs was observed in 38 of 89 (42.7%) patients; 3 LAEs at baseline (3.4% of all baseline scans), 39 (13.9%) LAEs during IL-2-based therapies, and 4 (2.9%) LAEs after termination of IL-based therapies. There was no difference in progression-free survival, overall survival, and treatment response in patients experiencing LAEs compared to patients without LAEs (P = 0.2, P = 0.5, and P = 0.6, respectively). CONCLUSION: Patients with mRCC demonstrated a higher incidence of LAEs after administration of iodinated contrast during ongoing IL-2 therapy, indicating that iodinated contrast media may cause a recall phenomenon of IL-2 toxicities in patients with mRCC. Treatment with IL-2 should not be a contraindication for contrast-enhanced scans in patients with mRCC but expertise and vigilance are required.

Association between Parenchymal B-Cell Follicles and CT-Proven Enlarged Mediastinal Lymph Nodes in Patients with Pulmonary Fibrosis.

BACKGROUND: Mediastinal lymph node enlargement (MLNE) is a finding described in a subset of patients with idiopathic pulmonary fibrosis (IPF) and other interstitial lung diseases (ILDs) and is associated with accelerated disease progression and increased mortality. The cause of MLNE is still not known. Our hypothesis is that there is an association between MLNE and B-cell follicles in lung tissue, another aspect detectable in the lung tissue of patients with IPF and other ILDs. OBJECTIVES: The aim of this study was to assess if there is an association between MLNE and B-cell follicles in lung tissue in patients with IPF and other ILDs. METHOD: Patients having transbronchial cryobiopsies performed as part of an investigation for ILD were included in this prospective observational study. MLNE (smallest diameter ≥10 mm) were assessed in station 7, 4R, and 4L on high-resolution computed tomography scans. B-cell follicles were assessed on haematoxylin-eosin-stained specimens. Lung function, 6-minute walk test, acute exacerbation, and mortality were registered after 2 years. In addition, we investigated if the finding of B-cell follicles was consistent in patients who underwent both surgical lung biopsies (SLBs) and cryobiopsies. RESULTS: In total, 93 patients were included for analysis (46% diagnosed with IPF, 54% diagnosed with other ILDs). MLNE was found in 26 (60%) of the IPF patients and in 23 (46%) of the non-IPF patients (p = 0.164). Diffusing capacity for carbon monoxide was significantly lower (p = 0.03) in patients with MLNE compared to patients without MLNE. B-cell follicles were found in 11 (26%) of the IPF patients and in 22 (44%) of the non-IPF patients (p = 0.064). Germinal centres were not seen in any of the patients. There was no association between MLNE and B-cell follicles (p = 0.057). No significant difference in change of pulmonary function test was seen at 2-year follow-up when comparing the patients with and without MLNE or B-cell follicles. In 13 patients, both SLBs and cryobiopsies were performed. The presence of B-cell follicles was not consistent when comparing the two different methods. CONCLUSION: MLNE is evident in a substantial part of patients with ILD and is associated with lower DLCO at inclusion. We could not demonstrate an association between histological B-cell follicles in biopsies and MLNE. A possible explanation for this is that the cryobiopsies might not have captured the changes we sought.

Autonomous Chest Radiograph Reporting Using AI: Estimation of Clinical Impact.

Background Automated interpretation of normal chest radiographs could alleviate the workload of radiologists. However, the performance of such an artificial intelligence (AI) tool compared with clinical radiology reports has not been established. Purpose To perform an external evaluation of a commercially available AI tool for (a) the number of chest radiographs autonomously reported, (b) the sensitivity for AI detection of abnormal chest radiographs, and (c) the performance of AI compared with that of the clinical radiology reports. Materials and Methods In this retrospective study, consecutive posteroanterior chest radiographs from adult patients in four hospitals in the capital region of Denmark were obtained in January 2020, including images from emergency department patients, in-hospital patients, and outpatients. Three thoracic radiologists labeled chest radiographs in a reference standard based on chest radiograph findings into the following categories: critical, other remarkable, unremarkable, or normal (no abnormalities). AI classified chest radiographs as high confidence normal (normal) or not high confidence normal (abnormal). Results A total of 1529 patients were included for analysis (median age, 69 years [IQR, 55-69 years]; 776 women), with 1100 (72%) classified by the reference standard as having abnormal radiographs, 617 (40%) as having critical abnormal radiographs, and 429 (28%) as having normal radiographs. For comparison, clinical radiology reports were classified based on the text and insufficient reports excluded (n = 22). The sensitivity of AI was 99.1% (95% CI: 98.3, 99.6; 1090 of 1100 patients) for abnormal radiographs and 99.8% (95% CI: 99.1, 99.9; 616 of 617 patients) for critical radiographs. Corresponding sensitivities for radiologist reports were 72.3% (95% CI: 69.5, 74.9; 779 of 1078 patients) and 93.5% (95% CI: 91.2, 95.3; 558 of 597 patients), respectively. Specificity of AI, and hence the potential autonomous reporting rate, was 28.0% of all normal posteroanterior chest radiographs (95% CI: 23.8, 32.5; 120 of 429 patients), or 7.8% (120 of 1529 patients) of all posteroanterior chest radiographs. Conclusion Of all normal posteroanterior chest radiographs, 28% were autonomously reported by AI with a sensitivity for any abnormalities higher than 99%. This corresponded to 7.8% of the entire posteroanterior chest radiograph production. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Park in this issue.

Potential impact of texture analysis in contrast enhanced CT in non-small cell lung cancer as a marker of survival: A retrospective feasibility study.

The objective of this feasibility study was to assess computed tomography (CT) texture analysis (CTTA) of pulmonary lesions as a predictor of overall survival in patients with suspected lung cancer on contrast-enhanced computed tomography (CECT). In a retrospective pilot study, 94 patients (52 men and 42 women; mean age, 67.2 ±â€…10.8 yrs) from 1 center with non-small cell lung cancer (NSCLC) underwent CTTA on the primary lesion by 2 individual readers. Both simple and multivariate Cox regression analyses correlating textural parameters with overall survival were performed. Statistically significant parameters were selected, and optimal cutoff values were determined. Kaplan-Meier plots based on these results were produced. Simple Cox regression analysis showed that normalized uniformity had a hazard ratio (HR) of 16.059 (3.861-66.788, P < .001), and skewness had an HR of 1.914 (1.330-2.754, P < .001). The optimal cutoff values for both parameters were 0.8602 and 0.1554, respectively. Normalized uniformity, clinical stage, and skewness were found to be prognostic factors for overall survival in multivariate analysis. Tumor heterogeneity, assessed by normalized uniformity and skewness on CECT may be a prognostic factor for overall survival.