Comparison of Commercial AI Software Performance for Radiograph Lung Nodule Detection and Bone Age Prediction.

Background Multiple commercial artificial intelligence (AI) products exist for assessing radiographs; however, comparable performance data for these algorithms are limited. Purpose To perform an independent, stand-alone validation of commercially available AI products for bone age prediction based on hand radiographs and lung nodule detection on chest radiographs. Materials and Methods This retrospective study was carried out as part of Project AIR. Nine of 17 eligible AI products were validated on data from seven Dutch hospitals. For bone age prediction, the root mean square error (RMSE) and Pearson correlation coefficient were computed. The reference standard was set by three to five expert readers. For lung nodule detection, the area under the receiver operating characteristic curve (AUC) was computed. The reference standard was set by a chest radiologist based on CT. Randomized subsets of hand (n = 95) and chest (n = 140) radiographs were read by 14 and 17 human readers, respectively, with varying experience. Results Two bone age prediction algorithms were tested on hand radiographs (from January 2017 to January 2022) in 326 patients (mean age, 10 years ± 4 [SD]; 173 female patients) and correlated strongly with the reference standard (r = 0.99; P < .001 for both). No difference in RMSE was observed between algorithms (0.63 years [95% CI: 0.58, 0.69] and 0.57 years [95% CI: 0.52, 0.61]) and readers (0.68 years [95% CI: 0.64, 0.73]). Seven lung nodule detection algorithms were validated on chest radiographs (from January 2012 to May 2022) in 386 patients (mean age, 64 years ± 11; 223 male patients). Compared with readers (mean AUC, 0.81 [95% CI: 0.77, 0.85]), four algorithms performed better (AUC range, 0.86-0.93; P value range, <.001 to .04). Conclusions Compared with human readers, four AI algorithms for detecting lung nodules on chest radiographs showed improved performance, whereas the remaining algorithms tested showed no evidence of a difference in performance. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Omoumi and Richiardi in this issue.

Performance of AI to exclude normal chest radiographs to reduce radiologists' workload.

INTRODUCTION: This study investigates the performance of a commercially available artificial intelligence (AI) system to identify normal chest radiographs and its potential to reduce radiologist workload. METHODS: Retrospective analysis included consecutive chest radiographs from two medical centers between Oct 1, 2016 and Oct 14, 2016. Exclusions comprised follow-up exams within the inclusion period, bedside radiographs, incomplete images, imported radiographs, and pediatric radiographs. Three chest radiologists categorized findings into normal, clinically irrelevant, clinically relevant, urgent, and critical. A commercial AI system processed all radiographs, scoring 10 chest abnormalities on a 0-100 confidence scale. AI system performance was evaluated using the area under the ROC curve (AUC), assessing the detection of normal radiographs. Sensitivity was calculated for the default and a conservative operating point. the detection of negative predictive value (NPV) for urgent and critical findings, as well as the potential workload reduction, was calculated. RESULTS: A total of 2603 radiographs were acquired in 2141 unique patients. Post-exclusion, 1670 radiographs were analyzed. Categories included 479 normal, 332 clinically irrelevant, 339 clinically relevant, 501 urgent, and 19 critical findings. The AI system achieved an AUC of 0.92. Sensitivity for normal radiographs was 92% at default and 53% at the conservative operating point. At the conservative operating point, NPV was 98% for urgent and critical findings, and could result in a 15% workload reduction. CONCLUSION: A commercially available AI system effectively identifies normal chest radiographs and holds the potential to lessen radiologists' workload by omitting half of the normal exams from reporting. CLINICAL RELEVANCE STATEMENT: The AI system is able to detect half of all normal chest radiographs at a clinically acceptable operating point, thereby potentially reducing the workload for the radiologists by 15%. KEY POINTS: The AI system reached an AUC of 0.92 for the detection of normal chest radiographs. Fifty-three percent of normal chest radiographs were identified with a NPV of 98% for urgent findings. AI can reduce the workload of chest radiography reporting by 15%.

Influence of device-assisted suction against resistance (Mueller maneuver) on image quality in CTPA for suspected lung embolism.

