Low-dose CT Research Articles

Concordance-based Predictive Uncertainty (CPU)-Index: Proof-of-concept with application towards improved specificity of lung cancers on low dose screening CT.

In this paper, we introduce a novel concordance-based predictive uncertainty (CPU)-Index, which integrates insights from subgroup analysis and personalized AI time-to-event models. Through its application in refining lung cancer screening (LCS) predictions generated by an individualized AI time-to-event model trained with fused data of low dose CT (LDCT) radiomics with patient demographics, we demonstrate its effectiveness, resulting in improved risk assessment compared to the Lung CT Screening Reporting & Data System (Lung-RADS). Subgroup-based Lung-RADS faces challenges in representing individual variations and relies on a limited set of predefined characteristics, resulting in variable predictions. Conversely, personalized AI time-to-event models are hindered by transparency issues and biases from censored data. By measuring the prediction consistency between subgroup analysis and AI time-to-event models, the CPU-Index framework offers a nuanced evaluation of the bias-variance trade-off and improves the transparency and reliability of predictions. Consistency was estimated by the concordance index of subgroup analysis-based similarity rank and model prediction similarity rank. Subgroup analysis-based similarity loss was defined as the sum-of-the-difference between Lung-RADS and feature-level 0-1 loss. Model prediction similarity loss was defined as squared loss. To test our approach, we identified 3,326 patients who underwent LDCT for LCS from 1/1/2015 to 6/30/2020 with confirmation of lung cancer on pathology within one year. For each LDCT image, the lesion associated with a Lung-RADS score was detected using a pretrained deep learning model from Medical Open Network for AI (MONAI), from which radiomic features were extracted. Radiomics were optimally fused with patient demographics via a positional encoding scheme and used to train a neural multi-task logistic regression time-to-event model that predicts malignancy. Performance was maximized when radiomics features were fused with positionally encoded demographic features. In this configuration, our algorithm raised the AUC from 0.81 ± 0.04 to 0.89 ± 0.02. Compared to standard Lung-RADS, our approach reduced the False-Positive-Rate from 0.41 ± 0.02 to 0.30 ± 0.12 while maintaining the same False-Negative-Rate. Our methodology enhances lung cancer risk assessment by estimating prediction uncertainty and adjusting accordingly. Furthermore, the optimal integration of radiomics and patient demographics improved overall diagnostic performance, indicating their complementary nature.

Fully automated coronary artery calcium score and risk categorization from chest CT using deep learning and multiorgan segmentation: A validation study from National Lung Screening Trial (NLST).

The National Lung Screening Trial (NLST) has shown that screening with low dose CT in high-risk population was associated with reduction in lung cancer mortality. These patients are also at high risk of coronary artery disease, and we used deep learning model to automatically detect, quantify and perform risk categorisation of coronary artery calcification score (CACS) from non-ECG gated Chest CT scans. Automated calcium quantification was performed using a neural network based on Mask regions with convolutional neural networks (R-CNN) for multiorgan segmentation. Manual evaluation of calcium was carried out using proprietary software. This study used 80 patients to train the segmentation model and randomly selected 1442 patients were used for the validation of the algorithm. We compared the model generated results with Ground Truth. Automatic cardiac and aortic segmentation model worked well (Mean Dice score: 0.91). Cohen's kappa coefficient between the reference actual and the interclass computed predictive categories on the test set is 0.72 (95% CI: 0.61-0.83). Our method correctly classifies the risk group in 78.8% of the cases and classifies the subjects in the same group. F-score is measured as 0.78; 0.71; 0.81; 0.82; 0.92 in calcium score categories 0(CS:0), I (1-99), II (100-400), III (400-1000), IV (>1000), respectively. 79% of the predictive scores lie in the same categories, 20% of the predictive scores are one category up or down, and only 1.2% patients were more than one category off. For the presence/absence of coronary artery calcifications, our deep learning model achieved a sensitivity of 90% and a specificity of 94%. Fully automated model shows good correlation compared with reference standards. Automating the process could improve diagnostic ability, risk categorization, facilitate primary prevention intervention, improve morbidity and mortality, and decrease healthcare costs.

Fast, Present and Future of the Concept of Spondyloarthritis.

Axial spondyloarthritis (axSpA) is a rather prevalent chronic inflammatory rheumatic disease that affects already relatively young patients. It has been known better since the end of the nineteenth century but quite a lot has been learned since the early 60ies when the first classification (diagnostic) criteria for ankylosing spondylitis (AS) were agreed on. I have been part of many developments in the last 30years, and I'm happy to have been able to contribute to the scientific progress in terms of diagnosis, imaging, pathophysiology and therapy. When I was asked to write a manuscript about the SpA concept I felt honored. Thus, the purpose of this extensive review was, on the one hand, to describe the history of AS and axSpA, and on the other hand, to reason about the concept and the gestalt of axSpA, and finally to deliver some ideas what future researchers could possibly do to further study the disease. The last 3 decades were full of innovations for both, classification and treatment of axSpA which also helped us to learn about the pathophysiology. Thus, TNFa, IL-17, IL-23 and Janus kinase are established targets to reduce inflammation. IL-17 and IL-23 are very special in that regard because they both work for psoriasis but only anti-IL-17 agents which don't work in IBD are approved for axSpA, while IL 23 inhibitors are approved for both, psoriasis and IBD, but they don't work in axSpA. New imaging techniques such as low dose CT and synthetic MRI are likely to improve the detection of both active and structural lesions of axSpA. This manuscript tries to describe the most important findings about axSpA. The main aim of research remains to discover the pathophysiology and to further improve treatment options in order to reduce and abolish inflammation and prevent new bone formation to increase the quality of life of our patients. The differences between male and female disease and the role of the immune system in axSpA are now the main challenges, and the role of special T-cell receptors seem to deserve special interest.

