Abstract"Radiomics" refers to the extraction and analysis of large amounts of advanced quantitative imaging features with high throughput from medical images obtained with computed tomography (CT), positron emission tomography (PET) or magnetic resonance imaging (MRI). Importantly, these data are designed to be extracted from standard-of-care images, leading to a very large potential subject pool. Radiomic data are in a mineable form that can be used to build descriptive and predictive models relating image features to phenotypes or gene-protein signatures. The core hypothesis of radiomics is that these models, which can include biological or medical data, can provide valuable diagnostic, prognostic or predictive information. The radiomics enterprise can be divided into distinct processes, each with its own challenges that need to be overcome: (i) image acquisition and reconstruction (ii) image segmentation and rendering (iii) feature extraction and feature qualification (iv) databases and data sharing for eventual (v) ad hoc informatic analyses.Each of these individual processes poses unique challenges. For example, optimum protocols for image acquisition and reconstruction have to be identified and harmonized. Also, segmentations have to be robust and involve minimal operator input. Features have to be generated that robustly reflect the complexity of the individual volumes, but cannot be overly complex or redundant. Furthermore, informatics databases that allow incorporation of image features and image annotations, along with medical and genetic data have to be generated. Finally, the statistical approaches to analyze these data have to be optimized, as radiomics is not a mature field of study. Each of these processes will be discussed in turn, as well as some of their unique challenges and

Pulmonary function is an easily measurable and reliable index of the physiological state of the lungs and airways 1 . Pulmonary function also predicts mortality in the general population, even among people who have never smoked (never-smokers) who have only modestly reduced pulmonary function and no respiratory symptoms 2,3 . The peak level of pulmonary function attained in early adulthood and its subsequent decline with age are likely influenced by genetic and environmental factors. Tobacco smoking is a major environmental cause of accelerated decline in pulmonary function with age. Other inhaled pollutants also appear to contribute. Familial aggregation studies suggest a genetic contribution to lung function, with heritability estimates exceeding 40% 4,5 , but little is known about the specific genetic factors involved. A relatively uncommon deficiency of α1-antitrypsin is the only established genetic risk factor for accelerated decline in pulmonary function and for development of chronic obstructive pulmonary disease (COPD), especially in smokers 4,6 . However, α1-antitrypsin accounts for little of the population variability in pulmonary function 4 . Candidate gene studies suggest that other genetic variants may influence the time course of pulmonary function and its decline in relation to smoking, but these putative genetic risk factors remain unknown 4 .Forced expiratory volume in the first second (FEV 1 ) and its ratio to forced vital capacity (FEV 1 /FVC) are two clinically relevant pulmonary function measures. Although both FEV 1 and FVC are influenced by lung size and can be reduced by restrictive lung diseases, obstructive lung disease leads to proportionately greater reduction in FEV 1 than FVC. Therefore, reduced FEV 1 /FVC, an indicator of airflow obstruction that is independent of lung size, is the primary criterion for defining an obstructive ventilatory defect 1 . Whereas low FEV 1 /FVC indicates the presence of airflow obstruction, FEV 1 is used to classify the severity and follow the progression of obstructive lung disease over time 5,7,8 .The first genome-wide association study (GWAS) for pulmonary function evaluating 70,987 SNPs in about 1,220 Framingham Heart Study (FHS) participants revealed no genome-wide significant loci 9 . Recently, a GWAS of FEV 1 /FVC using 2,540,223 SNPs in 7,691 FHS participants identified several SNPs on chromosome 4q31 near HHIP with genome-wide significance 10 . A GWAS of COPD 11 also implicated the HHIP region along with CHRNA3-CHRNA5 on chromosome 15, a region previously associated with nicotine dependence 12,13 .We conducted meta-analyses of GWAS results for a cross-sectional analysis of pulmonary function (FEV 1 /FVC and FEV 1 ) in 20,890 individuals of European ancestry from four Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium 14 studies: Atherosclerosis Risk in Communities (ARIC), Cardiovascular Health Study (CHS), FHS and Rotterdam Study (RS-I and RS-II). Given that cigarette smoking is a major risk factor for pulmonary fun...

SummaryWhile several lung cancer susceptibility loci have been identified, much of lung cancer heritability remains unexplained. Here, 14,803 cases and 12,262 controls of European descent were genotyped on the OncoArray and combined with existing data for an aggregated GWAS analysis of lung cancer on 29,266 patients and 56,450 controls. We identified 18 susceptibility loci achieving genome wide significance, including 10 novel loci. The novel loci highlighted the striking heterogeneity in genetic susceptibility across lung cancer histological subtypes, with four loci associated with lung cancer overall and six with lung adenocarcinoma. Gene expression quantitative trait analysis (eQTL) in 1,425 normal lung tissues highlighted RNASET2, SECISBP2L and NRG1 as candidate genes. Other loci include genes such as a cholinergic nicotinic receptor, CHRNA2, and the telomere-related genes, OFBC1 and RTEL1. Further exploration of the target genes will continue to provide new insights into the etiology of lung cancer.

Purpose Many radiomics features were originally developed for non-medical imaging applications and therefore original assumptions may need to be reexamined. In this study, we investigated the impact of slice thickness and pixel spacing (or pixel size) on radiomics features extracted from Computed Tomography (CT) phantom images acquired with different scanners as well as different acquisition and reconstruction parameters. The dependence of CT texture features on gray level discretization was also evaluated. Methods and Materials A texture phantom composed of 10 different cartridges of different materials was scanned on eight different CT scanners from three different manufacturers. The images were reconstructed for various slice thicknesses. For each slice thickness, the reconstruction Field Of View (FOV) was varied to render pixel sizes ranging from 0.39 to 0.98 mm. A fixed spherical region of interest (ROI) was contoured on the images of the shredded rubber cartridge and the 3D printed, 20% fill, acrylonitrile butadiene styrene plastic cartridge (ABS20) for all phantom imaging sets. Radiomics features were extracted from the ROIs using an in-house program. Features categories were: shape (10), intensity (16), GLCM (24), GLZSM (11), GLRLM (11), and NGTDM (5), fractal dimensions (8) and first order wavelets (128), for a total of 213 features. Voxel size resampling was performed to investigate the usefulness of extracting features using a suitably chosen voxel size. Acquired phantom image sets were resampled to a voxel size of 1 × 1 × 2 mm3 using linear interpolation. Image features were therefore extracted from resampled and original data sets and the absolute value of the percent coefficient of variation (%COV) for each feature was calculated. Based on %COV values, features were classified in 3 groups: 1) features with large variations before and after resampling (%COV > 50); 2) features with diminished variation (%COV < 30) after resampling; and 3) features that had originally moderate variation (%COV < 50%) and were negligibly affected by resampling. Group 2 features were further studied by modifying feature definitions to include voxel size. Original and voxel-size normalized features were used for interscanner comparisons. A subsequent analysis investigated feature dependency on gray level discretization by extracting 51 texture features from ROIs from each of the 10 different phantom cartridges using 16, 32, 64, 128 and 256 gray levels. Results Out of the 213 features extracted, 150 were reproducible across voxel sizes, 42 improved significantly (%COV < 30, Group 2) after resampling, and 21 had large variations before and after resampling (Group 1). Ten features improved significantly after definition modification effectively removed their voxel size dependency. Interscanner comparison indicated that feature variability among scanners nearly vanished for 8 of these 10 features. Furthermore, 17 out of 51 texture features were found to be dependent on the number of gray levels. These features were redef...

No abstract