OBJECTIVES: To investigate the effect of a device-assisted suction against resistance Mueller maneuver (MM) on transient interruption of contrast (TIC) in the aorta and pulmonary trunk (PT) on computed tomography pulmonary angiogram (CTPA). METHODS: In this prospective single-center study, 150 patients with suspected pulmonary artery embolism were assigned randomly with two different breathing maneuvers (Mueller maneuver (MM) or standard end-inspiratory breath-hold command (SBC)) during routine CTPA. The MM was performed using a patented prototype (Contrast Booster™) which allows both the patient by means of visual feedback and the medical staff in the CT scanning room to monitor whether the patient is sucking sufficiently or not. Mean Hounsfield attenuation in descending aorta and PT was measured and compared. RESULTS: Overall, patients with MM showed an attenuation of 338.24 HU in the pulmonary trunk, compared to 313.71 HU in SBC (p = 0.157). In the aorta, the values for MM were lower compared to SBC (134.42 HU vs. 177.83 HU, p = 0.001). The TP-aortic ratio was significantly higher in the MM group at 3.86 compared to the SBC group at 2.26, p = 0.001. TIC phenomenon was absent in the MM group, whereas it was present in 9 patients (12.3%) in the SBC group (p = 0.005). Overall contrast was better on all levels for MM (p < 0.001). The presence of breathing artifacts was higher in the MM group (48.1% vs. 30.1%, p = 0.038), without clinical consequence. CONCLUSIONS: Performing the MM with the application of the prototype is an effective way of preventing the TIC phenomenon during i.v. contrast-enhanced CTPA scanning compared to the standard end-inspiratory breathing command. CLINICAL RELEVANCE: Compared to standard end-inspiratory breathing command, the device-assisted Mueller maneuver (MM) improves contrast enhancement and prevents the transient interruption of contrast (TIC) phenomenon in CTPA. Therefore, it may offer optimized diagnostic workup and timely treatment for patients with pulmonary embolism. KEY POINTS: • Transient interruption of contrast (TIC) may impair image quality in CTPA. • Mueller Maneuver using a device prototype could lower the rate of TIC. • Device application in clinical routine may increase diagnostic accuracy.

Automated Assessment of COVID-19 Reporting and Data System and Chest CT Severity Scores in Patients Suspected of Having COVID-19 Using Artificial Intelligence.

Background The coronavirus disease 2019 (COVID-19) pandemic has spread across the globe with alarming speed, morbidity, and mortality. Immediate triage of patients with chest infections suspected to be caused by COVID-19 using chest CT may be of assistance when results from definitive viral testing are delayed. Purpose To develop and validate an artificial intelligence (AI) system to score the likelihood and extent of pulmonary COVID-19 on chest CT scans using the COVID-19 Reporting and Data System (CO-RADS) and CT severity scoring systems. Materials and Methods The CO-RADS AI system consists of three deep-learning algorithms that automatically segment the five pulmonary lobes, assign a CO-RADS score for the suspicion of COVID-19, and assign a CT severity score for the degree of parenchymal involvement per lobe. This study retrospectively included patients who underwent a nonenhanced chest CT examination because of clinical suspicion of COVID-19 at two medical centers. The system was trained, validated, and tested with data from one of the centers. Data from the second center served as an external test set. Diagnostic performance and agreement with scores assigned by eight independent observers were measured using receiver operating characteristic analysis, linearly weighted κ values, and classification accuracy. Results A total of 105 patients (mean age, 62 years ± 16 [standard deviation]; 61 men) and 262 patients (mean age, 64 years ± 16; 154 men) were evaluated in the internal and external test sets, respectively. The system discriminated between patients with COVID-19 and those without COVID-19, with areas under the receiver operating characteristic curve of 0.95 (95% CI: 0.91, 0.98) and 0.88 (95% CI: 0.84, 0.93), for the internal and external test sets, respectively. Agreement with the eight human observers was moderate to substantial, with mean linearly weighted κ values of 0.60 ± 0.01 for CO-RADS scores and 0.54 ± 0.01 for CT severity scores. Conclusion With high diagnostic performance, the CO-RADS AI system correctly identified patients with COVID-19 using chest CT scans and assigned standardized CO-RADS and CT severity scores that demonstrated good agreement with findings from eight independent observers and generalized well to external data. © RSNA, 2020 Supplemental material is available for this article.

