Skip to main content
SpringerLink
Log in

A Gradient Boosted Regression Tree Ensemble Model Using Wavelet Features for Post-acquisition Macromolecular Baseline Isolation from Brain MR Spectra

  • Original Paper
  • Published:
Applied Magnetic Resonance Aims and scope Submit manuscript

Abstract

Broad macromolecular baseline (MMBL) is present throughout the magnetic resonance spectroscopy (MRS) spectrum of brain at short echo-time (TE) acquisitions. These variations cause the metabolite peak quantification and processing difficult in diagnostics. MMBL can provide information for specific disease, as an important biomarker. This presents a requirement of an efficient MMBL isolation method in post-acquisition scenario. The volume of medical dataset available are mostly small- or medium-sized along with the constraint of ground truth. The estimation of MM baseline from a noisy spectrum was treated as an ill-conditioned inverse problem. To address both issues, a novel approach of gradient boosted wavelet-feature tree model in a multioutput-regression framework for MRS spectral fitting was adopted to isolate macromolecular baseline from noisy metabolite spectra, where the inverse problem was learned by training over wavelet coefficients of noisy spectral dataset simulated using basis-set of metabolites and macromolecules. The proposed method performed almost perfectly for the simulated dataset with smaller margins of error, compared to an equivalent CNN model. For the simulated test set, RMSE and SSIM of 0.1623 and 0.9571 respectively were obtained and RMSE of 0.2263 was obtained for in-vivo test set. The fitted peak amplitude for individual MM component within ± 4% of error range over the simulated dataset.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Data Availability

The in-vivo dataset used during the current study are not publicly available due to ethical issues, but data simulated during the study are available from the corresponding author on reasonable request.

References

  1. C. Cudalbu et al., NMR Biomed. (2021). https://doi.org/10.1002/nbm.4393

    Article  Google Scholar 

  2. M. Hoefemann, C.S. Bolliger, D.G. Chong, J.W. van der Veen, R. Kreis, NMR Biomed. (2020). https://doi.org/10.1002/nbm.4328

    Article  Google Scholar 

  3. A.M. Wright, S. Murali-Manohar, T. Borbath, N.I. Avdievich, A. Henning, Magn. Reson. Med. (2022). https://doi.org/10.1002/mrm.28958

    Article  Google Scholar 

  4. K. Landheer, M. Gajdošík, M. Treacy, C. Juchem, Magn. Reson. Med. (2020). https://doi.org/10.1002/mrm.28282

    Article  Google Scholar 

  5. S. Murali-Manohar, T. Borbath, A.M. Wright, B. Soher, R. Mekle, A. Henning, Magn. Reson. Med. (2020). https://doi.org/10.1002/mrm.28174

    Article  Google Scholar 

  6. D.K. Deelchand, E.J. Auerbach, N. Kobayashi, M. Marjańska, Magn. Reson. Med. (2018). https://doi.org/10.1002/mrm.26826

    Article  Google Scholar 

  7. M. Marjańska, E.J. Auerbach, R. Valabrègue, P.F. Van de Moortele, G. Adriany, M. Garwood, NMR Biomed. (2012). https://doi.org/10.1002/nbm.1754

    Article  Google Scholar 

  8. R.A. de Graaf, P.B. Brown, S. McIntyre, T.W. Nixon, K.L. Behar, D.L. Rothman, Magn. Reson. Med. (2006). https://doi.org/10.1002/mrm.20946

    Article  Google Scholar 

  9. K.L. Behar, T. Ogino, Magn. Reson. Med. (1991). https://doi.org/10.1002/mrm.1910170202

    Article  Google Scholar 

  10. K.L. Behar, D.L. Rothman, D.D. Spencer, O.A. Petroff, Magn. Reson. Med. (1994). https://doi.org/10.1002/mrm.1910320304

    Article  Google Scholar 

  11. D.G. Chong, R. Kreis, C.S. Bolliger, C. Boesch, J. Slotboom, MAGMA (2011). https://doi.org/10.1007/s10334-011-0246-y

    Article  Google Scholar 

  12. A. Henning, in Handbook of Magnetic Resonance Spectroscopy In Vivo: MRS Theory, Practice and Applications. ed. by P.A. Bottomley, J.R. Griffiths (Wiley, 2016), p.303

