egmentation of 3D images using EM algorithms

1 2 3 4 |

`img` |
is a 3d arrary representing an image. |

`nclust` |
is the number of clusters/classes to be segmented. |

`beta` |
is a matrix of size nclust x nclust, representing the prior weight of classes neighbouring each other. |

`z.scale` |
ratio of voxel dimension in x/y direction and z direction. Will be multiplied on beta for neighbouring voxel in z direction. |

`method` |
only "cem" classification EM algorithm implemented. |

`varfixed` |
is a logical variable. If TRUE, variacne is equal in each class. |

`maxit` |
is the maximum number of iterations. |

`mask` |
is a logical array, representing the voxels to be used in the segmentation. |

`priormu` |
is a vector with mean of the normal prior of the expected values of all classes. Default is NA, which represents no prior assumption. |

`priormusd` |
is a vector with standard deviations of the normal prior of the expected values of all classes. |

`min.eps` |
stop criterion. Minimal change in sum of squared estimate of mean in order to stop. |

`inforce.nclust` |
if TRUE enforces number of clusters to be nclust. Otherwise classes might be removed during algorithm. |

`start` |
? |

`silent` |
if TRUE, function remains silent during running time |

A list with "class": 3d array of class per voxel; "mu" estimated means; "sigma": estimated standard deviations.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | ```
## Not run:
original<-array(1,c(300,300,50))
for (i in 1:5)original[(i*60)-(0:20),,]<-original[(i*60)-(0:20),,]+1
for (i in 1:10)original[,(i*30)-(0:15),]<-original[,(i*30)-(0:15),]+1
original[,,26:50]<-4-aperm(original[,,26:50],c(2,1,3))
img<-array(rnorm(300*300*50,original,.2),c(300,300,50))
img<-img-min(img)
img<-img/max(img)
try1<-segment(img,3,beta=0.5,z.scale=.3)
print(sum(try1$class!=original)/prod(dim(original)))
beta<-matrix(rep(-.5,9),nrow=3)
beta<-beta+1.5*diag(3)
try2<-segment(img,3,beta,z.scale=.3)
print(sum(try2$class!=original)/prod(dim(original)))
par(mfrow=c(2,2))
img(original)
img(img)
img(try1$class)
img(try2$class)
## End(Not run)
``` |

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

Please suggest features or report bugs with the GitHub issue tracker.

All documentation is copyright its authors; we didn't write any of that.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.