CO-RADS: A Categorical CT Assessment Scheme for Patients Suspected of Having COVID-19-Definition and Evaluation.

Background A categorical CT assessment scheme for suspicion of pulmonary involvement of coronavirus disease 2019 (COVID-19 provides a basis for gathering scientific evidence and improved communication with referring physicians. Purpose To introduce the COVID-19 Reporting and Data System (CO-RADS) for use in the standardized assessment of pulmonary involvement of COVID-19 on unenhanced chest CT images and to report its initial interobserver agreement and performance. Materials and Methods The Dutch Radiological Society developed CO-RADS based on other efforts for standardization, such as the Lung Imaging Reporting and Data System or Breast Imaging Reporting and Data System. CO-RADS assesses the suspicion for pulmonary involvement of COVID-19 on a scale from 1 (very low) to 5 (very high). The system is meant to be used in patients with moderate to severe symptoms of COVID-19. The system was evaluated by using 105 chest CT scans of patients admitted to the hospital with clinical suspicion of COVID-19 and in whom reverse transcription-polymerase chain reaction (RT-PCR) was performed (mean, 62 years ± 16 [standard deviation]; 61 men, 53 with positive RT-PCR results). Eight observers used CO-RADS to assess the scans. Fleiss κ value was calculated, and scores of individual observers were compared with the median of the remaining seven observers. The resulting area under the receiver operating characteristics curve (AUC) was compared with results from RT-PCR and clinical diagnosis of COVID-19. Results There was absolute agreement among observers in 573 (68.2%) of 840 observations. Fleiss κ value was 0.47 (95% confidence interval [CI]: 0.45, 0.47), with the highest κ value for CO-RADS categories 1 (0.58, 95% CI: 0.54, 0.62) and 5 (0.68, 95% CI: 0.65, 0.72). The average AUC was 0.91 (95% CI: 0.85, 0.97) for predicting RT-PCR outcome and 0.95 (95% CI: 0.91, 0.99) for clinical diagnosis. The false-negative rate for CO-RADS 1 was nine of 161 cases (5.6%; 95% CI: 1.0%, 10%), and the false-positive rate for CO-RADS category 5 was one of 286 (0.3%; 95% CI: 0%, 1.0%). Conclusion The coronavirus disease 2019 (COVID-19) Reporting and Data System (CO-RADS) is a categorical assessment scheme for pulmonary involvement of COVID-19 at unenhanced chest CT that performs very well in predicting COVID-19 in patients with moderate to severe symptoms and has substantial interobserver agreement, especially for categories 1 and 5. © RSNA, 2020 Online supplemental material is available for this article.

Predictive factors for treatment-related mortality and major adverse events after autologous haematopoietic stem cell transplantation for systemic sclerosis: results of a long-term follow-up multicentre study.

BACKGROUND: Autologous haematopoietic stem cell transplantation (HSCT) improves survival in systemic sclerosis (SSc) with poor prognosis, but is hampered by treatment-related mortality (TRM). OBJECTIVE: To evaluate event-free survival (EFS), TRM, response to treatment, disease progression and patient characteristics associated with events. METHODS: All patients treated with HSCT for SSc in The Netherlands until 2017 (n=92) were included. Data on skin involvement (modified Rodnan skin score (mRSS), pulmonary function (forced vital capacity (FVC) and diffusion capacity of the lungs for carbon monoxide (DLCO)), extent of interstitial lung disease on high-resolution CT using Goh scores and left ventricular ejection fraction (LVEF) were collected at baseline, 1, 2 and 5 years. Occurrence of events, defined as death or major organ failure, were collected until 2019. As control, a comparison between patients treated with cyclophosphamide (CYC) and patients with HSCT who participated in the Autologous Stem Cell Transplantation International Scleroderma (ASTIS) trial was performed. RESULTS: Median follow-up was 4.6 years. EFS estimates at 5, 10 and 15 years were 78%, 76% and 66%, respectively. Twenty deaths occurred. Mean FVC, DLCO, mRSS and Goh scores all improved significantly. Disease progression occurred in 22 patients. Frequency of TRM decreased over time and occurred more often in males. Events were independently associated with male sex, LVEF <50% and older age. In ASTIS, patients treated with HSCT (n=23) 7 events occurred versus 13 in the CYC group (n=22). CONCLUSION: Our data confirm long-term efficacy of HSCT in improving survival, skin and lung involvement in SSc. Male sex, lower LVEF and older age at baseline were identified as risk factors for events.

