Interpolator for gradient calculation is linear, i.e. not using the interpolator set for the cost function

[romangrothausmann]

Trying to solve #61 I found that the interpolator that is used to calculate the gradient is always linear

ITKMinimalPathExtraction/include/itkPhysicalCentralDifferenceImageFunction.h

Line 86 in 2ff1ab5

using InterpolateImageFunctionType = LinearInterpolateImageFunction<TInputImage, TCoordRep>;

and not using the interpolator that is (possibly) set for the cost function

ITKMinimalPathExtraction/include/itkSingleImageCostFunction.h

Lines 105 to 107 in 2ff1ab5

	/** Get/set the Interpolator. */
	itkSetObjectMacro( Interpolator, InterpolatorType );
	itkGetConstObjectMacro( Interpolator, InterpolatorType );

It seems the cost function interpolator is used in case iterateNeighborhoodOptimizer (which uses GetValue) is employed

ITKMinimalPathExtraction/include/itkSingleImageCostFunction.hxx

Line 102 in 2ff1ab5

return static_cast<MeasureType>( m_Interpolator->Evaluate(point) );

but not when the gradient is computed (which uses GetDerivative)

ITKMinimalPathExtraction/include/itkSingleImageCostFunction.hxx

Line 137 in 2ff1ab5

output = m_GradientImageFunction->Evaluate(point);

This however does not seem to be the issue of #61 (where a linear interpolator was used anyway), so posting as an issue on its own.