Impact of frailty and comorbidity on initial response to lung cancer screening invitation and low-dose CT screening uptake: Findings from the Yorkshire Lung Screening Trial.

Low-dose computed tomography screening reduces lung cancer-specific mortality in high-risk individuals. Lung cancer risk factors overlap with comorbid diseases, highlighting the significance of frailty and comorbidities for lung cancer screening (LCS). Here, we describe the prevalence of frailty and comorbidity in those invited for LCS and evaluate their associations with response to telephone risk assessment invitation and subsequent uptake of LCS. Analysis was based on the intervention arm of the Yorkshire Lung Screening Trial, where ever-smoked individuals aged 55-80 were invited to telephone risk assessment followed by community-based LCS if at higher risk. The electronic frailty index (eFI) was used to compute individual frailty scores (categorised as fit, mild, moderate and severe) and derive comorbidity data. Of 27,761 individuals invited, 24.1% (n = 6702), 8.5% (n = 2353) and 1.7% (n = 459) had mild, moderate and severe frailty, respectively. Over half responded to the invitation to telephone risk assessment (n = 14,523, 52.5%) with frailty associated with a higher response rate compared to fit individuals: adjusted odds ratio (ORadj) 1.34, 95% confidence interval (CI) 1.26-1.42 for mild frailty; ORadj 1.28, 95%CI 1.16-1.40 for moderate frailty; and ORadj 1.32, 95%CI 1.08-1.61 for severe frailty. Similar patterns were seen with comorbidity counts. After assessment, moderate (ORadj 0.75, 95%CI 0.59-0.96) and severe (ORadj 0.67, 95%CI 0.43-1.04) frailty were associated with reduced screening uptake. The presence of frailty was associated with increased response to LCS invitation. Given the strong association between frailty and reduced life expectancy, these results suggest that people with potentially more life years to be gained from LCS may be less inclined to take part. Further research is needed to explore the interactions between frailty and LCS decision-making to inform future invitation strategies.

Feasibility analysis of non-electrocardiogram-triggered chest low-dose computed tomography using a kV-independent reconstruction algorithm for predicting cardiovascular disease risk in patients receiving maintenance hemodialysis

ObjectivesThis study aimed to evaluate the feasibility and accuracy of non-electrocardiogram (ECG)-triggered chest low-dose computed tomography (LDCT) with a kV-independent reconstruction algorithm in assessing coronary artery calcification (CAC) degree and cardiovascular disease risk in patients receiving maintenance hemodialysis (MHD).MethodsIn total, 181 patients receiving MHD who needed chest CT and coronary artery calcium score (CACS) scannings sequentially underwent non-ECG-triggered, automated tube voltage selection, high-pitch chest LDCT with a kV-independent reconstruction algorithm and ECG-triggered standard CACS scannings. Then, the image quality, radiation doses, Agatston scores (ASs), and cardiac risk classifications of the two scans were compared.ResultsOf the 181 patients, 89, 83, and 9 were scanned at 100, 110, and 120 kV, respectively. Excluding those scanned at 120 kV, 172 patients were enrolled. Although the ASs of non-ECG-triggered LDCT were lower than those of the standard CACS, the agreement and correlation of ASs of the two scans were excellent, and both intraclass correlation coefficients (ICCs) and Pearson’s correlation coefficients were > 0.96. Cardiac risk classifications did not significantly differ between the non-ECG-triggered LDCT and standard CACS (χ2 = 3.933, P = 0.269), and the agreement was excellent (weighted kappa value = 0.936; 95% confidence interval (CI): 0.903–0.970). The effective radiation doses of standard CACS and non-ECG-triggered chest LDCT scannings were 1.34 ± 0.74 and 1.04 ± 0.35 mSv, respectively.ConclusionsThe non-ECG-triggered, automated tube voltage selection, high-pitch chest LDCT protocol with a kV-independent reconstruction algorithm can obtain chest scans and ASs simultaneously and significantly reduce patients’ radiation exposure.

Usefulness of four-dimensional noise reduction filtering using a similarity algorithm in low-dose dynamic computed tomography for the evaluation of breast cancer: a preliminary study.

