Skip to main content

Table 1 Key studies in automated parenchymal texture analysis for breast cancer risk assessment

From: Beyond breast density: a review on the advancing role of parenchymal texture analysis in breast cancer risk assessment

Study Mammograms Dataset Breast sampling Texture features
Year Participating institutions F/D View A B S1 S2 T1 T2 T3 T4 T5
Distinguishing or predicting cancer cases from controls
Byng et al. (1997) [60] University of Toronto, Sunnybrook Health Science Centre, Ontario Cancer Institute F CC 354P 354 x   x    x  
Torres-Mejia et al. (2005) [72] LSHTM, Guy’s Hospital, UNAM, IPOFG F CC/MLO 111P 3100 x   x    x  
Wu et al. (2008) [76] University of Michigan F CC 128C 549 x     x   
Manduca et al. (2009) [66] Mayo Clinic, Moffitt F CC/MLO 246P 522 x   x x x x x
Wei et al. (2011) [73] University of Michigan F CC 136C 246 x     x   
Nielsen et al. (2011) [61] University of Copenhagen, Nordic Bioscience, Delft University of Technology, RadboudUMC, Mayo Clinic F MLO 245P 250 x      x x
Brandt et al. (2011) [74] University of Copenhagen, RadboudUMC, Synarc Imaging Technologies F MLO 245P 245   x     x  
Häberle et al. (2012) [56] Erlangen University Hospital, Fraunhofer Institute for Integrated Circuits IIS, IMPRS, UCLA F CC 864C 418 x   x x x x x
Li et al. (2012) [84] University of Chicago D CC 75C 328 x   x x   x x
Chen et al. (2014) [75] University of Manchester D MLO 50C 50   x      x
Nielsen et al. (2014) [64] ​University of Copenhagen, Nordic Bioscience, Biomediq, RadboudUMC, Mayo Clinic F CC/MLO 471P,C 692 x      x x
Li et al. (2014) [71] University of Chicago D CC 75C 328 x   x x    x
Karemore et al. (2014) [89] ​University of Copenhagen, RadboudUMC F MLO 245P 250   x     x x
Zheng et al. (2015) [51] University of Pennsylvania D MLO 106C 318   x x x x x  
Sun et al. (2015) [53] University of Texas, China Northeastern University, University of Oklahoma, TTUHS, Guiyang Medical University D CC 141P 199   x x x    x
Tan et al. (2015) [77] University of Texas, University of Oklahoma, University of Pittsburgh D CC/MLO 812P 1084 x   x x x x  
Tan et al. (2015) [78] University of Oklahoma, University of Pittsburgh D CC/MLO 430P 440 x    x x x x
Predicting the risk of carrying a high-risk genetic mutation
Huo et al. (2000) [80] University of Chicago F CC 15 143 x   x x    x
Huo et al. (2002) [55] University of Chicago, University of Pennsylvania F CC 30 142 x   x x    x
Li et al. (2004) [54] University of Chicago, University of Pennsylvania F CC 30 60 x   x x   x x
Li et al. (2005) [81] University of Chicago F CC 30 142 x   x x   x x
Li et al. (2007) [82] University of Chicago F CC 30 142 x      x  
Li et al. (2008) [83] University of Chicago F CC 30 142 x       x
Li et al. (2012) [84] University of Chicago D CC 53 328 x   x x   x x
Li et al. (2014) [71] University of Chicago D CC 53 328 x   x x    x
Gierach et al. (2014) [85] University of Chicago, NCI-NIH, Washington Radiology Associates, Genentech, USUHS, UCL, WRNMC, Westat Inc. F CC 137 100 x   x x   x x
  1. The Table describes the image data used in each study, including type of mammograms and dataset size, as well as methodological details for the computerized texture analysis, the technique of breast sampling, and algorithm implementation of texture features
  2. IMPRS International Max Planck Research School for Optics and Imaging, IPOFG Instituto Português de Oncologia Francisco Gentil, LSHTM London School of Hygiene and Tropical Medicine, Moffitt Moffitt Cancer Center and Research Institute, NCI-NIH National Cancer Institute, National Institutes of Health, RadboudUMC Radboud University Nijmegen Medical Centre, TTUHS Texas Tech University Health Sciences, UCL University College London, UCLA University of California at Los Angeles, UNAM Universidad Nacional Autónoma de México, USUHS Uniformed Services University of the Health Sciences, WRNMC Walter Reed National Military Medical Center
  3. Mammograms: F Digitized screen-film, D Full-field digital, CC cranio-caudal, MLO mediolateral-oblique; Dataset: A cancer cases (Pprior, unaffected, images, Cimages from the contralateral, unaffected, breast at the time of cancer diagnosis) or other high-risk population (i.e., BRCA1/2 carriers), B controls; Breast sampling: S1 retro-areolar region or the entire breast/dense tissue as a single region of interest (ROI), S2 multiple ROIs covering the entire breast; Types of texture features: T1 gray-level histogram, T2 co-occurrence, T3 run-length, T4 structural/pattern, T5 multi-resolution/spectral