Radiomics and stacking regression model for measuring bone mineral density using abdominal computed tomography.

BACKGROUND: Measurement of bone mineral density (BMD) is the most important method to diagnose osteoporosis. However, current BMD measurement is always performed after a fracture has occurred. PURPOSE: To explore whether a radiomic model based on abdominal computed tomography (CT) can predict the BMD of lumbar vertebrae. MATERIAL AND METHODS: A total of 245 patients who underwent both dual-energy X-ray absorptiometry (DXA) and abdominal CT examination (training cohort, n = 196; validation cohort, n = 49) were included in our retrospective study. In total, 1218 image features were extracted from abdominal CT images for each patient. Combined with clinical information, three steps including least absolute shrinkage and selection operator (LASSO) regression were used to select key features. A two-tier stacking regression model with multi-algorithm fusion was used for BMD prediction, which can integrate the advantages of linear model and non-linear model. The prediction results of this model were compared with those using a single regressor. The degree-of-freedom adjusted coefficient of determination (Adjusted-R2), root mean square error (RMSE), and mean absolute error (MAE) were used to evaluate the regression performance. RESULTS: Compared with other regression methods, the two-tier stacking regression model has a higher regression performance, with Adjusted-R2, RMSE, and MAE of 0.830, 0.077, and 0.06, respectively. Pearson correlation analysis and Bland-Altman analysis showed that the BMD predicted by the model had a high correlation with the DXA results (r = 0.932, difference = -0.01 ± 0.1412 mg/cm2). CONCLUSION: Using radiomics, the BMD of lumbar vertebrae could be predicted from abdominal CT images.

Preoperative clinical-radiomics nomogram for microvascular invasion prediction in hepatocellular carcinoma using [Formula: see text]F-FDG PET/CT.

PURPOSE: To develop a clinical-radiomics nomogram by incorporating radiomics score and clinical predictors for preoperative prediction of microvascular invasion in hepatocellular carcinoma. METHODS: A total of 97 HCC patients were retrospectively enrolled from Shanghai Universal Medical Imaging Diagnostic Center and Changhai Hospital Affiliated to the Second Military Medical University. 909 CT and 909 PET slicers from 97 HCC patients were divided into a training cohort (N = 637) and a validation cohort (N = 272). Radiomics features were extracted from each CT or PET slicer, and features selection was performed with least absolute shrinkage and selection operator regression and radiomics score was also generated. The clinical-radiomics nomogram was established by integrating radiomics score and clinical predictors, and the performance of the models were evaluated from its discrimination ability, calibration ability, and clinical usefulness. RESULTS: The radiomics score consisted of 45 selected features, and age, the ratio of maximum to minimum tumor diameter, and [Formula: see text]F-FDG uptake status were independent predictors of microvascular invasion. The clinical-radiomics nomogram showed better performance for MVI detection (0.890 [0.854, 0.927]) than the clinical nomogram (0.849 [0.804, 0.893]) ([Formula: see text]). Both nomograms showed good calibration and the clinical-radiomics nomogram's clinical practicability outperformed the clinical nomogram. CONCLUSIONS: With the combination of radiomics score and clinical predictors, the clinical-radiomics nomogram can significantly improve the predictive efficacy of microvascular invasion in hepatocellular carcinoma ([Formula: see text]) compared with clinical nomogram.

Blood vessel segmentation of fundus images via cross-modality dictionary learning.

Automated retinal blood vessel segmentation is important for the early computer-aided diagnosis of some ophthalmological diseases and cardiovascular disorders. Traditional supervised vessel segmentation methods are usually based on pixel classification, which categorizes all pixels into vessel and non-vessel pixels. In this paper, we propose a new retinal vessel segmentation method with the motivation to extract vessels based on vessel block segmentation via cross-modality dictionary learning. For this, we first enhance the structural information of vessels using multi-scale filtering. Then, cross-modality description and segmentation dictionaries are learned to build the intrinsic relationship between the enhanced vessels and the labeled ground truth vessels for the purpose of vessel segmentation. Also, effective pre-processing and post-processing are adopted to promote the performance. Experimental results on three benchmark data sets demonstrate that the proposed method can achieve good segmentation results.

Optimizing multiview video plus depth retargeting technique for stereoscopic 3D displays.

In a multiview video plus depth (MVD) based three-dimensional (3D) video system, the generation of the contents with simultaneous resolution and depth adjustments is very challenging. In this paper, we have presented a Multiview Video plus Depth ReTargeting (MVDRT) technique for stereoscopic 3D (S3D) displays. The main motivation of this work is to optimize the resolution and depth of original MVD data so that it is suitable for view synthesis. Our method takes shape preservation, line bending and visual comfort constraints into account, and simultaneously optimizes the horizontal, vertical and depth coordinates in display space. The retargeted MVD data is used to generate the contents for S3D displays. Experimental results demonstrate our method can achieve a better view synthesis performance than other approaches that still preserve the original depth information after retargeting, leading to promising S3D experience.

Optimizing visual comfort for stereoscopic 3D display based on color-plus-depth signals.

Visual comfort is a long-facing problem in stereoscopic 3D (S3D) display. In this paper, targeting to produce S3D content based on color-plus-depth signals, a general framework for depth mapping to optimize visual comfort for S3D display is proposed. The main motivation of this work is to remap the depth range of color-plus-depth signals to a new depth range that is suitable to comfortable S3D display. Towards this end, we first remap the depth range globally based on the adjusted zero disparity plane, and then present a two-stage global and local depth optimization solution to solve the visual comfort problem. The remapped depth map is used to generate the S3D output. We demonstrate the power of our approach on perceptually uncomfortable and comfortable stereoscopic images.