To evaluate the effects of four-dimensional noise reduction filtering using a similarity algorithm (4D-SF) on the image quality and tumor visibility of low-dose dynamic computed tomography (CT) in evaluating breast cancer. Thirty-four patients with 38 lesions who underwent low-dose dynamic breast CT and were pathologically diagnosed with breast cancer were enrolled. Dynamic CT images were reconstructed using iterative reconstruction alone or in combination with 4D-SF. We selected the peak enhancement phase image of breast cancer for each patient for quantitative and qualitative evaluations of image quality and measurement of the maximum diameter of breast cancer. The signal-to-noise and contrast-to-noise ratios were calculated for quantitative evaluation. The maximum diameters of the breast cancer were measured from the images obtained with and without 4D-SF (4D-SF ±) (size-4D-SF + and size-4D-SF-) and for the pathological specimen (size-PS) and compared. The median and interquartile ranges of the signal-to-noise ratio [4D-SF-: 3.03 (2.54-4.17) vs 4D-SF + : 5.52 (4.75-6.66)] and contrast-to-noise ratio [4D-SF-: 2.88 (2.00-3.60) vs 4D-SF + : 7.84 (4.65-10.35)] were significantly higher for 4D-SF + than for 4D-SF- (p < 0.001). The overall image quality (Observer 1, p < 0.001; Observer 2, p < 0.001) and tumor margin sharpness scores (Observer 1, p = 0.003; Observer 2, p < 0.001) were significantly higher for 4D-SF + than for 4D-SF-. The tumor contrast scores for 4D-SF + and 4D-SF- were not significantly different (Observers 1, 2; p = 0.083). Size-4D-SF- was significantly smaller than size-PS (p < 0.001); size-4D-SF + was also smaller than size-PS, but the difference was not significant (p = 0.088). The Spearman's rank correlation coefficient was 0.65 for size-PS and size-4D-SF- and 0.77 for size-PS and size-4D-SF + . The 4D-SF can improve the image quality and tumor visibility of low-dose dynamic CT in evaluating breast cancer extent due to noise reduction.

An early lung cancer diagnosis model for non-smokers incorporating ct imaging analysis and circulating genetically abnormal cells (CACs)

BackgroundAn increase in the prevalence of lung cancer that is not smoking-related has been noticed in recent years. Unfortunately, these patients are not included in low dose computer tomography (LDCT) screening programs and are not actually considered in early diagnosis. Therefore, improved early diagnosis methods are urgently needed for non-smokers. It is necessary to establish a prediction model for non-smoking individuals at intermediate to high risk of developing lung cancer (LC) and develop a tool to address the significant gap in evaluating pulmonary nodules in non-smokers.MethodsWe retrospectively investigated 1121 patients with pulmonary nodules, who underwent LDCT examinations between September 2019 and March 2023. Five artificial intelligence (AI) algorithms were used to build two kinds of models and identify which one was better at diagnosing non-smoking pulmonary nodules patients. In the first model, we assigned 554 non-smoking individuals to a training cohort and 150 non-smoking patients to an independent validation cohort. The second model included 971 patients for the training set and 150 non-smoking patients for an independent validation set. All LDCT images of participants were obtained for AI analysis. AI of LDCT scans, liquid biopsy, and clinical characteristics were collected for model building.ResultsAmong LC patients, 58,4% were non-smokers. Non-smoking patients had a high incidence of LC (71.4%), and women showed a significant excess risk compared with non-smoking men in terms of LC risk. Furthermore, our results indicated that the model built using random forest (RF) method, which integrates clinical characteristics (age, extra-thoracic cancer history, gender), radiological characteristics of pulmonary nodules (nodule diameter, nodule count, upper lobe location, malignant sign at the nodule edge, subsolid status), the artificial intelligence analysis of LDCT data, and liquid biopsy achieved the best diagnostic performance in the independent external non-smokers validation cohort (sensitivity 92%, specificity 97%, area under the curve [AUC] = 0.99).ConclusionsThese results could significantly improve early non-smoker LC diagnosis and treatment for non-smoker patients with malignant nodules. The established multi-omics model is a noninvasive prediction tool for non-smoking malignant pulmonary nodule diagnosis. Validation revealed that these models exhibited excellent discrimination and calibration capacities, especially the first model built using the RF method, suggesting their clinical utility in the early screening and diagnosis of non-smoking LC.

A low-dose CT reconstruction method using sub-pixel anisotropic diffusion.

