Generalization Evaluation of Machine Learning Numerical Observers for Image Quality Assessment

In this paper, we present two new numerical observers (NO) based on machine learning for image quality assessment. The proposed NOs aim to predict human observer performance in a cardiac perfusion-defect detection task for single-photon emission computed tomography (SPECT) images. Human observer (HumO) studies are now considered to be the gold standard for task-based evaluation of medical images. However such studies are impractical for use in early stages of development for imaging devices and algorithms, because they require extensive involvement of trained human observers who must evaluate a large number of images.

[1]  Miguel P Eckstein,et al.  Task-based model/human observer evaluation of SPIHT wavelet compression with human visual system-based quantization. , 2005, Academic radiology.

[2]  Miguel P. Eckstein,et al.  The effect of nonlinear human visual system components on performance of a channelized Hotelling observer in structured backgrounds , 2006, IEEE Transactions on Medical Imaging.

[3]  Yongyi Yang,et al.  Learning a Channelized Observer for Image Quality Assessment , 2009, IEEE Transactions on Medical Imaging.

[4]  Eric Clarkson,et al.  Experimental determination of object statistics from noisy images. , 2003, Journal of the Optical Society of America. A, Optics, image science, and vision.

[5]  Craig K. Abbey,et al.  Optimal shifted estimates of human-observer templates in two-alternative forced-choice experiments , 2002, IEEE Transactions on Medical Imaging.

[6]  Jovan G. Brankov,et al.  Personalized numerical observer , 2010, Medical Imaging.

[7]  Harrison H Barrett,et al.  Validating the use of channels to estimate the ideal linear observer. , 2003, Journal of the Optical Society of America. A, Optics, image science, and vision.

[8]  M. King,et al.  A mathematical model of motion of the heart for use in generating source and attenuation maps for simulating emission imaging. , 1999, Medical physics.

[9]  Yongyi Yang,et al.  Channelized relevance vector machine as a numerical observer for cardiac perfusion defect detection task , 2011, Medical Imaging.

[10]  Jovan G. Brankov,et al.  Comparison of the internal noise models for channelized Hotelling observer , 2011, 2011 IEEE Nuclear Science Symposium Conference Record.

[11]  Miles N. Wernick,et al.  Optimization of iterative reconstructions of /sup 99m/Tc cardiac SPECT studies using numerical observers , 2001 .

[12]  M. Ljungberg,et al.  A Monte Carlo program for the simulation of scintillation camera characteristics. , 1989, Computer methods and programs in biomedicine.

[13]  O. Mangasarian,et al.  Robust linear programming discrimination of two linearly inseparable sets , 1992 .

[14]  George Eastman House,et al.  Sparse Bayesian Learning and the Relevance Vector Machine , 2001 .

[15]  L B Lusted,et al.  Signal detectability and medical decision-making. , 1971, Science.

[16]  Jie Yao,et al.  Predicting human performance by a channelized Hotelling observer model , 1992, Optics & Photonics.

[17]  John A. Swets,et al.  Evaluation of diagnostic systems : methods from signal detection theory , 1982 .

[18]  Kyle J Myers,et al.  Channelized-ideal observer using Laguerre-Gauss channels in detection tasks involving non-Gaussian distributed lumpy backgrounds and a Gaussian signal. , 2007, Journal of the Optical Society of America. A, Optics, image science, and vision.

[19]  C. Metz.  Basic principles of ROC analysis. , 1978, Seminars in nuclear medicine.

[20]  Jovan G. Brankov,et al.  Optimization of the internal noise models for channelized Hotelling observer , 2011, 2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[21]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[22]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[23]  David J. C. MacKay,et al.  Bayesian Interpolation , 1992, Neural Computation.

[24]  H H Barrett,et al.  Human- and model-observer performance in ramp-spectrum noise: effects of regularization and object variability. , 2001, Journal of the Optical Society of America. A, Optics, image science, and vision.

[25]  W. W. Peterson,et al.  The theory of signal detectability , 1954, Trans. IRE Prof. Group Inf. Theory.

[26]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[27]  Dr. M. G. Worster.  Methods of Mathematical Physics , 1947, Nature.

[28]  Cyril Castella,et al.  Human linear template with mammographic backgrounds estimated with a genetic algorithm. , 2007, Journal of the Optical Society of America. A, Optics, image science, and vision.

[29]  H H Barrett,et al.  Addition of a channel mechanism to the ideal-observer model. , 1987, Journal of the Optical Society of America. A, Optics and image science.

[30]  Miguel P Eckstein,et al.  Evaluation of internal noise methods for Hotelling observer models. , 2007, Medical physics.

[31]  M P Eckstein,et al.  Visual signal detection in structured backgrounds. II. Effects of contrast gain control, background variations, and white noise. , 1997, Journal of the Optical Society of America. A, Optics, image science, and vision.