Iodine Maps from Subtraction CT or Dual-Energy CT to Detect Pulmonary Emboli with CT Angiography: A Multiple-Observer Study.

Background Dual-energy CT iodine maps are used to detect pulmonary embolism (PE) with CT angiography but require dedicated hardware. Subtraction CT, a software-only solution, results in iodine maps with high contrast-to-noise ratios. Purpose To compare the use of subtraction CT versus dual-energy CT iodine maps to CT angiography for PE detection. Materials and Methods In this prospective study ( https://clinicaltrials.gov , NCT02890706), 274 participants suspected of having PE underwent precontrast CT followed by contrast material-enhanced dual-energy CT angiography between July 2016 and April 2017. Iodine maps from dual-energy CT were derived. Subtraction maps (contrast-enhanced CT minus precontrast CT) were calculated after motion correction. Truth was established by expert consensus. A total of 75 randomly selected participants with and without PE (1:1 ratio) were evaluated by three radiologists and six radiology residents (blinded to final diagnosis) for the presence of PE using three types of CT: CT angiography alone, dual-energy CT, and subtraction CT. The partial area under the receiver operating characteristic curve (AUC) for the clinically relevant specificity region (maximum partial AUC, 0.11) was compared by using multireader multicase variance. A P value less than or equal to .025 was considered indicative of a significant difference due to multiple comparisons. Results There were 35 men and 40 women in the reader study (mean age, 63 years ± 12 [standard deviation]). The pooled sensitivities were not different (P ≥ .31 among techniques) (95% confidence intervals [CIs]: 67%, 89% for CT angiography; 72%, 91% for dual-energy CT; 70%, 91% for subtraction CT). However, pooled specificity was higher for subtraction CT (95% CI: 100%, 100%) than for CT angiography (95% CI: 89%, 97%) or dual-energy CT (95% CI: 89%, 98%) (P < .001). Partial AUCs for the average observer improved equally when adding iodine maps (subtraction CT [0.093] vs CT angiography [0.088], P = .03; dual-energy CT [0.094] vs CT angiography, P = .01; dual-energy CT vs subtraction CT, P = .68). Average reading times were equivalent (range, 97-101 seconds; P ≥ .41) among techniques. Conclusion Subtraction CT iodine maps had greater specificity than CT angiography alone in pulmonary embolism detection. Subtraction CT had comparable diagnostic performance to that of dual-energy CT, without the need for dedicated hardware. © RSNA, 2019 Online supplemental material is available for this article.

Observer variability for Lung-RADS categorisation of lung cancer screening CTs: impact on patient management.

OBJECTIVES: Lung-RADS represents a categorical system published by the American College of Radiology to standardise management in lung cancer screening. The purpose of the study was to quantify how well readers agree in assigning Lung-RADS categories to screening CTs; secondary goals were to assess causes of disagreement and evaluate its impact on patient management. METHODS: For the observer study, 80 baseline and 80 follow-up scans were randomly selected from the NLST trial covering all Lung-RADS categories in an equal distribution. Agreement of seven observers was analysed using Cohen's kappa statistics. Discrepancies were correlated with patient management, test performance and diagnosis of malignancy within the scan year. RESULTS: Pairwise interobserver agreement was substantial (mean kappa 0.67, 95% CI 0.58-0.77). Lung-RADS category disagreement was seen in approximately one-third (29%, 971) of 3360 reading pairs, resulting in different patient management in 8% (278/3360). Out of the 91 reading pairs that referred to scans with a tumour diagnosis within 1 year, discrepancies in only two would have resulted in a substantial management change. CONCLUSIONS: Assignment of lung cancer screening CT scans to Lung-RADS categories achieves substantial interobserver agreement. Impact of disagreement on categorisation of malignant nodules was low. KEY POINTS: • Lung-RADS categorisation of low-dose lung screening CTs achieved substantial interobserver agreement. • Major cause for disagreement was assigning a different nodule as risk-dominant. • Disagreement led to a different follow-up time in 8% of reading pairs.