We present a new low-dose CT reconstruction method using sub-pixel and anisotropic diffusion. The sub-pixel intensity values and their second-order differences were obtained using linear interpolation techniques, and the new gradient information was then embedded into an anisotropic diffusion process, which was introduced into a penalty-weighted least squares model to reduce the noise in low-dose CT projection data. The high-quality CT image was finally reconstructed using the classical filtered back-projection (FBP) algorithm from the estimated data. In the Shepp-Logan phantom experiments, the structural similarity (SSIM) index of the CT image reconstructed by the proposed algorithm, as compared with FBP, PWLS-Gibbs and PWLS-TV algorithms, was increased by 28.13%, 5.49%, and 0.91%, the feature similarity (FSIM) index was increased by 21.08%, 1.78%, and 1.36%, and the root mean square error (RMSE) was reduced by 69.59%, 18.96%, and 3.90%, respectively. In the digital XCAT phantom experiments, the SSIM index of the CT image reconstructed by the proposed algorithm, as compared with FBP, PWLS-Gibbs and PWLS-TV algorithms, was increased by 14.24%, 1.43% and 7.89%, the FSIM index was increased by 9.61%, 1.78% and 5.66%, and the RMSE was reduced by 26.88%, 9.41% and 18.39%, respectively. In clinical experiments, the SSIM index of the image reconstructed using the proposed algorithm was increased by 19.24%, 15.63% and 3.68%, the FSIM index was increased by 4.30%, 2.92% and 0.43%, and the RMSE was reduced by 44.60%, 36.84% and 15.22% in comparison with FBP, PWLS-Gibbs and PWLS-TV algorithms, respectively. The proposed method can effectively reduce the noises and artifacts while maintaining the structural details in low-dose CT images.

Dual-domain Wasserstein Generative Adversarial Network with Hybrid Loss for Low-dose CT Imaging

Objective.Low-dose computed tomography (LDCT) has gained significant attention in hospitals and clinics as a popular imaging modality for reducing the risk of x-ray radiation. However, reconstructed LDCT images often suffer from undesired noise and artifacts, which can negatively impact diagnostic accuracy. This study aims to develop a novel approach to improve LDCT imaging performance.Approach.A dual-domain Wasserstein generative adversarial network (DWGAN) with hybrid loss is proposed as an effective and integrated deep neural network (DNN) for LDCT imaging. The DWGAN comprises two key components: a generator (G) network and a discriminator (D) network. TheGnetwork is a dual-domain DNN designed to predict high-quality images by integrating three essential components: the projection-domain denoising module, filtered back-projection-based reconstruction layer, and image-domain enhancement module. TheDnetwork is a shallow convolutional neural network used to differentiate between real (label) and generated images. To prevent the reconstructed images from becoming excessively smooth and to preserve both structural and textural details, a hybrid loss function with weighting coefficients is incorporated into the DWGAN.Main results.Numerical experiments demonstrate that the proposed DWGAN can effectively suppress noise and better preserve image details compared with existing methods. Moreover, its application to head CT data confirms the superior performance of the DWGAN in restoring structural and textural details.Significance.The proposed DWGAN framework exhibits excellent performance in recovering structural and textural details in LDCT images. Furthermore, the framework can be applied to other tomographic imaging techniques that suffer from image distortion problems.

A comparative analysis of image harmonization techniques in mitigating differences in CT acquisition and reconstruction.

The study aims to systematically characterize the effect of CT parameter variations on images and lung radiomic and deep features, and to evaluate the ability of different image harmonization methods to mitigate the observed variations. A retrospective in-house sinogram dataset of 100 low-dose chest CT scans was reconstructed by varying radiation dose (100%, 25%, 10%) and reconstruction kernels (smooth, medium, sharp). A set of image processing, convolutional neural network (CNNs), and generative adversarial network-based (GANs) methods were trained to harmonize all image conditions to a reference condition (100% dose, medium kernel). Harmonized scans were evaluated for image similarity using peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and learned perceptual image patch similarity (LPIPS), and for the reproducibility of radiomic and deep features using concordance correlation coefficient (CCC). CNNs consistently yielded higher image similarity metrics amongst others; for Sharp/10%, which exhibited the poorest visual similarity, PSNR increased from a mean ± CI of 17.763 ± 0.492 to 31.925 ± 0.571, SSIM from 0.219 ± 0.009 to 0.754 ± 0.017, and LPIPS decreased from 0.490 ± 0.005 to 0.275 ± 0.016. Texture-based radiomic features exhibited a greater degree of variability across conditions, i.e. a CCC of 0.500 ± 0.332, compared to intensity-based features (0.972 ± 0.045). GANs achieved the highest CCC (0.969 ± 0.009 for radiomic and 0.841 ± 0.070 for deep features) amongst others. Convolutional neural networks are suitable if downstream applications necessitate visual interpretation of images, whereas generative adversarial networks are better alternatives for generating reproducible quantitative image features needed for machine learning applications. Understanding the efficacy of harmonization in addressing multi-parameter variability is crucial for optimizing diagnostic accuracy and a critical step toward building generalizable models suitable for clinical use.

A Method for Validating PET and SPECT Cameras for Quantitative Clinical Imaging Trials Using Novel Radionuclides.

