Methods used for measuring atrophy in Multiple Sclerosis

Location/generic methodology	Method	Advantages	Disadvantages
Brain/linear and regional measures	Third ventricle width	Easy to implement; rapid analysis; standard acquisition methods; enables targeting of eloquent regions	Limited anatomical scope; may miss subtle effects; may exhibit user bias; high user input
	Third ventricle volume
	Brain width
	Corpus callosum width
	Volume on central brain slices
	Stereology
Brain/whole brain segmentation approaches	CSF volumes	Increased automation reduces user bias and user input; generally higher measurement precision	Complex analysis methods; possibly more complex acquisition schemes
	BPF
	WBR
	BICCR
	Fuzzy connectedness
	Probabilistic segmentation (SPM)
	TDS
	SIENAX
Brain/registration‐based methods	MIDAS	Regional atrophy may become apparent	Complex analysis methods; limited application to multiple sclerosis to date
	Voxel‐based morphometry
	SIENA
Spinal cord	Manual outlining	Straightforward	Possible user bias; high user input
	Semi‐automated outline of 3D axial images	Precise
	Automated whole cord volume measurement	Little user input	Complex analysis methods; limited application to date
Optic nerve	Manual outlining	Straightforward	Possible user bias; high user input
	Semi‐automated outline of 3D axial images	Precise
	Automated whole nerve volume measurement	Little user input	Complex analysis methods; limited application to date

TDS = template‐driven segmentation; MIDAS = Medical Image Display and Analysis Software.

Source: David H. Miller, Frederik Barkhof, Joseph A. Frank, Geoffrey J. M. Parker, Alan J. Thompson. Measurement of atrophy in multiple sclerosis: pathological basis, methodological aspects and clinical relevance . Brain 2002. DOI: http://dx.doi.org/10.1093/brain/awf177