[Master Index] [Index for export/lyngby]

lyngby_batch_bmr

(export/lyngby/lyngby_batch_bmr.m)


Function Synopsis

lyngby_batch_bmr(filenames, varargin)

Help text

 lyngby_batch_bmr    - Batch for Binomial mixture ROC curve

       lyngby_batch_bmr(filenames, 'PropertyName', 'PropertyValue')
     
       Input:    filenames   ANALYZE filenames as cell array strings
 
       Property: LambdaCommon [ Replications | {ReplicationsMethods} ]
                 NumberOfThresholds [ {8} | an integer ]
                 Path         { {''} ] filepath for result files
                 Prefix       [ {lyngby_bmr} | a string ] Filename
                              prefix for the result files.
                 VolMask      'Vol' structure with volume mask

       This functions is a batch for the binomial mixture estimate of
       the ROC-curve (receiver operating characteristics curve).

       Input is filenames for summary images of replicated
       experiments. Filenames should be organized in a cell array of
       string with replications in rows. It is possible to have
       several methods and this should be organized in columns of the
       'filenames' input argument. 

       'LambdaCommon' determines whether the mixing coefficient in
       the binomial mixture model (lmabda) should be common only across
       replications or common both across replications and methods.

       'NumberOfThresholds' determine the number of thresholds
       applied to the summary images. Each threshold will correspond
       to a point on the ROC-curve. This number should be larger than
       two. 

       The output of this function is a text file with results and a
       EPS-file with a plot of the ROC-curves. 

       Ref: Genovese, Noll, Eddy, MRM 38:497+, 1997.

       Example:
         % A specific data set
         Runs = [ 1 2 3 8 9 10 14 15 ]; 
         Methods = { 'fir_strength' 'fg_zscore' 'cc_energy' 'ica_image1'};
         for m = 1:length(Methods)
           method = Methods{m};
           k = 1;
           for r = 1:length(Runs) 
             filenames{r,m} = sprintf('psn_jerom25_run%02d_%s', Runs(r), Methods{m});
             k = k + 1;
           end
         end
         vmask = lyngby_read_analyze(filenames{1,1})
         lyngby_batch_bmr(filenames, 'VolMask', vmask);
         !gv lyngby_bmr.eps
                 
       See also LYNGBY, LYNGBY_PDE_BINMIX, LYNGBY_READ_ANALYZE,
                LYNGBY_READ_VOLUMES. 
                      
 $Id: lyngby_batch_bmr.m,v 1.6 2007/09/21 17:05:44 fn Exp $

Cross-Reference Information

This function calls

Produced by mat2html on Wed Jul 29 15:43:40 2009
Cross-Directory links are: OFF