Our aim is to report methodology that has been developed to calibrate and verify PET and SPECT quantitative image accuracy and quality assurance for use with nonstandard radionuclides, especially with longer half-lives, in clinical imaging trials. Methods: Procedures have been developed for quantitative PET and SPECT image calibration for use in clinical trials. The protocol uses a 3-step approach: check quantitative accuracy with a previously calibrated radionuclide in a simple geometry, check the novel trial radionuclide in the same geometry, and check the novel radionuclide in a more challenging, complex geometry (the National Electrical Manufacturers Association [NEMA] NU-2 International Electrotechnical Commission [IEC] image-quality phantom). The radionuclides used in the trial as an example are 124I (PET) and 131I (SPECT). In both cases, whole-body tomographic SPECT and PET imaging with accompanying low-dose CT are required. PET accuracy is based on calibrating the dose calibrator to produce quantitative images for radionuclides other than 18F, with all images reconstructed on each individual site's PET systems. For SPECT, an independent sensitivity measurement is made and then used to calibrate the SPECT images reconstructed at the core laboratory. After calibration, the final testing for both PET and SPECT uses the NEMA NU-2 IEC image-quality phantom to derive several metrics including quantitative accuracy based on an average SUV (SUVavg). Results: Using the method described, 7 sites in Australia have been qualified for 10 PET/CT scanners using 124I imaging and 8 SPECT/CT systems for 131I. One PET/CT system was found to give a result outside the specification of an SUVavg of 1.0 ± 0.05. All SPECT/CT systems gave an SUVavg accurate to within ±10% (SUVmean, 1.0 ± 0.1) of the true value for reconstructed radioactivity concentration in Bq/cm3 Conclusion: A general methodology has been developed to calibrate and validate PET and SPECT systems for quantitative imaging in clinical trials. The preparation of the test objects and the procedures is relatively simple and can generally be implemented by the staff at the site of the imaging center with the equipment supplied by the clinical trials organization.

Lung cancer screening in India: Preparing for the future using smart tools &amp; biomarkers to identify highest risk individuals

There is a growing burden of lung cancer cases in India, incidence projected to increase from 63,708 cases (2015) to 81,219 cases (2025). The increasing numbers are attributed to smoking (India currently has nearly 100 million adult smokers) and environmental pollution. Most patients present with advanced disease (80-85% are incurable), causing nearly 60,000 annual deaths from lung cancer. Early detection through lung cancer screening (LCS) can result in curative therapies for earlier stages of lung cancer and improved survival. Annual low-dose computerized tomography (LDCT) is the standard method for LCS. Usually, high-risk populations (age&gt;50 yr and &gt;20 pack-years of smoking) are considered for LCS, but even such focused screening may be challenging in resource-limited countries like India. However, developing a smart LCS programme with high yield may be possible by leveraging demographic and genomic data, use of smart tools, and judicious use of blood-based biomarkers. Developing this model over the next several years will facilitate a structured cancer screening programme for populations at the highest risk of lung cancer. In this paper, we discuss the demographics of lung cancer in India and its relation to smoking patterns. Further, we elaborate on the potential applications and challenges of bringing a smart approach to LCS in high-risk populations in India.

Gender Disparities and Lung Cancer Screening Outcomes Among Individuals Who Have Never Smoked

Lung cancer in individuals who have never smoked (INS) is a growing global concern, with a rapidly increasing incidence and proportion among all lung cancer cases. Particularly in East Asia, opportunistic lung cancer screening (LCS) programs targeting INS have gained popularity. However, the sex-specific outcomes and drawbacks of screening INS remain unexplored, with data predominantly focused on women. To compare LCS outcomes between Asian women and men with no smoking history. This multicenter cohort study was conducted at health checkup centers in South Korea from 2009 to 2021. Participants included individuals aged 50 to 80 years with no smoking history who underwent low-dose computed tomography (LDCT) screening. Data were retrospectively analyzed from November 2023 to June 2024. Opportunistic LDCT screening for lung cancer. Participants were followed up until December 2022 for the outcome of death. Lung cancer diagnosis, diagnostic characteristics, clinical course, and lung cancer-specific deaths (LCSD) were compared between women and men. A total of 21 062 participants (16 133 [76.6%] women and 4929 [23.4%] men) with a mean (SD) age of 59.8 (7.2) years were included. From baseline screening, 176 participants (139 women [0.9%] and 37 men [0.8%]) were diagnosed with lung cancer (screen-detected); 131 of 139 women (94.3%) and 33 of 37 men (89.2%) were diagnosed with stage 0 to I disease, with 133 of 139 women (95.7%) and 36 of 37 men (97.3%) having adenocarcinoma. There were no significant sex-based differences in stage or histologic type distribution. Among the 21 062 screened individuals, LCSD was reported in 8 women and 3 men during a mean (SD) follow-up of 83.8 (41.7) months. Multivariable analyses found no significant association between sex and cumulative hazards of lung cancer diagnosis (adjusted hazard ratio [aHR], 0.90 [95% CI, 0.64-1.26] for men vs women) or LCSD (aHR, 1.06 [95% CI, 0.28-4.00] for men vs women). The estimated 5-year lung cancer-specific survival rate was 97.7% for women and 100% for men with screen-detected lung cancer, showing no significant sex differences. In this cohort study of Asian individuals with no smoking history who underwent LDCT screening, no significant sex-based differences were detected in lung cancer diagnosis, stage distribution, or LCSD. These findings suggest that men and women who have never smoked would experience similar risks of overdiagnosis with little to no benefit when exposed to indiscriminate screening.