    Google Scholar 

  13. M. Považan, B. Strasser, G. Hangel, E. Heckova, S. Gruber, S. Trattnig, W. Bogner, Magn. Reson. Med. (2018). https://doi.org/10.1002/mrm.26778

    Article  Google Scholar 

  14. R. Birch, A.C. Peet, H. Dehghani, M. Wilson, Magn. Reson. Med. (2017). https://doi.org/10.1002/mrm.26103

    Article  Google Scholar 

  15. I.A. Giapitzakis, T. Borbath, S. Murali-Manohar, N. Avdievich, A. Henning, Magn. Reson. Med. (2019). https://doi.org/10.1002/mrm.27467

    Article  Google Scholar 

  16. U. Seeger, U. Klose, I. Mader, W. Grodd, T. Nägele, Magn. Reson. Med. (2003). https://doi.org/10.1002/mrm.10332

    Article  Google Scholar 

  17. R. Bartha, D.J. Drost, P.C. Williamson, NMR Biomed. (1999). https://doi.org/10.1002/(SICI)1099-1492(199906)12:4%3c205::AID-NBM558%3e3.0.CO;2-1

    Article  Google Scholar 

  18. W.W.F. Pijnappel, A. Van den Boogaart, R. De Beer, D. Van Ormondt, J. Magn. Reson. (1969). (1992). https://doi.org/10.1016/0022-2364(92)90241-X

    Article  Google Scholar 

  19. T. Laudadio, N. Mastronardi, L. Vanhamme, P. Van Hecke, S. Van Huffel, J. Magn. Reson. (2002). https://doi.org/10.1006/jmre.2002.2593

    Article  Google Scholar 

  20. H. Barkhuijsen, R. De Beer, D. Van Ormondt, J Magn Reson. (1969) (1987). https://doi.org/10.1016/0022-2364(87)90023-0

    Article  Google Scholar 

  21. L. Vanhamme, A. van den Boogaart, S. Van Huffel, J Magn Reson. (1997). https://doi.org/10.1006/jmre.1997.1244

    Article  Google Scholar 

  22. H.H. Lee, H. Kim, Magn. Reson. Med. (2019). https://doi.org/10.1002/mrm.27727

    Article  Google Scholar 

  23. H.H. Lee, H. Kim, Magn. Reson. Med. (2017). https://doi.org/10.1002/mrm.26502

    Article  Google Scholar 

  24. S.S. Gurbani, S. Sheriff, A.A. Maudsley, H. Shim, L.A. Cooper, Magn. Reson. Med. (2019). https://doi.org/10.1002/mrm.27641

    Article  Google Scholar 

  25. M. Marjanska, D. K. Deelchand, R. Kreis, MRS fitting challenge data setup by ISMRM MRS study group. Data Repository for the University of Minnesota. https://experts.umn.edu/en/datasets/mrs-fitting-challenge-data-setup-by-ismrm-mrs-study-group. (2021). Accessed 12 May, 2022.

  26. I.W. Selesnick, R.G. Baraniuk, N.C. Kingsbury, IEEE Signal Process. Mag. (2005). https://doi.org/10.1109/MSP.2005.1550194

    Article  Google Scholar 

  27. T. Chen, C. Guestrin, In: Proceedings of the 22nd acm sigkdd International Conference on Knowledge Discovery and Data Mining. (2016), p. 785–794. https://doi.org/10.48550/arXiv.1603.02754.

