1. The document discusses medical image processing and analysis techniques. It covers topics like image enhancement, segmentation, quantification, and registration.
2. Segmentation is described as an essential analysis function that partitions images into homogeneous regions. It is important for feature extraction, measurements, and display. A variety of segmentation techniques have been developed.
3. Bayesian image restoration uses probabilistic models of the imaging process and prior knowledge to obtain an optimal estimate of the true image from observations. Markov random field models can incorporate spatial dependencies and allow introduction of expert knowledge.