Early diagnosis of lung cancer using a sensor gas analysis complex: case report

Background. Currently, low-dose computed tomography (LDCT) is the only screening test that reduces the risk of death from lung cancer. However, there are a number of disadvantages, such as lack of widespread use, high cost, high false-positive rate and the need to conduct studies only in high-risk groups, which significantly limit mass screening. exhaled breath analysis, which uses sensitive breath sensors, is a promising method to improve early diagnosis of lung cancer. Cancer Research Institute of Tomsk National Research Medical Center together with Tomsk State University and Tomsk Polytechnic Research Institute has developed a gas analysis complex capable of analyzing the gas composition of exhaled air with remote sampling from bags. during the study, data obtained by digitizing signals from gas analysis system sensors and patient metadata are recorded in a database for subsequent automated processing and analysis using a neural network. Case description. A 48-year-old female patient with a long history of smoking came to the clinic of the Cancer Research Institute for consultation with suspected pathological infiltration around the celiac trunk detected by abdominal CT. As a clinical trial of the developed gas analytical complex for cancer detection, a sample of exhaled air was taken, and the comparison of the composition of volatile organic compounds (VOCs) with that in the control group (healthy individuals) revealed abnormalities characteristic of lung cancer. the patient underwent a chest CT scan, which revealed stage IIB peripheral cancer of the lower lobe of the left lung. the original sensor gas analysis complex, which has no analogues in Russia, was used for the first time in the detection of lung cancer. the data obtained allowed us to suspect the presence of lung tumor in the patient and perform radical surgical treatment. the composition of VOCs in exhaled air was assessed on day 10 after surgery, and no significant changes in the composition of exhaled air were observed. Conclusion. Machine learning algorithms are actively used to diagnose socially significant diseases. the platforms being developed based on arrays of chemical sensors with data analysis using a neural network are promising candidates for implementation in screening activities.

Deep learning-based image domain reconstruction enhances image quality and pulmonary nodule detection in ultralow-dose CT with adaptive statistical iterative reconstruction-V.

To evaluate the image quality and lung nodule detectability of ultralow-dose CT (ULDCT) with adaptive statistical iterative reconstruction-V (ASiR-V) post-processed using a deep learning image reconstruction (DLIR)-based image domain compared to low-dose CT (LDCT) and ULDCT without DLIR. A total of 210 patients undergoing lung cancer screening underwent LDCT (mean ± SD, 0.81 ± 0.28 mSv) and ULDCT (0.17 ± 0.03 mSv) scans. ULDCT images were reconstructed with ASiR-V (ULDCT-ASiR-V) and post-processed using DLIR (ULDCT-DLIR). The quality of the three CT images was analyzed. Three radiologists detected and measured pulmonary nodules on all CT images, with LDCT results serving as references. Nodule conspicuity was assessed using a five-point Likert scale, followed by further statistical analyses. A total of 463 nodules were detected using LDCT. The image noise of ULDCT-DLIR decreased by 60% compared to that of ULDCT-ASiR-V and was lower than that of LDCT (p < 0.001). The subjective image quality scores for ULDCT-DLIR (4.4 [4.1, 4.6]) were also higher than those for ULDCT-ASiR-V (3.6 [3.1, 3.9]) (p < 0.001). The overall nodule detection rates for ULDCT-ASiR-V and ULDCT-DLIR were 82.1% (380/463) and 87.0% (403/463), respectively (p < 0.001). The percentage difference between diameters > 1 mm was 2.9% (ULDCT-ASiR-V vs. LDCT) and 0.5% (ULDCT-DLIR vs. LDCT) (p = 0.009). Scores of nodule imaging sharpness on ULDCT-DLIR (4.0 ± 0.68) were significantly higher than those on ULDCT-ASiR-V (3.2 ± 0.50) (p < 0.001). DLIR-based image domain improves image quality, nodule detection rate, nodule imaging sharpness, and nodule measurement accuracy of ASiR-V on ULDCT. Question Deep learning post-processing is simple and cheap compared with raw data processing, but its performance is not clear on ultralow-dose CT. Findings Deep learning post-processing enhanced image quality and improved the nodule detection rate and accuracy of nodule measurement of ultralow-dose CT. Clinical relevance Deep learning post-processing improves the practicability of ultralow-dose CT and makes it possible for patients with less radiation exposure during lung cancer screening.

Peripheral Extracellular Vesicles for Diagnosis and Prognosis of Resectable Lung Cancer: The LUCEx Study Protocol.

