Learning a nonlinear channelized observer for image quality assessment

We propose two algorithms for task-based image quality assessment based on machine learning. The channelized Hotelling observer (CHO) is a well-known numerical observer, which is used as a surrogate for human observers in assessments of lesion detectability. We explore the possibility of replacing the linear CHO with nonlinear algorithms that learn the relationship between measured image features and lesion detectability obtained from human observer studies. Our results suggest that both support vector machines and neural networks can offer improved performance over the CHO in predicting the human-observer performance.

[1]  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.

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

[3]  B.M.W. Tsui,et al.  Comparison of radially-symmetric versus oriented channel. Models using channelized hotelling observers for myocardial defect detection in parallel-hole SPECT , 1998, 1998 IEEE Nuclear Science Symposium Conference Record. 1998 IEEE Nuclear Science Symposium and Medical Imaging Conference (Cat. No.98CH36255).

[4]  D. Signorini,et al.  Neural networks , 1995, The Lancet.

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

[6]  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.

[7]  Martin Brown,et al.  Network Performance Assessment for Neurofuzzy Data Modelling , 1997, IDA.

[8]  Wilson S. Geisler,et al.  Image quality assessment based on a degradation model , 2000, IEEE Trans. Image Process..

[9]  M.A. King,et al.  A comparison of human observer LROC and numerical observer ROC for tumor detection in SPECT images , 1998, 1998 IEEE Nuclear Science Symposium Conference Record. 1998 IEEE Nuclear Science Symposium and Medical Imaging Conference (Cat. No.98CH36255).

[10]  T K Narayan,et al.  Prediction of human observer performance by numerical observers: an experimental study. , 1999, Journal of the Optical Society of America. A, Optics, image science, and vision.

[11]  Craig K. Abbey,et al.  Estimates of human-observer templates for a simple detection task in correlated noise , 2000, Medical Imaging.

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