Search

Background: Chest x-ray is widely utilized for the evaluation of pulmonary conditions due to its technical simplicity, cost-effectiveness, and portability. However, as a two-dimensional (2-D) imaging modality, chest x-ray images depict limited anatomical details and are challenging to interpret., Purpose: To validate the feasibility of reconstructing three-dimensional (3-D) lungs from a single 2-D chest x-ray image via Vision Transformer (ViT)., Methods: We created a cohort of 2525 paired chest x-ray images (scout images) and computed tomography (CT) acquired on different subjects and we randomly partitioned them as follows: (1) 1800 - training set, (2) 200 - validation set, and (3) 525 - testing set. The 3-D lung volumes segmented from the chest CT scans were used as the ground truth for supervised learning. We developed a novel model termed XRayWizard that employed ViT blocks to encode the 2-D chest x-ray image. The aim is to capture global information and establish long-range relationships, thereby improving the performance of 3-D reconstruction. Additionally, a pooling layer at the end of each transformer block was introduced to extract feature information. To produce smoother and more realistic 3-D models, a set of patch discriminators was incorporated. We also devised a novel method to incorporate subject demographics as an auxiliary input to further improve the accuracy of 3-D lung reconstruction. Dice coefficient and mean volume error were used as performance metrics as the agreement between the computerized results and the ground truth., Results: In the absence of subject demographics, the mean Dice coefficient for the generated 3-D lung volumes achieved a value of 0.738 ± 0.091. When subject demographics were included as an auxiliary input, the mean Dice coefficient significantly improved to 0.769 ± 0.089 (p < 0.001), and the volume prediction error was reduced from 23.5 ± 2.7%. to 15.7 ± 2.9%., Conclusion: Our experiment demonstrated the feasibility of reconstructing 3-D lung volumes from 2-D chest x-ray images, and the inclusion of subject demographics as additional inputs can significantly improve the accuracy of 3-D lung volume reconstruction., (© 2023 American Association of Physicists in Medicine.)

Purpose: To clarify the causal relationship between factors contributing to the postoperative survival of patients with esophageal cancer., Methods: A cohort of 195 patients who underwent surgery for esophageal cancer between 2008 and 2021 was used in the study. All patients had preoperative chest computed tomography (CT) and positron emission tomography-CT (PET-CT) scans prior to receiving any treatment. From these images, high throughput and quantitative radiomic features, tumor features, and various body composition features were automatically extracted. Causal relationships among these image features, patient demographics, and other clinicopathological variables were analyzed and visualized using a novel score-based directed graph called "Grouped Greedy Equivalence Search" (GGES) while taking prior knowledge into consideration. After supplementing and screening the causal variables, the intervention do-calculus adjustment (IDA) scores were calculated to determine the degree of impact of each variable on survival. Based on this IDA score, a GGES prediction formula was generated. Ten-fold cross-validation was used to assess the performance of the models. The prediction results were evaluated using the R-Squared Score (R 2 score)., Results: The final causal graphical model was formed by two PET-based image variables, ten body composition variables, four pathological variables, four demographic variables, two tumor variables, and one radiological variable (Percentile 10). Intramuscular fat mass was found to have the most impact on overall survival month. Percentile 10 and overall TNM (T: tumor, N: nodes, M: metastasis) stage were identified as direct causes of overall survival (month). The GGES casual model outperformed GES in regression prediction (R 2  = 0.251) (p < 0.05) and was able to avoid unreasonable causality that may contradict common sense., Conclusion: The GGES causal model can provide a reliable and straightforward representation of the intricate causal relationships among the variables that impact the postoperative survival of patients with esophageal cancer., (© 2023 American Association of Physicists in Medicine.)

Objective: To develop and validate a novel deep learning architecture to classify retinal vein occlusion (RVO) on color fundus photographs (CFPs) and reveal the image features contributing to the classification., Methods: The neural understanding network (NUN) is formed by two components: (1) convolutional neural network (CNN)-based feature extraction and (2) graph neural networks (GNN)-based feature understanding. The CNN-based image features were transformed into a graph representation to encode and visualize long-range feature interactions to identify the image regions that significantly contributed to the classification decision. A total of 7062 CFPs were classified into three categories: (1) no vein occlusion ("normal"), (2) central RVO, and (3) branch RVO. The area under the receiver operative characteristic (ROC) curve (AUC) was used as the metric to assess the performance of the trained classification models., Results: The AUC, accuracy, sensitivity, and specificity for NUN to classify CFPs as normal, central occlusion, or branch occlusion were 0.975 (± 0.003), 0.911 (± 0.007), 0.983 (± 0.010), and 0.803 (± 0.005), respectively, which outperformed available classical CNN models., Conclusion: The NUN architecture can provide a better classification performance and a straightforward visualization of the results compared to CNNs., (© 2022 American Association of Physicists in Medicine.)