Background/Objectives: Lung cancer is the primary cause of cancer-related deaths. Most patients are typically diagnosed at advanced stages. Low-dose computed tomography (LDCT) has been proven to reduce lung cancer mortality, but screening programs using LDCT are associated with a high number of false positives and unnecessary thoracotomies. It is therefore imperative that a certain diagnosis is refined, especially in cases of solitary pulmonary nodules that are difficult to technically access for an accurate preoperative diagnosis. Extracellular vesicles (EVs) involved in intercellular communication may be an innovative biomarker for diagnosis and therapeutic strategies in lung cancer, regarding their ability to carry tumor-specific cargo. The aim of the LUCEx study is to determine if extracellular vesicle cargoes from both lung tissue and blood could provide complementary information to screen lung cancer patients and enable personalized follow-up after the surgery. Methods: The LUCEx study is a prospective study aiming to recruit 600 patients with lung cancer and 50 control subjects (false positives) undergoing surgery after diagnostic imaging for suspected pulmonary nodules using computed tomography (CT) scans. These patients will undergo curative surgery at the Department of Thoracic Surgery of the Miguel Servet Hospital in Zaragoza, Spain, and will be followed-up for at least 5 years. At baseline, samples from both tumor distal lung tissue and preoperative peripheral blood will be collected and processed to compare the quantity and content of EVs, particularly their micro-RNA (miRNA) cargo. At the third and fifth years of follow-up, CT scans, functional respiratory tests, and blood extractions will be performed. Discussion: Extracellular vesicles and their miRNA have emerged as promising tools for the diagnosis and prognosis of several diseases, including cancer. The LUCEx study, based on an observational clinical cohort, aims to understand the role of these vesicles and their translational potential as complementary tools for imaging diagnosis and prognosis.

Impact of comorbidities on the mortality benefits of lung cancer screening: a post-hoc analysis of PLCO and NLST trials.

To evaluate how comorbidities affect mortality benefits of lung cancer screening (LCS) with low-dose computed-tomography (LDCT). We developed a comorbidity index (PLCO-ci) using LCS-eligible participants' data from the Prostate Lung Colorectal and Ovarian (PLCO) trial (training set) and the National Lung Screening Trial (NLST) (validation set). PLCO-ci predicts 5-year non-lung cancer (LC) mortality using a regularized Cox model; with performance evaluated by the area under the ROC curve (ROCAUC). In NLST, LC mortality (per original publication) was compared between LDCT and chest X-ray arms across PLCO-ci quintile (Q1-5) using cause-specific hazard ratio (csHR) with 95% confidence intervals [95%CI]. Analyses included 34,690 PLCO and 53,452 NLST participants (mean age: 62 (±5) and 61 (±5) years old, 58% and 59% males, 39% and 41% active smokers, respectively). PLCO-ci predicted 5-year non-LC mortality with ROCAUC [95%CI] of 0.72 [0.71-0.74] in PLCO and 0.69 [0.67-0.70] in NLST. In NLST, at median follow-up of 6.5 years, LC mortality was significantly reduced for participants with intermediate comorbidity (Q2-Q3-Q4): csHR [95% CI] 0.62 [0.41-0.95], 0.68 [0.48-0.96] and 0.72 [0.54-0.96] respectively, with a non-statistically significant reduction for Q1 (csHR=0.72 [0.45-1.17]) and no reduction for Q5 participants (csHR=0.99 [0.79-1.23]). Participants in Q2-Q3-Q4 (60%) accounted for 89% of LC deaths averted among all NLST participants. Q1 participants had low LC incidence, while Q5 had higher localized LC lethality, more squamous cell carcinomas, and untreated LC. The PLCO-ci developed in this work shows that individuals with intermediate comorbidity benefited the most from LCS, highlighting the need of addressing comorbidities to achieve LC mortality benefits.

Comparison of the CT-based micromotion analysis method versus marker-based RSA in measuring femoral head translation and evaluation of its intra- and interobserver reliability: a prospective agreement diagnostic study on 27 patients up to 1 year.

Computed tomography radiostereometric analysis (CT-RSA) assesses implant micromovements using low-dose CT scans. We aimed to investigate whether CT-RSA is comparable to marker-based radiostereometric analysis (RSA) measuring early femoral head migration in cemented stems. We hypothesized that CT-RSA is comparable to marker-based RSA in evaluating femoral head subsidence. We prospectively included 31 patients undergoing cemented total hip arthroplasty (THA), of which 27 were eligible for the analysis. Femoral head migration at 1 year was measured with marker-based RSA and CT-RSA. Comparison was performed using paired analysis and Bland-Altman plots, and the intra- and interobserver reliability of CT-RSA was assessed Results:The median (interquartile range [IQR]) translation on the Y-axis measured with marker-based RSA was -0.86 mm (-1.10 to -0.37) and -0.83 mm (-1.11 to -0.48) for CT-RSA (i.e. subsidence), with a median difference of -0.03 mm (95% confidence interval [CI] -0.08 to 0.18). The minimal important difference in translation was set to 0.2 mm. This value was excluded from the CI of the differences. No statistical difference was found between marker-based RSA and CT-RSA regarding assessment of subsidence of the femoral head. The Bland-Altman plots showed good agreement between the 2 methods in measuring subsidence of the femoral head. The intra- and interobserver reliability of the CT-RSA method was excellent with intraclass correlation coefficient (ICC) = 1 (0.99-1) and ICC = 0.99 (0.99-1), respectively. We showed that CT-RSA was comparable to marker-based RSA in measuring femoral head subsidence. Moreover, the intra- and interobserver reliability of the CT-RSA method was excellent, suggesting that the method is assessor independent.

