Dataset Information

Deep learning predicts cardiovascular disease risks from lung cancer screening low dose computed tomography.

ABSTRACT: Cancer patients have a higher risk of cardiovascular disease (CVD) mortality than the general population. Low dose computed tomography (LDCT) for lung cancer screening offers an opportunity for simultaneous CVD risk estimation in at-risk patients. Our deep learning CVD risk prediction model, trained with 30,286 LDCTs from the National Lung Cancer Screening Trial, achieves an area under the curve (AUC) of 0.871 on a separate test set of 2,085 subjects and identifies patients with high CVD mortality risks (AUC of 0.768). We validate our model against ECG-gated cardiac CT based markers, including coronary artery calcification (CAC) score, CAD-RADS score, and MESA 10-year risk score from an independent dataset of 335 subjects. Our work shows that, in high-risk patients, deep learning can convert LDCT for lung cancer screening into a dual-screening quantitative tool for CVD risk estimation.

SUBMITTER: Chao H

PROVIDER: S-EPMC8137697 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ImportanceScreening for lung cancer has the potential to reduce mortality, but in addition to detecting aggressive tumors, screening will also detect indolent tumors that otherwise may not cause clinical symptoms. These overdiagnosis cases represent an important potential harm of screening because they incur additional cost, anxiety, and morbidity associated with cancer treatment.ObjectiveTo estimate overdiagnosis in the National Lung Screening Trial (NLST).Design, setting, and participantsWe used data from the NLST, a randomized trial comparing screening using low-dose computed tomography (LDCT) vs chest radiography (CXR) among 53 452 persons at high risk for lung cancer observed for 6.4 years, to estimate the excess number of lung cancers in the LDCT arm of the NLST compared with the CXR arm.Main outcomes and measuresWe calculated 2 measures of overdiagnosis: the probability that a lung cancer detected by screening with LDCT is an overdiagnosis (PS), defined as the excess lung cancers detected by LDCT divided by all lung cancers detected by screening in the LDCT arm; and the number of cases that were considered overdiagnosis relative to the number of persons needed to screen to prevent 1 death from lung cancer.ResultsDuring follow-up, 1089 lung cancers were reported in the LDCT arm and 969 in the CXR arm of the NLST. The probability is 18.5% (95% CI, 5.4%-30.6%) that any lung cancer detected by screening with LDCT was an overdiagnosis, 22.5% (95% CI, 9.7%-34.3%) that a non-small cell lung cancer detected by LDCT was an overdiagnosis, and 78.9% (95% CI, 62.2%-93.5%) that a bronchioalveolar lung cancer detected by LDCT was an overdiagnosis. The number of cases of overdiagnosis found among the 320 participants who would need to be screened in the NLST to prevent 1 death from lung cancer was 1.38.Conclusions and relevanceMore than 18% of all lung cancers detected by LDCT in the NLST seem to be indolent, and overdiagnosis should be considered when describing the risks of LDCT screening for lung cancer.

Project description:BackgroundThe imaging dose of cone-beam computed tomography (CBCT) in image-guided radiotherapy (IGRT) poses adverse effects on patient health. To improve the quality of sparse-view low-dose CBCT images, a projection synthesis convolutional neural network (SynCNN) model is proposed.MethodsIncluded in this retrospective, single-center study were 223 patients diagnosed with brain tumours from Beijing Cancer Hospital. The proposed SynCNN model estimated two pairs of orthogonally direction-separable spatial kernels to synthesize the missing projection in between the input neighboring sparse-view projections via local convolution operations. The SynCNN model was trained on 150 real patients to learn patterns for inter-view projection synthesis. CBCT data from 30 real patients were used to validate the SynCNN, while data from a phantom and 43 real patients were used to test the SynCNN externally. Sparse-view projection datasets with 1/2, 1/4, and 1/8 of the original sampling rate were simulated, and the corresponding full-view projection datasets were restored using the SynCNN model. The tomographic images were then reconstructed with the Feldkamp-Davis-Kress algorithm. The root-mean-square error (RMSE), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) metrics were measured in both the projection and image domains. Five experts were invited to grade the image quality blindly for 40 randomly selected evaluation groups with a four-level rubric, where a score greater than or equal to 2 was considered acceptable image quality. The running time of the SynCNN model was recorded. The SynCNN model was directly compared with the three other methods on 1/4 sparse-view reconstructions.ResultsThe phantom and patient studies showed that the missing projections were accurately synthesized. In the image domain, for the phantom study, compared with images reconstructed from sparse-view projections, images with SynCNN synthesis exhibited significantly improved qualities with decreased values in RMSE and increased values in PSNR and SSIM. For the patient study, between the results with and without the SynCNN synthesis, the averaged RMSE decreased by 3.4×10-4, 10.3×10-4, and 21.7×10-4, the averaged PSNR increased by 3.4, 6.6, and 9.4 dB, and the averaged SSIM increased by 5.2×10-2, 18.9×10-2 and 33.9×10-2, for the 1/2, 1/4, and 1/8 sparse-view reconstructions, respectively. In expert subjective evaluation, both the median scores and acceptance rates of the images with SynCNN synthesis were higher than those reconstructed from sparse-view projections. It took the model less than 0.01 s to synthesize an inter-view projection. Compared with the three other methods, the SynCNN model obtained the best scores in terms of the three metrics in both domains.ConclusionsThe proposed SynCNN model effectively improves the quality of sparse-view CBCT images at a low time cost. With the SynCNN model, the CBCT imaging dose in IGRT could be reduced potentially.