Background: Estimating whole-body composition from limited region-computed tomography (CT) scans has many potential applications in clinical medicine; however, it is challenging., Purpose: To investigate if whole-body composition based on several tissue types (visceral adipose tissue [VAT], subcutaneous adipose tissue [SAT], intermuscular adipose tissue [IMAT], skeletal muscle [SM], and bone) can be reliably estimated from a chest CT scan only., Methods: A cohort of 97 lung cancer subjects who underwent both chest CT scans and whole-body positron emission tomography-CT scans at our institution were collected. We used our in-house software to automatically segment and quantify VAT, SAT, IMAT, SM, and bone on the CT images. The field-of-views of the chest CT scans and the whole-body CT scans were standardized, namely, from vertebra T1 to L1 and from C1 to the bottom of the pelvis, respectively. Multivariate linear regression was used to develop the computer models for estimating the volumes of whole-body tissues from chest CT scans. Subject demographics (e.g., gender and age) and lung volume were included in the modeling analysis. Ten-fold cross-validation was used to validate the performance of the prediction models. Mean absolute difference (MAD) and R-squared (R 2 ) were used as the performance metrics to assess the model performance., Results: The R 2 values when estimating volumes of whole-body SAT, VAT, IMAT, total fat, SM, and bone from the regular chest CT scans were 0.901, 0.929, 0.900, 0.933, 0.928, and 0.918, respectively. The corresponding MADs (percentage difference) were 1.44 ± 1.21 L (12.21% ± 11.70%), 0.63 ± 0.49 L (29.68% ± 61.99%), 0.12 ± 0.09 L (16.20% ± 18.42%), 1.65 ± 1.40 L (10.43% ± 10.79%), 0.71 ± 0.68 L (5.14% ± 4.75%), and 0.17 ± 0.15 L (4.32% ± 3.38%), respectively., Conclusion: Our algorithm shows promise in its ability to estimate whole-body compositions from chest CT scans. Body composition measures based on chest CT scans are more accurate than those based on vertebra third lumbar., (© 2022 The Authors. Medical Physics published by Wiley Periodicals LLC on behalf of American Association of Physicists in Medicine.)

Purpose: To investigate the relationship between macrovasculature features and the standardized uptake value (SUV) of positron emission tomography (PET), which is a surrogate for the metabolic activity of a lung tumor., Methods: We retrospectively analyzed a cohort of 90 lung cancer patients who had both chest CT and PET-CT examinations before receiving cancer treatment. The SUVs in the medical reports were used. We quantified three macrovasculature features depicted on CT images (i.e., vessel number, vessel volume, and vessel tortuosity) and several tumor features (i.e., volume, maximum diameter, mean diameter, surface area, and density). Tumor size (e.g., volume) was used as a covariate to adjust for possible confounding factors. Backward stepwise multiple regression analysis was performed to develop a model for predicting PET SUV from the relevant image features. The Bonferroni correction was used for multiple comparisons., Results: PET SUV was positively correlated with vessel volume (R = 0.44, p < 0.001) and vessel number (R = 0.44, p < 0.001) but not with vessel tortuosity (R = 0.124, p > 0.05). After adjusting for tumor size, PET SUV was significantly correlated with vessel tortuosity (R = 0.299, p = 0.004) and vessel number (R = 0.224, p = 0.035), but only marginally correlated with vessel volume (R = 0.187, p = 0.079). The multiple regression model showed a performance with an R-Squared of 0.391 and an adjusted R-Squared of 0.355 (p < 0.001)., Conclusions: Our investigations demonstrate the potential relationship between macrovasculature and PET SUV and suggest the possibility of inferring the metabolic activity of a lung tumor from chest CT images., (© 2021 American Association of Physicists in Medicine.)

Purpose: The potential to compute volume metrics of emphysema from planar scout images was investigated in this study. The successful implementation of this concept will have a wide impact in different fields, and specifically, maximize the diagnostic potential of the planar medical images., Methods: We investigated our premise using a well-characterized chronic obstructive pulmonary disease (COPD) cohort. In this cohort, planar scout images from computed tomography (CT) scans were used to compute lung volume and percentage of emphysema. Lung volume and percentage of emphysema were quantified on the volumetric CT images and used as the "ground truth" for developing the models to compute the variables from the corresponding scout images. We trained two classical convolutional neural networks (CNNs), including VGG19 and InceptionV3, to compute lung volume and the percentage of emphysema from the scout images. The scout images (n = 1,446) were split into three subgroups: (1) training (n = 1,235), (2) internal validation (n = 99), and (3) independent test (n = 112) at the subject level in a ratio of 8:1:1. The mean absolute difference (MAD) and R-square (R2) were the performance metrics to evaluate the prediction performance of the developed models., Results: The lung volumes and percentages of emphysema computed from a single planar scout image were significantly linear correlated with the measures quantified using volumetric CT images (VGG19: R2 = 0.934 for lung volume and R2 = 0.751 for emphysema percentage, and InceptionV3: R2 = 0.977 for lung volume and R2 = 0.775 for emphysema percentage). The mean absolute differences (MADs) for lung volume and percentage of emphysema were 0.302 ± 0.247L and 2.89 ± 2.58%, respectively, for VGG19, and 0.366 ± 0.287L and 3.19 ± 2.14, respectively, for InceptionV3., Conclusions: Our promising results demonstrated the feasibility of inferring volume metrics from planar images using CNNs., (© 2021 American Association of Physicists in Medicine.)

Objectives: To develop and validate a deep learning algorithm to automatically detect and segment an orbital abscess depicted on computed tomography (CT)., Methods: We retrospectively collected orbital CT scans acquired on 67 pediatric subjects with a confirmed orbital abscess in the setting of infectious orbital cellulitis. A context-aware convolutional neural network (CA-CNN) was developed and trained to automatically segment orbital abscess. To reduce the requirement for a large dataset, transfer learning was used by leveraging a pre-trained model for CT-based lung segmentation. An ophthalmologist manually delineated orbital abscesses depicted on the CT images. The classical U-Net and the CA-CNN models with and without transfer learning were trained and tested on the collected dataset using the 10-fold cross-validation method. Dice coefficient, Jaccard index, and Hausdorff distance were used as performance metrics to assess the agreement between the computerized and manual segmentations., Results: The context-aware U-Net with transfer learning achieved an average Dice coefficient and Jaccard index of 0.78 ± 0.12 and 0.65 ± 0.13, which were consistently higher than the classical U-Net or the context-aware U-Net without transfer learning (P < 0.01). The average differences of the abscess between the computerized results and the experts in terms of volume and Hausdorff distance were 0.10 ± 0.11 mL and 1.94 ± 1.21 mm, respectively. The context-aware U-Net detected all orbital abscess without false positives., Conclusions: The deep learning solution demonstrated promising performance in detecting and segmenting orbital abscesses on CT images in strong agreement with a human observer., (© 2021 American Association of Physicists in Medicine.)

Purpose: Knowledge of pulmonary interlobar fissure integrity is of interest in a number of clinical and investigational applications. The authors developed and tested a high resolution CT based automated computerized scheme for this purpose., Methods: The fissure integrity assessment scheme consists of the following steps: (1) Fissure detection, (2) individual fissure identification, (3) fissure type determination, (4) "complete" interlobe surface estimation, and (5) fissure integrity estimation. For evaluation purposes, 50 anonymized chest CT examinations were ascertained and the complete and "incomplete" regions of the fissures of interest were manually marked by two experienced radiologists. After applying the scheme to the same examinations, differences among fissure percent completeness estimates based on the radiologists' manual markings and the automated computerized scheme were computed and compared., Results: Average differences in estimated fissure percent completeness (integrity) between the results of the computerized scheme and that based on each of the two radiologists' markings were 6.88% +/- 5.86%, 9.57% +/- 7.77%, and 4.19% +/- 5.64% for the right major fissures, the right minor fissures, and the left major fissures, respectively. The differences between results based on radiologists' markings for the same fissures were 4.27% +/- 3.32%, 7.02% +/- 5.54%, and 4.23% +/- 4.93%, respectively. The difference among the three matched measurement sets for each fissure were statistically significant (Friedman's test, p < or = 0.005) but paired comparisons showed that much of the observed difference was related to inter-reader differences rather than reader-computerized scheme differences. Computerized estimates were correlated with each of the radiologist's estimates (Spearman, p < 0.0001)., Conclusions: While variability between readers-based estimates of fissure integrity was smaller than differences between the computerized scheme and each of the readers, the result reported here are quite encouraging in that the magnitude of these differences were in the same magnitude, demonstrating the feasibility of using a computerized scheme for this purpose.

Purpose: Lung Image Database Consortium (LIDC) is the largest public CT image database of lung nodules. In this study, the authors present a comprehensive and the most updated analysis of this dynamically growing database under the help of a computerized tool, aiming to assist researchers to optimally use this database for lung cancer related investigations., Methods: The authors developed a computer scheme to automatically match the nodule outlines marked manually by radiologists on CT images. A large variety of characteristics regarding the annotated nodules in the database including volume, spiculation level, elongation, interobserver variability, as well as the intersection of delineated nodule voxels and overlapping ratio between the same nodules marked by different radiologists are automatically calculated and summarized. The scheme was applied to analyze all 157 examinations with complete annotation data currently available in LIDC dataset., Results: The scheme summarizes the statistical distributions of the abovementioned geometric and diagnosis features. Among the 391 nodules, (1) 365 (93.35%) have principal axis length < or =20 mm; (2) 120, 75, 76, and 120 were marked by one, two, three, and four radiologists, respectively; and (3) 122 (32.48%) have the maximum volume overlapping ratios -80% for the delineations of two radiologists, while 198 (50.64%) have the maximum volume overlapping ratios <60%. The results also showed that 72.89% of the nodules were assessed with malignancy score between 2 and 4, and only 7.93% of these nodules were considered as severely malignant (malignancy > or =4)., Conclusions: This study demonstrates that LIDC contains examinations covering a diverse distribution of nodule characteristics and it can be a useful resource to assess the performance of the nodule detection and/or segmentation schemes.

The authors present a new computerized scheme to automatically detect lung nodules depicted on computed tomography (CT) images. The procedure is performed in the signed distance field of the CT images. To obtain an accurate signed distance field, CT images are first interpolated linearly along the axial direction to form an isotropic data set. Then a lung segmentation strategy is applied to smooth the lung border aiming to include as many juxtapleural nodules as possible while minimizing over segmentations of the lung regions. Potential nodule regions are then detected by locating local maximas of signed distances in each subvolume with values and the sizes larger than the smallest nodule of interest in the three-dimensional space. Finally, all detected candidates are scored by computing the similarity distance of their medial axis-like shapes obtained through a progressive clustering strategy combined with a marching cube algorithm from a sphere based shape. A free-response receiver operating characteristics curve is computed to assess the scheme performance. A performance test on 52 low-dose CT screening examinations that depict 184 verified lung nodules showed that during the initial stage the scheme achieved an asymptotic maximum sensitivity of 95.1% (175/184) with an average of 1200 suspicious voxels per CT examination. The nine missed nodules included two small solid nodules (with a diameter < or =3.1 mm) and seven nonsolid nodules. The final performance level after the similarity scoring stage was an absolute sensitivity level, namely, including the nine missed during the initial stage, of 81.5% (150/184) with 6.5 false-positive identifications per CT examination. This preliminary study demonstrates the feasibility of applying a simple and robust geometric model using the signed distance field to identify suspicious lung nodules. In the authors' data set the sensitivity of this scheme is not affected by nodule size. In addition to potentially being a stand alone approach, the signed distance field based method can be easily implemented as an initial filtering step in other computer-aided detection schemes.

When reading mammograms, radiologists routinely search for and compare suspicious breast lesions identified on two corresponding craniocaudal (CC) and mediolateral oblique (MLO) views. Automatically identifying and matching the same true-positive breast lesions depicted on two views is an important step for developing successful multiview based computer-aided detection (CAD) schemes. The authors developed a method to automatically register breast areas and detect matching strips of interest used to identify the matched mass regions depicted on CC and MLO views. The method uses an ellipse based model to fit the breast boundary contour (skin line) and set a local Cartesian coordinate system for each view. One intersection point between the major/minor axis and the fitted ellipse perimeter passed through breast boundary is selected as the origin and the majoraxis and the minoraxis of the ellipse are used as the two axis of the Cartesian coordinate system. When a mass is identified on one view, the scheme computes its position in the local coordinate system. Then, the distance is mapped onto the local coordinate of the other view. At the end of the mapped distance a registered centerline of the matching strip is created. The authors established an image database that includes 200 test examinations each depicting one verified mass visible on the two views. They tested whether the registered centerline identified on another view can be used to locate the matched mass region. The experiments show that the average distance between the mass region centers and the registered centerlines was +/- 8.3 mm and in 91% of testing cases the registered centerline actually passes through the matched mass regions. A matching strip width of 47 mm was required to achieve 100% sensitivity for the test database. The results demonstrate the feasibility of the proposed method to automatically identify masses depicted on CC and MLO views, which may improve future development of multiview based CAD schemes.

We developed and tested a fully automated computerized scheme that identifies pulmonary airway sections depicted on computed tomography (CT) images and computes their sizes including the lumen and airway wall areas. The scheme includes four processing modules that (1) segment left and right lung areas, (2) identify airway locations, (3) segment airway walls from neighboring pixels, and (4) compute airway sizes. The scheme uses both a raster scanning and a labeling algorithm complemented by simple classification rules for region size and circularity to automatically search for and identify airway sections of interest. A profile tracking method is used to segment airway walls from neighboring pixels including those associated with dense tissue (i.e., pulmonary arteries) along scanning radial rays. A partial pixel membership method is used to compute airway size. The scheme was tested on ten randomly selected CT studies that included 26 sets of CT images acquired using both low and conventional dose CT examinations with one of four reconstruction algorithms (namely, "bone," "lung," "soft," and "standard" convolution kernels). Three image section thicknesses (1.25, 2.5, and 5 mm) were evaluated. The scheme detected a large number of quantifiable airway sections when the CT images were reconstructed using high spatial frequency convolution kernels. The detection results demonstrated a consistent trend for all test image sets in that as airway lumen size increases, on average the airway wall area increases as well and the wall area percentage decreases. The study suggested that CT images reconstructed using high spatial frequency convolution kernels and thin-section thickness were most amenable to automated detection, reasonable segmentation, and quantified assessment when the airways are close to being perpendicular to the CT image plane.

In this study, we developed and tested a new multiview-based computer-aided detection (CAD) scheme that aims to maintain the same case-based sensitivity level as a single-image-based scheme while substantially increasing the number of masses being detected on both ipsilateral views. An image database of 450 four-view examinations (1800 images) was assembled. In this database, 250 cases depicted malignant masses, of which 236 masses were visible on both views and 14 masses were visible only on one view. First, we detected suspected mass regions depicted on each image in the database using a single-image-based CAD. For each identified region (with detection score > or = 0.55), we then identified a matching strip of interest on the ipsilateral view based on the projected distance to the nipple along the centerline. By lowering CAD operating threshold inside the matching strip, we searched for a region located inside the strip and paired it with the original region. A multifeature-based artificial neural network scored the likelihood of the paired "matched" regions representing true-positive masses. All single (unmatched) regions except for those either with very high detection scores (> or = 0.85) or those located near the chest wall that cannot be matched on the other view were discarded. The original single-image-based CAD scheme detected 186 masses (74.4% case-based sensitivity) and 593 false-positive regions. Of the 186 identified masses, 91 were detected on two views (48.9%) and 95 were detected only on one view (51.1%). Of the false-positive detections, 54 were paired on the ipsilateral view inside the corresponding matching strips and the remaining 485 were not, which represented 539 case-based false-positive detections (0.3 per image). Applying the multiview-based CAD scheme, the same case-based sensitivity was maintained while cueing 169 of 186 masses (90.9%) on both views and at the same time reducing the case-based false-positive detection rate by 23.7% (from 539 to 411). The study demonstrated that the new multiview-based CAD scheme could substantially increase the number of masses being cued on two ipsilateral views while reducing the case-based false-positive detection rate.

A method for quantitatively estimating lesion "size" from mammographic images was developed and evaluated. The main idea behind the measure, termed "integrated density" (ID), is that the total x-ray attenuation attributable to an object is theoretically invariant with respect to the projected view and object deformation. Because it is possible to estimate x-ray attenuation of a lesion from relative film densities, after appropriate corrections for background, the invariant property of the measure is expected to result in an objective method for evaluating the "sizes" of breast lesions. ID was calculated as the integral of the estimated image density attributable to a lesion, relative to surrounding background, over the area of the lesion and after corrections for the nonlinearity of the film characteristic curve. This effectively provides a measure proportional to lesion volume. We computed ID and more traditional measures of size (such as "mass diameter" and "effective size") for 100 pairs of ipsilateral mammographic views, each containing a lesion that was relatively visible in both views. The correlation between values calculated for each measure from corresponding pairs of ipsilateral views were computed and compared. All three size-related measures (mass diameter, effective size, and ID) exhibited reasonable linear relationship between paired views (r2>0.7, P<0.001). Specifically, the ID measures for the 100 masses were found to be highly correlated (r2=0.9, P<0.001) between ipsilateral views of the same mass. The correlation increased substantially (r2=0.95), when a measure with linear dimensions of length was defined as the cube root of ID. There is a high degree of correlation between ID-based measures obtained from different views of the same mass. ID-based measures showed a higher degree of invariance than mass diameter or effective size.