Imaging of nontraumatic thoracic emergencies.

PURPOSE OF REVIEW: Acute chest symptoms form an important incentive for imaging in the emergency setting. This review discusses the radiologic features of various vascular and pulmonary diseases leading to acute respiratory distress and recent developments on important emergency radiologic examinations. RECENT FINDINGS: Recently, triple-rule-out computed tomography protocol was introduced in diagnosis of chest pain, and advancing computed tomography technology and knowledge have led to discussion on treatment of pulmonary embolism. Diffuse pulmonary opacities remain a diagnostic dilemma in the emergency setting and although imaging findings can often be nonspecific, they help in guiding toward accurate diagnosis and timely management. SUMMARY: Though promising, triple-rule-out is not yet justified because of low incidence of additional findings compared with conventional computed tomography angiography in chest pain, but it might be suited for clinical practice in the near future. Relevance of isolated subsegmental pulmonary embolism is unknown and research on this topic is needed and on its way. We provided some key findings in differentiating diffuse pulmonary opacities and describe the additional value of chest ultrasound in this clinical dilemma. A brief sidestep to pneumothorax is made, as this is also a frequent finding in the acute dyspneic patient, as well as in patients with acute chest pain.

Clofazimine Is a Safe and Effective Alternative for Rifampicin in Mycobacterium avium Complex Pulmonary Disease Treatment: Outcomes of a Randomized Trial.

BACKGROUND: Results of retrospective studies have suggested clofazimine as an alternative for rifampicin in the treatment of Mycobacterium avium complex pulmonary disease (MAC-PD). RESEARCH QUESTION: Is a treatment regimen consisting of clofazimine-ethambutol-macrolide noninferior to the standard treatment regimen (rifampicin-ethambutol-macrolide) in the treatment of MAC-PD? STUDY DESIGN AND METHODS: In this single-center, nonblinded clinical trial, adult patients with MAC-PD were randomly assigned in a 1:1 ratio to receive rifampicin or clofazimine as adjuncts to an ethambutol-macrolide regimen. The primary outcome was sputum culture conversion following 6 months of treatment. RESULTS: Forty patients were assigned to receive either rifampicin (n = 19) or clofazimine (n = 21) in addition to ethambutol and a macrolide. Following 6 months of treatment, both arms showed similar percentages of sputum culture conversion based on an intention-to-treat analysis: 58% (11 of 19) for rifampicin and 62% (13 of 21) for clofazimine. Study discontinuation, mainly due to adverse events, was equal in both arms (26% vs 33%). Based on an on-treatment analysis, sputum culture conversion following 6 months of treatment was 79% in both groups. In the clofazimine arm, diarrhea was more prevalent (76% vs 37%; P = .012), while arthralgia was more frequent in the rifampicin arm (37% vs 5%; P = .011). No difference in the frequency of QTc prolongation was seen between groups. INTERPRETATION: A clofazimine-ethambutol-macrolide regimen showed similar results to the standard rifampicin-ethambutol-macrolide regimen and should be considered in the treatment of MAC-PD. The frequency of adverse events was similar in both arms, but their nature was different. Individual patient characteristics and possible drug-drug interactions should be taken into consideration when choosing an antibiotic regimen for MAC-PD. CLINICAL TRIAL REGISTRATION: EudraCT; No.: 2015-003786-28; URL: https://eudract.ema.europa.eu.

The reliability of magnetic resonance imaging in traumatic brain injury lesion detection.