  28. R.A. De Graaf, In-vivo NMR spectroscopy: principles and techniques, 3rd edn. (Wiley, New York, 2007)

    Book  Google Scholar 

  29. V. Govindaraju, K. Young, A.A. Maudsley, NMR Biomed. (2000). https://doi.org/10.1002/1099-1492(200005)13:3%3c129::AID-NBM619%3e3.0.CO;2-

    Article  Google Scholar 

  30. T.L. Perry, S. Hansen, K. Berry, C. Mok, D. Lesk, J. Neurochem. (1971). https://doi.org/10.1111/j.1471-4159.1971.tb11980.x

    Article  Google Scholar 

  31. C. Sagar, D.K. Singh, N. Sharma, J. Appl. Spectrosc. (2022). https://doi.org/10.1007/s10812-022-01393-7

    Article  Google Scholar 

  32. J. Bergstra, D. Yamins, D. Cox, In PMLR (2013) p. 115–123. https://doi.org/10.48550/arXiv.1209.5111.

  33. D. P. Kingma, J. Ba, arXiv preprint. (2014) https://doi.org/10.48550/arXiv.1412.6980.

  34. R. Kreis, V. Boer, I.Y. Choi, C. Cudalbu et al., NMR Biomed. (2021). https://doi.org/10.1002/nbm.4347

    Article  Google Scholar 

  35. I.A. Giapitzakis, N. Avdievich, A. Henning, Magn Reson Med. (2018). https://doi.org/10.1002/mrm.27070

    Article  Google Scholar 

  36. R. Shwartz-Ziv, A. Armon, Inf Fusion. (2022) https://doi.org/10.48550/arXiv.2106.03253.

  37. S. Elsayed, D. Thyssens, A. Rashed, H. S. Jomaa, L. Schmidt-Thieme, arXiv preprint. (2021) https://doi.org/10.48550/arXiv.2101.02118.

Download references

Acknowledgements

The authors express their sincere gratitude to the Indian Institute of Technology (BHU), Varanasi, UP, India, and Apex Super-speciality Hospital and Post Graduate Institute, Varanasi, Uttar Pradesh, India, for supporting the work.

Funding

None.

Author information

Authors and Affiliations

  1. School of Biomedical Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, UP, 221005, India

    Chiranjeev Sagar & Neeraj Sharma

  2. Department of Oncology, Apex Super-Speciality Hospital and Post Graduate Institute, Varanasi, Uttar Pradesh, India

    Deepak Kumar Singh

Authors
  1. Chiranjeev Sagar
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Deepak Kumar Singh
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Neeraj Sharma
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

CS: principal author, conceptualization, methodology, data curation, software, formal analysis, investigation, writing—original draft. DKS: conceptualization, resources, validation, review & editing, supervision. NS: conceptualization, validation, resources, review & editing, supervision, project administration.

Corresponding author

Correspondence to Chiranjeev Sagar.

Ethics declarations

Conflict of Interest

The author(s) declared no potential conflicts of interest concerning the research, authorship, and/or publication of this article.

Ethical Approval

This is a retrospective dataset study. So, the requirement of patient consent was waived off by the institute ethical committee.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

See Table 4, Figs. 5, 6

Table 4 Parameters used to generate Macromolecular spectral baseline (expected at B0 = 3 T). (Small variations in values can be found in different literature)
Full size table
Fig. 5
figure 5

CNN model architecture used for comparative study

Full size image
Fig. 6
figure 6

MM spectrum isolation from residual water peak included noisy MR spectra. Top: Noisy spectra with residual waterpeak; Middle: Superimposed individual components contribution of Metabolites, Macromolecular Baseline (GT) and Residual waterpeak, [Blue: Noisy spectrum, Green: GT-MM, Orange: Residual waterpeak]; Bottom: Isolated MM from nosiy spectra with residual waterpeak (GT vs Proposed method) [Blue: GT-MM, Orange: Fitted MM from the proposed method],

Full size image

Results for MM isolation from spectra with residual waterpeak:

Noisy data with Res. Water content

Res. waterpeak height <  = 3 × Highest peak of training spectra

Res. waterpeak height > 3 × Highest peak of training spectra

RMSE

SSIM

R2-value

RMSE

SSIM

R2-value

Proposed method

0.2988

0.8973

0.7836

 > 0.55

 < 0.74

 < 10

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sagar, C., Singh, D.K. & Sharma, N. A Gradient Boosted Regression Tree Ensemble Model Using Wavelet Features for Post-acquisition Macromolecular Baseline Isolation from Brain MR Spectra. Appl Magn Reson 54, 637–655 (2023). https://doi.org/10.1007/s00723-023-01537-8

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00723-023-01537-8

Search

Navigation