//******************************************************************************
//
// File: Test05.java
// Package: edu.rit.mri.test
// Unit: Class edu.rit.mri.test.Test05
//
// This Java source file is copyright (C) 2008 by Alan Kaminsky. All rights
// reserved. For further information, contact the author, Alan Kaminsky, at
// ark@cs.rit.edu.
//
// This Java source file is part of the Parallel Java Library ("PJ"). PJ is free
// software; you can redistribute it and/or modify it under the terms of the GNU
// General Public License as published by the Free Software Foundation; either
// version 3 of the License, or (at your option) any later version.
//
// PJ is distributed in the hope that it will be useful, but WITHOUT ANY
// WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
// A PARTICULAR PURPOSE. See the GNU General Public License for more details.
//
// A copy of the GNU General Public License is provided in the file gpl.txt. You
// may also obtain a copy of the GNU General Public License on the World Wide
// Web at http://www.gnu.org/licenses/gpl.html.
//
//******************************************************************************
package edu.rit.mri.test;
import edu.rit.mri.PixelSignal;
import edu.rit.mri.SignalDataSetReader;
import edu.rit.mri.SpinSignal;
import edu.rit.mri.SpinSignalDifference;
import edu.rit.numeric.AggregateXYSeries;
import edu.rit.numeric.ListXYSeries;
import edu.rit.numeric.NonLinearLeastSquares;
import edu.rit.numeric.NonNegativeLeastSquares;
import edu.rit.numeric.Series;
import edu.rit.numeric.TooManyIterationsException;
import edu.rit.numeric.XYSeries;
import edu.rit.numeric.plot.Dots;
import edu.rit.numeric.plot.Plot;
import edu.rit.numeric.plot.Strokes;
import java.awt.Color;
import java.io.File;
import java.util.ArrayList;
import java.util.Collections;
/**
* Class Test05 is a unit test main program for class {@linkplain
* edu.rit.mri.SignalDataSetReader edu.rit.mri.SignalDataSetReader}. The program
* reads the data for a specified pixel index. The program does a spin
* relaxometry analysis on the data set. The program displays a plot of the data
* set along with the reconstructed spin signal.
*
* Usage: java edu.rit.mri.test.Test05 signalfile R1_lower
* R1_upper N index
*
signalfile = Input spin signal data set file
*
R1_lower = Lower R1 spin-lattice relaxation rate (1/sec)
*
R1_upper = Upper R1 spin-lattice relaxation rate (1/sec)
*
N = Number of R1 intervals
*
index = Pixel index
*
* @author Alan Kaminsky
* @version 24-Jun-2008
*/
public class Test05
{
// Prevent construction.
private Test05()
{
}
// Main program.
/**
* Main program.
*/
public static void main
(String[] args)
throws Exception
{
// Parse command line arguments.
if (args.length != 5) usage();
File signalfile = new File (args[0]);
double R1_lower = Double.parseDouble (args[1]);
double R1_upper = Double.parseDouble (args[2]);
int N = Integer.parseInt (args[3]);
int index = Integer.parseInt (args[4]);
// Set up spin signal data set reader.
SignalDataSetReader reader = new SignalDataSetReader (signalfile);
// Get data X-Y series.
Series tseries = reader.getTimeSeries();
int M = tseries.length();
PixelSignal signal = reader.getPixelSignal (index);
if (signal == null)
{
System.err.println ("Test05: Pixel "+index+" has no data");
System.exit (0);
}
Series dseries = signal.S_measured();
AggregateXYSeries dataseries = new AggregateXYSeries (tseries, dseries);
// // Print data X-Y series.
// System.out.println ("Measured t, S(t)");
// for (int i = 0; i < M; ++ i)
// {
// System.out.print (tseries.x(i));
// System.out.print ('\t');
// System.out.print (dseries.x(i));
// System.out.println();
// }
// Do a spin relaxometry analysis using nonnegative linear least
// squares. N spin relaxation rate intervals logarithmically spaced from
// R1_lower to R1_upper.
// Compute spin relaxation rates.
double[] Rseries = new double [N+1];
double logRlower = Math.log (R1_lower);
double logRupper = Math.log (R1_upper);
double interval = (logRupper - logRlower)/N;
for (int j = 0; j <= N; ++ j)
{
Rseries[j] = Math.exp (logRlower + j*interval);
}
// Create nonnegative linear least squares solver.