OBJECTIVE: This study compares inter-rater-reliability, lesion detection and clinical relevance of T2-weighted imaging (T2WI), Fluid Attenuated Inversion Recovery (FLAIR), T2*-gradient recalled echo (T2*-GRE) and Susceptibility Weighted Imaging (SWI) in Traumatic Brain Injury (TBI). METHODS: Three raters retrospectively scored 56 TBI patients' MR images (12-76 years old, median TBI-MRI interval 7 weeks) on number, volume, location and intensity. Punctate lesions (diameter <10 mm) were scored separately from large lesions (diameter ≥ 10 mm). Injury severity was assessed with the Glasgow Coma Scale (GCS), outcome with the Glasgow Outcome Scale-Extended (GOSE). RESULTS: Inter-rater-reliability for lesion volume and punctate lesion count was good (ICC = 0.69-0.94) except for punctate lesion count on T2WI (ICC = 0.19) and FLAIR (ICC = 0.15). SWI showed the highest number of lesions (mean = 30.0), followed by T2*-GRE (mean = 15.4), FLAIR (mean = 3.1) and T2WI (mean = 2.2). Sequences did not differ in detected lesion volume. Punctate lesion count on T2*-GRE (r = -0.53) and SWI (r = -0.49) correlated with the GCS (p < 0.001). CONCLUSIONS: T2*-GRE and SWI are more sensitive than T2WI and FLAIR in detecting (haemorrhagic) traumatic punctate lesions. The correlation between number of punctate lesions on T2*-GRE/SWI and the GCS indicates that haemorrhagic lesions are clinically relevant. The considerable inter-rater-disagreement in this study advocates cautiousness in interpretation of punctate lesions using T2WI and FLAIR.

Assisted versus Manual Interpretation of Low-Dose CT Scans for Lung Cancer Screening: Impact on Lung-RADS Agreement.

Purpose To compare the inter- and intraobserver agreement and reading times achieved when assigning Lung Imaging Reporting and Data System (Lung-RADS) categories to baseline and follow-up lung cancer screening studies by using a dedicated CT lung screening viewer with integrated nodule detection and volumetric support with those achieved by using a standard picture archiving and communication system (PACS)-like viewer. Materials and Methods Data were obtained from the National Lung Screening Trial (NLST). By using data recorded by NLST radiologists, scans were assigned to Lung-RADS categories. For each Lung-RADS category (1 or 2, 3, 4A, and 4B), 40 CT scans (20 baseline scans and 20 follow-up scans) were randomly selected for 160 participants (median age, 61 years; interquartile range, 58-66 years; 61 women) in total. Seven blinded observers independently read all CT scans twice in a randomized order with a 2-week washout period: once by using the standard PACS-like viewer and once by using the dedicated viewer. Observers were asked to assign a Lung-RADS category to each scan and indicate the risk-dominant nodule. Inter- and intraobserver agreement was analyzed by using Fleiss κ values and Cohen weighted κ values, respectively. Reading times were compared by using a Wilcoxon signed rank test. Results The interobserver agreement was moderate for the standard viewer and substantial for the dedicated viewer, with Fleiss κ values of 0.58 (95% CI: 0.55, 0.60) and 0.66 (95% CI: 0.64, 0.68), respectively. The intraobserver agreement was substantial, with a mean Cohen weighted κ value of 0.67. The median reading time was significantly reduced from 160 seconds with the standard viewer to 86 seconds with the dedicated viewer (P < .001). Conclusion Lung-RADS interobserver agreement increased from moderate to substantial when using the dedicated CT lung screening viewer. The median reading time was substantially reduced when scans were read by using the dedicated CT lung screening viewer. Keywords: CT, Thorax, Lung, Computer Applications-Detection/Diagnosis, Observer Performance, Technology Assessment Supplemental material is available for this article. © RSNA, 2021.

Deep Learning for Lung Cancer Detection on Screening CT Scans: Results of a Large-Scale Public Competition and an Observer Study with 11 Radiologists.