Project description:ObjectivesDeep-learning image reconstruction (DLIR) offers unique opportunities for reducing image noise without degrading image quality or diagnostic accuracy in coronary CT angiography (CCTA). The present study aimed at exploiting the capabilities of DLIR to reduce radiation dose and assess its impact on stenosis severity, plaque composition analysis, and plaque volume quantification.MethodsThis prospective study includes 50 patients who underwent two sequential CCTA scans at normal-dose (ND) and lower-dose (LD). ND scans were reconstructed with Adaptive Statistical Iterative Reconstruction-Veo (ASiR-V) 100%, and LD scans with DLIR. Image noise (in Hounsfield units, HU) and quantitative plaque volumes (in mm3) were assessed quantitatively. Stenosis severity was visually categorized into no stenosis (0%), stenosis (< 20%, 20-50%, 51-70%, 71-90%, 91-99%), and occlusion (100%). Plaque composition was classified as calcified, non-calcified, or mixed.ResultsReduction of radiation dose from ND scans with ASiR-V 100% to LD scans with DLIR at the highest level (DLIR-H; 1.4 mSv vs. 0.8 mSv, p < 0.001) had no impact on image noise (28 vs. 27 HU, p = 0.598). Reliability of stenosis severity and plaque composition was excellent between ND scans with ASiR-V 100% and LD scans with DLIR-H (intraclass correlation coefficients of 0.995 and 0.974, respectively). Comparison of plaque volumes using Bland-Altman analysis revealed a mean difference of - 0.8 mm3 (± 2.5 mm3) and limits of agreement between - 5.8 and + 4.1 mm3.ConclusionDLIR enables a reduction in radiation dose from CCTA by 43% without significant impact on image noise, stenosis severity, plaque composition, and quantitative plaque volume.Key points•Deep-learning image reconstruction (DLIR) enables radiation dose reduction by over 40% for coronary computed tomography angiography (CCTA). •Image noise remains unchanged between a normal-dose CCTA reconstructed by ASiR-V and a lower-dose CCTA reconstructed by DLIR. •There is no impact on the assessment of stenosis severity, plaque composition, and quantitative plaque volume between the two scans.

Project description:ImportanceAnnual low-dose computed tomographic (LDCT) screening reduces lung cancer mortality, but harms could be reduced and cost-effectiveness improved by reusing the LDCT image in conjunction with deep learning or statistical models to identify low-risk individuals for biennial screening.ObjectiveTo identify low-risk individuals in the National Lung Screening Trial (NLST) and estimate, had they been assigned a biennial screening, how many lung cancers would have been delayed 1 year in diagnosis.Design, setting, and participantsThis diagnostic study included participants with a presumed nonmalignant lung nodule in the NLST between January 1, 2002, and December 31, 2004, with follow-up completed on December 31, 2009. Data were analyzed for this study from September 11, 2019, to March 15, 2022.ExposuresAn externally validated deep learning algorithm that predicts malignancy in current lung nodules using LDCT images (Lung Cancer Prediction Convolutional Neural Network [LCP-CNN]; Optellum Ltd) was recalibrated to predict 1-year lung cancer detection by LDCT for presumed nonmalignant nodules. Individuals with presumed nonmalignant lung nodules were hypothetically assigned annual vs biennial screening based on the recalibrated LCP-CNN model, Lung Cancer Risk Assessment Tool (LCRAT + CT [a statistical model combining individual risk factors and LDCT image features]), and the American College of Radiology recommendations for lung nodules, version 1.1 (Lung-RADS).Main outcomes and measuresPrimary outcomes included model prediction performance, the absolute risk of a 1-year delay in cancer diagnosis, and the proportion of people without lung cancer assigned a biennial screening interval vs the proportion of cancer diagnoses delayed.ResultsThe study included 10 831 LDCT images from patients with presumed nonmalignant lung nodules (58.7% men; mean [SD] age, 61.9 [5.0] years), of whom 195 were diagnosed with lung cancer from the subsequent screen. The recalibrated LCP-CNN had substantially higher area under the curve (0.87) than LCRAT + CT (0.79) or Lung-RADS (0.69) to predict 1-year lung cancer risk (P < .001). If 66% of screens with nodules were assigned to biennial screening, the absolute risk of a 1-year delay in cancer diagnosis would have been lower for recalibrated LCP-CNN (0.28%) than LCRAT + CT (0.60%; P = .001) or Lung-RADS (0.97%; P < .001). To delay only 10% of cancer diagnoses at 1 year, more people would have been safely assigned biennial screening under LCP-CNN than LCRAT + CT (66.4% vs 40.3%; P < .001).Conclusions and relevanceIn this diagnostic study evaluating models of lung cancer risk, a recalibrated deep learning algorithm was most predictive of 1-year lung cancer risk and had least risk of 1-year delay in cancer diagnosis among people assigned biennial screening. Deep learning algorithms could prioritize people for workup of suspicious nodules and decrease screening intensity for people with low-risk nodules, which may be vital for implementation in health care systems.

