Answer Happy

We discussed the problem of tissue classification in T1 MRI images. Explain how the Hidden Markov Random Field (HMRF) is implemented in DIPY and how and why it does what it does. More specifically read this tutorial and underlying DIPY code https://dipy.org/documentation/latest/examples built/tissue classification/#example-tissue-classification Read chapters 1 and 2.1 from Stan Li's book and the paper by Zhang 2001. Stan Li MRFs Chapters 1 and 2.1.pdf Zhang2001.pdf Explain how, why and where Expectation Maximization is used. Explain what the hidden variables represent. Explain why an MRF is also a Gibbs Random Field. Explain what is the purpose of Iterated Conditional Modes for this problem. Explain why one of the outputs of the method is a PVE map and why it is four dimensional (4D). Read carefully the code in DIPY and understand the input and outputs of all major functions of the HMRF class. In summary, please explain the what, why and how of the HMRF approach used in DIPY and why it is a good approximation of the tissue classification problem (distinguishing white matter, gray matter and corticospinal fluid areas in the brain).