PURPOSE: To determine whether deep learning algorithms developed in a public competition could identify lung cancer on low-dose CT scans with a performance similar to that of radiologists. MATERIALS AND METHODS: In this retrospective study, a dataset consisting of 300 patient scans was used for model assessment; 150 patient scans were from the competition set and 150 were from an independent dataset. Both test datasets contained 50 cancer-positive scans and 100 cancer-negative scans. The reference standard was set by histopathologic examination for cancer-positive scans and imaging follow-up for at least 2 years for cancer-negative scans. The test datasets were applied to the three top-performing algorithms from the Kaggle Data Science Bowl 2017 public competition: grt123, Julian de Wit and Daniel Hammack (JWDH), and Aidence. Model outputs were compared with an observer study of 11 radiologists that assessed the same test datasets. Each scan was scored on a continuous scale by both the deep learning algorithms and the radiologists. Performance was measured using multireader, multicase receiver operating characteristic analysis. RESULTS: The area under the receiver operating characteristic curve (AUC) was 0.877 (95% CI: 0.842, 0.910) for grt123, 0.902 (95% CI: 0.871, 0.932) for JWDH, and 0.900 (95% CI: 0.870, 0.928) for Aidence. The average AUC of the radiologists was 0.917 (95% CI: 0.889, 0.945), which was significantly higher than grt123 (P = .02); however, no significant difference was found between the radiologists and JWDH (P = .29) or Aidence (P = .26). CONCLUSION: Deep learning algorithms developed in a public competition for lung cancer detection in low-dose CT scans reached performance close to that of radiologists.Keywords: Lung, CT, Thorax, Screening, Oncology Supplemental material is available for this article. © RSNA, 2021.

Validation, comparison, and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: The LUNA16 challenge.

Automatic detection of pulmonary nodules in thoracic computed tomography (CT) scans has been an active area of research for the last two decades. However, there have only been few studies that provide a comparative performance evaluation of different systems on a common database. We have therefore set up the LUNA16 challenge, an objective evaluation framework for automatic nodule detection algorithms using the largest publicly available reference database of chest CT scans, the LIDC-IDRI data set. In LUNA16, participants develop their algorithm and upload their predictions on 888 CT scans in one of the two tracks: 1) the complete nodule detection track where a complete CAD system should be developed, or 2) the false positive reduction track where a provided set of nodule candidates should be classified. This paper describes the setup of LUNA16 and presents the results of the challenge so far. Moreover, the impact of combining individual systems on the detection performance was also investigated. It was observed that the leading solutions employed convolutional networks and used the provided set of nodule candidates. The combination of these solutions achieved an excellent sensitivity of over 95% at fewer than 1.0 false positives per scan. This highlights the potential of combining algorithms to improve the detection performance. Our observer study with four expert readers has shown that the best system detects nodules that were missed by expert readers who originally annotated the LIDC-IDRI data. We released this set of additional nodules for further development of CAD systems.

The Effect of Supplementary Bone-Suppressed Chest Radiographs on the Assessment of a Variety of Common Pulmonary Abnormalities: Results of an Observer Study.

PURPOSE: The aim of the study was to investigate the effect of bone-suppressed chest radiographs on the detection of common chest abnormalities. MATERIALS AND METHODS: A total of 261 posteroanterior and lateral chest radiographs were collected from 2 hospitals. Radiographs could contain single or multiple focal opacities <3 cm (n=66), single or multiple focal opacities >3 cm (n=33), diffuse lung disease (n=49), signs of cardiogenic congestion (n=26), or no abnormalities (n=110). Twenty-one cases contained >1 type of disease. All abnormalities were confirmed by a computed tomographic scan obtained within 4 weeks of the radiograph. Bone-suppressed images (BSIs) were generated from every posteroanterior radiograph (ClearRead BSI 3.2). All cases were read by 6 radiologists without BSI, followed by an evaluation of the same case with BSI. Presence or absence of each disease category and confidence (0-100) of the observers were documented for each interpretation. Differences in the number of correct detections without and with BSI were analyzed using the Wilcoxon signed-rank test. RESULTS: On average, 6 more cases with focal lesions were correctly identified with BSI (P=0.03), and 1 additional case with diffuse abnormalities was found with BSI (P=0.32). None of the observers demonstrated a decrease in the number of correctly detected cases with diffuse abnormalities or cardiogenic congestion with BSI. False positives in normal cases with availability of BSI mainly referred to the detection of small focal lesions (on average 7 per reader; P=0.04). CONCLUSIONS: BSI does not negatively affect the interpretation of diffuse lung disease, while improving visualization of focal lesions on chest radiographs. BSI leads to overcalling of focal abnormalities in normal radiographs.