Project description:ImportanceDespite the recommendations of lung cancer screening guidelines and the evidence supporting the effectiveness of population-based lung screening, a common barrier to effective lung cancer screening is that the participation rates of low-dose computed tomography (LDCT) screening among individuals with the highest risk are not large. There are limited data from clinical practice regarding whether opportunistic LDCT screening is associated with reduced lung-cancer mortality.ObjectiveTo evaluate whether opportunistic LDCT screening is associated with improved prognosis among adults with lung cancer in mainland China.Design, setting, and participantsThis cohort study included patients diagnosed with lung cancer at Weihai Municipal Hospital Healthcare Group, Weihai City, China, from 2016 to 2021. Data were analyzed from January 2022 to February 2023.ExposuresData collected included demographic indicators, tumor characteristics, comorbidities, blood indexes, and treatment information. Patients were classified into screened and nonscreened groups on the basis of whether or not their lung cancer diagnosis occurred through opportunistic screening.Main outcomes and measuresFollow-up outcome indicators included lung cancer-specific mortality and all-cause mortality. Propensity score matching (PSM) was adopted to account for potential imbalanced factors between groups. The associations between LDCT screening and outcomes were analyzed using Cox regression models based on the matched data. Propensity score regression adjustment and inverse probability treatment weighting were used for sensitivity analysis.ResultsA total of 5234 patients (mean [SD] baseline age, 61.8 [9.8] years; 2518 [48.1%] female) with complete opportunistic screening information were included in the analytical sample, with 2251 patients (42.91%) receiving their lung cancer diagnosis through opportunistic screening. After 1:1 PSM, 2788 patients (1394 in each group) were finally included. The baseline characteristics of the matched patients were balanced between groups. Opportunistic screening with LDCT was associated with a 49% lower risk of lung cancer death (HR, 0.51; 95% CI, 0.42-0.62) and 46% lower risk of all-cause death (HR, 0.54; 95% CI, 0.45-0.64).Conclusions and relevanceIn this cohort study of patients with lung cancer, opportunistic lung cancer screening with LDCT was associated with lower lung cancer mortality and all-cause mortality. These findings suggest that opportunistic screening is an important supplement to population screening to improve prognosis of adults with lung cancer.

Project description:ImportanceLung cancer screening has been widely implemented in Europe and the US. However, there is little evidence on participation and diagnostic yields in population-based lung cancer screening in China.ObjectiveTo assess the participation rate and detection rate of lung cancer in a population-based screening program and the factors associated with participation.Design, setting, and participantsThis cross-sectional study used data from the Cancer Screening Program in Urban China from October 2013 to October 2019, with follow-up until March 10, 2020. The program is conducted at centers in 8 cities in Henan Province, China. Eligible participants were aged 40 to 74 and were evaluated for a high risk for lung cancer using an established risk score system.Main outcomes and measuresOverall and group-specific participation rates by common factors, such as age, sex, and educational level, were calculated. Differences in participation rates between those groups were compared. The diagnostic yield of both screening and nonscreening groups was calculated.ResultsThe study recruited 282 377 eligible participants and included 55 428 with high risk for lung cancer; the mean (SD) age was 55.3 (8.1) years, and 34 966 participants (63.1%) were men. A total of 22 260 participants underwent LDCT (participation rate, 40.16%; 95% CI, 39.82%-40.50%). The multivariable logistic regression model showed that female sex (odds ratio [OR], 1.64; 95% CI, 1.52-1.78), former smoking (OR, 1.26; 95% CI, 1.13-1.41), lack of physical activity (OR, 1.19; 95% CI, 1.14-1.24), family history of lung cancer (OR, 1.73; 95% CI, 1.66-1.79), and 7 other factors were associated with increased participation of LDCT screening. Overall, at 6-year follow-up, 78 participants in the screening group (0.35%; 95% CI, 0.29%-0.42%) and 125 in the nonscreening group (0.38%; 95% CI, 0.33%-0.44%) had lung cancer detected, which resulted in an odds ratio of 0.93 (95% CI, 0.70-1.23; P = .61).Conclusions and relevanceThe low participations rate in the program studied suggests that an improved strategy is needed. These findings may provide useful information for designing effective population-based lung cancer screening strategies in the future.

Dataset Information

Deep learning predicts cardiovascular disease risks from lung cancer screening low dose computed tomography.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets