Rapid generation of biexponential and diffusional kurtosis maps using multi-layer perceptrons: a preliminary experience

  • Ludovico Minati
Short Communication



To investigate whether multi-layer perceptrons (MLPs) could be used to determine biexponential and diffusional kurtosis model parameters directly from diffusion-weighted images.

Materials and methods

Model parameters were determined with least-squares fitting and with MLPs. The corresponding estimates were compared with linear regressions, t tests and Levene’s tests. Residuals were also compared.


Strong linear correlation was found for all parameters. MLP estimates were unbiased for the biexponential but not for the kurtosis model, and generally had smaller variance. Residuals were smaller for MLP estimates. The maps generated by the two methods were visually very similar.


Multi-layer perceptrons are potentially useful as a curve fitting method for these models.


Diffusion-tensor imaging (DTI) Biexponential model Diffusional kurtosis imaging (DKI) Neural network (NN) Multi-layer perceptron (MLP) 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Niendorf T, Dijkhuizen RM, Norris DG, van Lookeren Campagne M, Nicolay K (1996) Biexponential diffusion attenuation in various states of brain tissue: implications for diffusion-weighted imaging. Magn Reson Med 36: 847–857PubMedCrossRefGoogle Scholar
  2. 2.
    Jensen JH, Helpern JA, Ramani A, Lu H, Kaczynski K (2005) Diffusional kurtosis imaging: the quantification of non-gaussian water diffusion by means of magnetic resonance imaging. Magn Reson Med 53: 1432–1440PubMedCrossRefGoogle Scholar
  3. 3.
    Minati L, Weglarz WP (2007) Physical foundations, models and methods of diffusion magnetic resonance imaging of the brain: a review. Conc Magn Reson A 30: 278–307CrossRefGoogle Scholar
  4. 4.
    Minati L, Aquino D, Rampoldi S, Papa S, Grisoli M, Bruzzone MG, Maccagnano E (2007) Biexponential and diffusional kurtosis imaging, and generalised diffusion-tensor imaging (GDTI) with rank-4 tensors: a study in a group of healthy subjects. Magn Reson Mater Phy 20: 241–253CrossRefGoogle Scholar
  5. 5.
    Kiselev VG, Il’yasov KA (2007) Is the “biexponential diffusion” biexponential?. Magn Reson Med 57: 464–469PubMedCrossRefGoogle Scholar
  6. 6.
    Maier SE, Bogner P, Bajzik G, Mamata H, Mamata Y, Repa I, Jolesz FA, Mulkern RV (2001) Normal brain and brain tumor: multicomponent apparent diffusion coefficient line scan imaging. Radiology 219: 842–849PubMedGoogle Scholar
  7. 7.
    Brugieres P, Thomas P, Maraval A, Hosseini H, Combes C, Chafiq A, Ruel L, Breil S, Peschanski M, Gaston A (2004) Water diffusion compartmentation at high b values in ischemic human brain. AJNR Am J Neuroradiol 25: 692–698PubMedGoogle Scholar
  8. 8.
    Haykin S (1999) Neural networks: a comprehensive foundation, 2nd edn. Prentice-Hall, New JerseyGoogle Scholar
  9. 9.
    Cybenko GV (1989) Approximation by superpositions of a sigmoidal function. Math Contr Sign Sys 2: 303–314CrossRefGoogle Scholar
  10. 10.
    Zell A, Mache N, Sommer T, Korb T (1991) Design of the SNNS neural network simulator. Streichische Artificial-Intelligence-Tagung, 93–102, Informatik-Fachberichte, 287. Springer, WienGoogle Scholar
  11. 11.
    Pfefferbaum A, Adalsteinsson E, Rohlfing T, Sullivan EV (2008) Diffusion tensor imaging of deep gray matter brain structures: effects of age and iron concentration. Neurobiol Aging. Epub ahead of printGoogle Scholar
  12. 12.
    MacKay AL, Whittall KP, Adler J, Li DKB, Paty DW, Graeb D (1994) In vivo visualization of myelin water in brain by magnetic resonance. Magn Reson Med 31: 673–677PubMedCrossRefGoogle Scholar

Copyright information

© ESMRMB 2008

Authors and Affiliations

  1. 1.Science Direction UnitFondazione Istituto Nazionale Neurologico “Carlo Besta” IRCCSMilanoItaly
  2. 2.Neuroradiology DepartmentFondazione Istituto Nazionale Neurologico “Carlo Besta” IRCCSMilanoItaly

Personalised recommendations