Classification of Clouds in Satellite Imagery Using Adaptive Fuzzy Sparse Representation.

Automatic cloud detection and classification using satellite cloud imagery have various meteorological applications such as weather forecasting and climate monitoring. Cloud pattern analysis is one of the research hotspots recently. Since satellites sense the clouds remotely from space, and different cloud types often overlap and convert into each other, there must be some fuzziness and uncertainty in satellite cloud imagery. Satellite observation is susceptible to noises, while traditional cloud classification methods are sensitive to noises and outliers; it is hard for traditional cloud classification methods to achieve reliable results. To deal with these problems, a satellite cloud classification method using adaptive fuzzy sparse representation-based classification (AFSRC) is proposed. Firstly, by defining adaptive parameters related to attenuation rate and critical membership, an improved fuzzy membership is introduced to accommodate the fuzziness and uncertainty of satellite cloud imagery; secondly, by effective combination of the improved fuzzy membership function and sparse representation-based classification (SRC), atoms in training dictionary are optimized; finally, an adaptive fuzzy sparse representation classifier for cloud classification is proposed. Experiment results on FY-2G satellite cloud image show that, the proposed method not only improves the accuracy of cloud classification, but also has strong stability and adaptability with high computational efficiency.

Developing analytical approaches to explore the connection between endocrine-active pharmaceuticals in water to effects in fish.

The emphasis of this research project was to develop and optimize a solid-phase extraction method and high-performance liquid chromatography-electrospray ionization-mass spectrometry method, such that a linkage between the detection of endocrine-active pharmaceuticals (EAPs) in the aquatic environment and subsequent effects on fish populations could eventually be studied. Four EAPs were studied: tamoxifen (TAM), exemestane (EXE), letrozole (LET), anastrozole (ANA); and three TAM metabolites: 4-hydroxytamoxifen, e/z endoxifen, and n-desmethyl tamoxifen. In aqueous matrices, the use of isotopically labeled standards for the EAPs allowed for the generation of good recoveries, greater than 80 %, and low relative standard deviations (% RSDs) (3 to 27 %). TAM metabolites had lower recoveries in the spiked water matrices: 35 to 93 % in waste/source water compared to 58 to 110 % in DI water. The precision in DI water was acceptable ranging from 8 to 38 % RSD. However, the precision in real environmental wastewaters could be poor, ranging from 15 to 120 % RSD, dependent upon unique matrix effects. In plasma, the overall recoveries of the EAPs were acceptable: 88 to 110 %, with %RSDs of 6 to 18 % (Table 3). The spiked recoveries of the TAM metabolites from plasma were good, ranging from 77 to 120 %, with %RSDs ranging from 27 to 32 %. Two of the TAM metabolites, 4-hydroxytamoxifen and n-desmethyl tamoxifen, were confirmed in most of the environmental aqueous samples. The discovery of TAM metabolites demonstrates that the source of the TAM metabolites, TAM, is constant, introducing a pseudo-persistence of this chemical into the environment.

Development of a nomograph integrating radiomics and deep features based on MRI to predict the prognosis of high grade Gliomas.

The purpose of this study was to assess the overall survival of patients with HGG using a nomogram which combines the optimized radiomics with deep signatures extracted from 3D Magnetic Resonance Images (MRI) as well as clinical predictors. One training cohort of 168 HGG patients and one validation cohort of 42 HGG patients were enrolled in this study. From each patient's 3D MRI, 1284 radiomics features were extracted, and 8192 deep features were extracted via transfer learning. By using Least Absolute Shrinkage and Selection Operator (LASSO) regression to select features, the radiomics signatures and deep signatures were generated. The radiomics and deep features were then analyzed synthetically to generate a combined signature. Finally, the nomogram was developed by integrating the combined signature and clinical predictors. The radiomics and deep signatures were significantly associated with HGG patients' survival time. The signature derived from the synthesized radiomics and deep features showed a better prognostic performance than those from radiomics or deep features alone. The nomogram we developed takes the advantages of both radiomics and deep signatures, and also integrates the predictive ability of clinical indicators. The calibration curve shows our predicted survival time by the nomogram was very close to the actual time.

Point sources of emerging contaminants along the Colorado River Basin: source water for the arid Southwestern United States.

Emerging contaminants (ECs) (e.g., pharmaceuticals, illicit drugs, personal care products) have been detected in waters across the United States. The objective of this study was to evaluate point sources of ECs along the Colorado River, from the headwaters in Colorado to the Gulf of California. At selected locations in the Colorado River Basin (sites in Colorado, Utah, Nevada, Arizona, and California), waste stream tributaries and receiving surface waters were sampled using either grab sampling or polar organic chemical integrative samplers (POCIS). The grab samples were extracted using solid-phase cartridge extraction (SPE), and the POCIS sorbents were transferred into empty SPEs and eluted with methanol. All extracts were prepared for, and analyzed by, liquid chromatography-electrospray-ion trap mass spectrometry (LC-ESI-ITMS). Log D(OW) values were calculated for all ECs in the study and compared to the empirical data collected. POCIS extracts were screened for the presence of estrogenic chemicals using the yeast estrogen screen (YES) assay. Extracts from the 2008 POCIS deployment in the Las Vegas Wash showed the second highest estrogenicity response. In the grab samples, azithromycin (an antibiotic) was detected in all but one urban waste stream, with concentrations ranging from 30ng/L to 2800ng/L. Concentration levels of azithromycin, methamphetamine and pseudoephedrine showed temporal variation from the Tucson WWTP. Those ECs that were detected in the main surface water channels (those that are diverted for urban use and irrigation along the Colorado River) were in the region of the limit-of-detection (e.g., 10ng/L), but most were below detection limits.