Repeatability of AI-based, automatic measurement of vertebral and cardiovascular imaging biomarkers in low-dose chest CT: the ImaLife cohort.

To evaluate the repeatability of AI-based automatic measurement of vertebral and cardiovascular markers on low-dose chest CT. We included participants of the population-based Imaging in Lifelines (ImaLife) study with low-dose chest CT at baseline and 3-4 month follow-up. An AI system (AI-Rad Companion chest CT prototype) performed automatic segmentation and quantification of vertebral height and density, aortic diameters, heart volume (cardiac chambers plus pericardial fat), and coronary artery calcium volume (CACV). A trained researcher visually checked segmentation accuracy. We evaluated the repeatability of adequate AI-based measurements at baseline and repeat scan using Intraclass Correlation Coefficient (ICC), relative differences, and change in CACV risk categorization, assuming no physiological change. Overall, 632 participants (63 ± 11 years; 56.6% men) underwent short-term repeat CT (mean interval, 3.9 ± 1.8 months). Visual assessment showed adequate segmentation in both baseline and repeat scan for 98.7% of vertebral measurements, 80.1-99.4% of aortic measurements (except for the sinotubular junction (65.2%)), and 86.0% of CACV. For heart volume, 53.5% of segmentations were adequate at baseline and repeat scans. ICC for adequately segmented cases showed excellent agreement for all biomarkers (ICC > 0.9). Relative difference between baseline and repeat measurements was < 4% for vertebral and aortic measurements, 7.5% for heart volume, and 28.5% for CACV. There was high concordance in CACV risk categorization (81.2%). In low-dose chest CT, segmentation accuracy of AI-based software was high for vertebral, aortic, and CACV evaluation and relatively low for heart volume. There was excellent repeatability of vertebral and aortic measurements and high concordance in overall CACV risk categorization. Question Can AI algorithms for opportunistic screening in chest CT obtain an accurate and repeatable result when applied to multiple CT scans of the same participant? Findings Vertebral and aortic analysis showed accurate segmentation and excellent repeatability; coronary calcium segmentation was generally accurate but showed modest repeatability due to a non-electrocardiogram-triggered protocol. Clinical relevance Opportunistic screening for diseases outside the primary purpose of the CT scan is time-consuming. AI allows automated vertebral, aortic, and coronary artery calcium (CAC) assessment, with highly repeatable outcomes of vertebral and aortic biomarkers and high concordance in overall CAC categorization.

Co-occurrence of bronchiectasis, airway wall thickening, and emphysema in Chinese low-dose CT screening.

To assess the co-occurrence of incidental CT lung findings (emphysema, bronchiectasis, and airway wall thickening) as well as associated risk factors in low-dose CT (LDCT) lung cancer screening in a Chinese urban population. Data from 978 participants aged 40-74 years from the Chinese NELCIN-B3 urban population study who underwent LDCT screening were selected. CT scans were reviewed for incidental lung findings: emphysema, bronchiectasis and airway wall thickness. Emphysema was defined in three ways (≥ trace, ≥ mild, or ≥ moderate) depending on severity. Participants were described and stratified by presence or absence of incidental lung findings. Logistic regression analyses were performed to examine the relationship between participant characteristics and CT findings. Mean age was 61.3 years ± 6.8 and 533 (54.6%) were female. 48% of participants had incidental lung findings: 19.9% had emphysema (≥ mild), 9.2% had bronchiectasis, and 35.7% had airway wall thickening. Among 978 participants, 2.1% showed all three findings. Multivariable analysis showed that higher age (OR: 1.06; 95% CI: 1.04-1.08; p < 0.001), male sex (OR: 1.68; 95% CI: 1.14-2.47; p = 0.008) smoking history (OR: 1.76; 95% CI: 1.02-3.03; p = 0.04 for former smokers; OR: 2.45; 95% CI: 1.59-3.78; p < 0.001 for current smokers), and the presence of respiratory symptoms (OR: 1.42; 95% CI: 1.01-2.00; p = 0.04) were associated with the risk of having at least one incidental lung findings. When different definitions of emphysema were used, the results remained consistent. In a Chinese urban population undergoing LDCT lung cancer screening, 48% had at least one incidental CT lung finding, which was associated with higher age, male sex, questionnaire-based respiratory symptoms and smoking history. Question Reporting of incidental lung findings that indicate lung disease risk lacks consensus in the cancer screening setting and needs evidence of co-occurrence in general populations. Findings Almost half of the 978 participants had at least one incidental lung CT finding; these were associated with older age, male sex, respiratory symptoms, and smoking history. Clinical relevance Incidental lung findings and related risk factors are often observed in low-dose CT lung cancer screening, and careful consideration of their relevance should be given to their inclusion in future screenings.