NonNegativeLeastSquares linsolver =
new NonNegativeLeastSquares (M, N+1);
// Find the solution.
for (int i = 0; i < M; ++ i)
{
double[] a_i = linsolver.a[i];
double t_i = dataseries.x(i);
for (int j = 0; j <= N; ++ j)
{
a_i[j] = SpinSignal.S (Rseries[j], t_i);
}
linsolver.b[i] = dataseries.y(i);
}
linsolver.solve();
double[] rhoseries = linsolver.x;
// Print the solution.
System.out.println ("Linear analysis: density, R1, T1");
for (int j = 0; j <= N; ++ j)
{
if (rhoseries[j] > 0.0)
{
System.out.print (j);
System.out.print (". ");
System.out.print (rhoseries[j]);
System.out.print ('\t');
System.out.print (Rseries[j]);
System.out.print ('\t');
System.out.print (1.0/Rseries[j]);
System.out.println();
}
}
System.out.print ("chi^2 = ");
System.out.print (linsolver.normsqr);
System.out.println();
// Find peaks in the solution. A peak occurs at index i if
// rho[i] > rho[i-1] and rho[i] > rho[i+1].
ArrayList rho = new ArrayList();
ArrayList R = new ArrayList();
ArrayList newrho = new ArrayList();
ArrayList newR = new ArrayList();
System.out.println ("Peaks in linear analysis: density, R1, T1");
for (int j = 0; j <= N; ++ j)
{
if (rhoseries[j] > (j == 0 ? 0.0 : rhoseries[j-1]) &&
rhoseries[j] > (j == N ? 0.0 : rhoseries[j+1]))
{
System.out.print (j);
System.out.print (". ");
System.out.print (rhoseries[j]);
System.out.print ('\t');
System.out.print (Rseries[j]);
System.out.print ('\t');
System.out.print (1.0/Rseries[j]);
System.out.println();
rho.add (rhoseries[j]);
R.add (Rseries[j]);
}
}
// Do a spin relaxometry analysis using nonlinear least squares. Peaks
// in the linear analysis give the initial vector of densities and
// rates.
// Repeat until the solution is plausible.
boolean plausible = false;
int L = rho.size();
while (L > 0 && ! plausible)
{
// Print solution before.
System.out.println ("Nonlinear analysis, before: density, R1, T1");
printSolution (dataseries, rho, R);
// Create spin signal difference function. L = number of tissues.
SpinSignalDifference fcn =
new SpinSignalDifference (dataseries, L);
// Create nonlinear least squares solver.
NonLinearLeastSquares nonlinsolver =
new NonLinearLeastSquares (fcn);
// Find the solution.
for (int i = 0; i < L; ++ i)
{
nonlinsolver.x[(i<<1)] = rho.get(i);
nonlinsolver.x[(i<<1)+1] = R.get(i);
}
try
{
nonlinsolver.solve();
newrho.clear();
newR.clear();
for (int i = 0; i < L; ++ i)
{
newrho.add (nonlinsolver.x[(i<<1)]);
newR.add (nonlinsolver.x[(i<<1)+1]);
}
// Print solution after.
System.out.println ("Nonlinear analysis, after: R1, T1");
printSolution (dataseries, newrho, newR);
// Decide if solution is plausible.
plausible = checkPlausibility (dataseries, newrho, newR);
}
// Couldn't find a solution.
catch (TooManyIterationsException exc)
{
System.out.println ("*** Too many iterations in nonlinear solver");
plausible = false;
}
// If solution is not plausible, eliminate tissue with smallest
// density and try again.
if (! plausible)
{
double minrho = Double.MAX_VALUE;
int mini = 0;
for (int i = 0; i < L; ++ i)
{
if (rho.get(i) < minrho)
{
minrho = rho.get(i);
mini = i;
}
}
rho.remove (mini);
R.remove (mini);
L = rho.size();
}
// If solution is plausible, stop.
else
{
rho = newrho;
R = newR;
}
}
// Create plot with data X-Y series.
Plot plot = new Plot();
plot.xAxisLength (1200)
.xAxisMajorDivisions (20)
.yAxisLength (600)
.yAxisMajorDivisions (10)
.seriesDots (Dots.circle (Color.BLACK, null, null, 5))
.seriesColor (Color.BLACK)
.seriesStroke (Strokes.solid (2))
.xySeries (dataseries);
// If we found a plausible solution, generate model spin signal X-Y
// series and add to plot.
if (L > 0)
{
ListXYSeries signalseries = new ListXYSeries();
for (int i = 0; i < M; ++ i)
{
double t_i = dataseries.x(i);
double s_i = 0.0;
for (int j = 0; j < rho.size(); ++ j)
{
s_i += SpinSignal.S (rho.get(j), R.get(j), t_i);
}
signalseries.add (t_i, s_i);
}
plot.seriesDots (null)
.seriesColor (Color.RED)
.seriesStroke (Strokes.solid (1))
.xySeries (signalseries);
}
else
{
System.out.println ("*** Could not find plausible solution");
}
// Display plot.
plot.getFrame().setVisible (true);
}
// Hidden operations.
/**
* Print the given parameters.
*/
private static void printSolution
(XYSeries dataseries,
ArrayList rho,
ArrayList R)
{
int L = rho.size();
for (int i = 0; i < L; ++ i)
{
System.out.print (rho.get(i));
System.out.print ('\t');
System.out.print (R.get(i));
System.out.print ('\t');
System.out.print (1.0/R.get(i));
System.out.println();
}
System.out.print ("chi^2 = ");
System.out.print (computeChiSqr (dataseries, rho, R));
System.out.println();
}
/**
* Calculate chi^2 between the measured data series and the model function
* determined by the given parameters.
*/
private static double computeChiSqr
(XYSeries dataseries,
ArrayList rho,
ArrayList R)
{
int M = dataseries.length();
int L = rho.size();
double chisqr = 0.0;
for (int i = 0; i < M; ++ i)
{
double t_i = dataseries.x(i);
double s_i = 0.0;
for (int j = 0; j < L; ++ j)
{
s_i += SpinSignal.S (rho.get(j), R.get(j), t_i);
}
double d = s_i - dataseries.y(i);
chisqr += d*d;
}
return chisqr;
}
/**
* Decide if the given solution is plausible.
*/
private static boolean checkPlausibility
(XYSeries dataseries,
ArrayList rho,
ArrayList R)
{
int M = dataseries.length();
int L = rho.size();
// If any density or rate is negative, solution is not plausible.
for (int i = 0; i < L; ++ i)
{
if (rho.get(i) < 0.0)
{
System.out.println ("*** Negative density: "+rho.get(i));
return false;
}
if (R.get(i) < 0.0)
{
System.out.println ("*** Negative rate: "+R.get(i));
return false;
}
}
// If relative difference between any two rates is too small, solution
// is not plausible.
for (int i = 0; i < L-1; ++ i)
{
double R_i = R.get(i);
for (int j = i+1; j < L; ++ j)
{
double R_j = R.get(j);
double reldiff = 2.0*Math.abs(R_i-R_j)/Math.abs(R_i+R_j);
if (reldiff <= 0.001)
{
System.out.println
("*** Rates too close: "+R_i+" and "+R_j+
", reldiff = "+reldiff);
return false;
}
}
}
// If sum of densities is too far from asymptotic measurement for large
// t, solution is not plausible.
double sumrho = 0.0;
for (int i = 0; i < L; ++ i)
{
sumrho += rho.get(i);
}
double S_last = 0.0;
int n = 0;
for (int i = M-1; i >=0 && n < 7; -- i)
{
S_last += dataseries.y(i);
++ n;
}
S_last /= n;
double reldiff = Math.abs(sumrho-S_last)/Math.abs(S_last);
if (reldiff >= 0.2)
{
System.out.println
("*** Sum of densities = "+sumrho+
" doesn't agree with asymptotic measurement = "+S_last+
", reldiff = "+reldiff);
return false;
}
// Solution is plausible.
return true;
}
/**
* Print a usage message and exit.
*/
private static void usage()
{
System.err.println ("Usage: java edu.rit.mri.test.Test05 ");
System.err.println (" = Input spin signal data set file");
System.err.println (" = Lower R1 spin-lattice relaxation rate (1/sec)");
System.err.println (" = Upper R1 spin-lattice relaxation rate (1/sec)");
System.err.println (" = Number of R1 intervals");
System.err.println (" = Pixel index");
System.exit (1);
}
}
Ads help maintain this website.