Skip to main content

Advertisement

Log in

A deep learning approach to evaluate intestinal fibrosis in magnetic resonance imaging models

  • Original Article
  • Published:
Neural Computing and Applications Aims and scope Submit manuscript

Abstract

Fibrosis may be introduced as a severe complication of inflammatory bowel disease (IBD). This is a particular disorder causing luminal narrowing and stricture formation in the inflamed bowel wall of a patient denoting, possibly, need for surgery. Thus, the development of treatments reducing fibrosis is an urgent issue to be addressed in IBD. In this context, we require the finding and development of biomarkers of intestinal fibrosis. Potential candidates such as microRNAs, gene variants or fibrocytes have shown controversial results on heterogeneous sets of IBD patients. Magnetic resonance imaging (MRI) has been already successfully proven in the recognition of fibrosis. Nevertheless, while there are no numerical models capable of systematically reproducing experiments, the usage of MRI could not be considered a standard in the inflammatory domain. Hence, there is an importance of deploying new sequence combinations in MRI methods that enable learning reproducible models. In this work, we provide reproducible deep learning models of intestinal fibrosis severity scores based on MRI novel radiation-induced rat model of colitis that incorporates some unexplored sequences such as the flow-sensitive alternating inversion recovery or diffusion imaging. The results obtained return an \(87.5\%\) of success in the prediction of MRI scores with an associated mean-square error of 0.12. This approach offers practitioners a valuable tool to evaluate antifibrotic treatments under development and to extrapolate such noninvasive MRI scores model to patients with the aim of identifying early stages of fibrosis improving patients’ management.

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

Access this article

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

Similar content being viewed by others

References

  1. Agarap A (2018) Deep learning using rectified linear units (ReLU). CoRR abs/1803.08375. arXiv:1803.08375

  2. Apley D (2001) Visualizing the effects of predictor variables in black box supervised learning models. Ann Stat 8(2):1189–1232

    Google Scholar 

  3. Baldi P (2012) Autoencoders, unsupervised learning, and deep architectures. In: JMLR: workshop and conference proceedings, vol 27, pp 37–50

  4. Barkmeier D, Dillman J, Al-Hawary M, Heider A, Davenport M, Smith E, Adler J (2016) Mr enterography–histology comparison in resected pediatric small bowel crohn disease strictures: can imaging predict fibrosis? Pediatr Radiol 4:498–507

    Article  Google Scholar 

  5. Ben Amar M, Bianca C (2016) Towards a unified approach in the modeling of fibrosis: a review with research perspectives. Phys Life Rev 3:917–932

    Google Scholar 

  6. Breiman L (2001) Random forests. Mach Learn 45:5–32

    Article  Google Scholar 

  7. de Boer P, Kroese D, Mannor S, Rubinstein R (2005) A tutorial on the cross-entropy method. Ann Oper Res 134(1):19–67. https://doi.org/10.1007/s10479-005-5724-z

    Article  MathSciNet  MATH  Google Scholar 

  8. Dillman J, Swanson S, Johnson L, Moons D, Adler J, Stidham R, Higgins P (2015) Comparison of noncontrast MRI magnetization transfer and t2-weighted signal intensity ratios for detection of bowel wall fibrosis in a crohn’s disease animal model. J Mag Res Imaging 42:801–810

    Article  Google Scholar 

  9. Ellmann S, Langer V, Britzen-Laurent N, Hildner K, Huber C, Tripal P, Seyler L, Waldner M, Uder M, Stürzl M, Bäuerle T (2018) Application of machine learning algorithms for multiparametric MRI-based evaluation of murine colitis. PLoS ONE 13(10):1–17. https://doi.org/10.1371/journal.pone.0206576

    Article  Google Scholar 

  10. Freund Y, Schapire R (1996) Experiments with a new boosting algorithm. Morgan Kaufmann Publishers Inc, Burlington

    Google Scholar 

  11. Friedman A, Hao W (2017) Mathematical modeling of liver fibrosis. Math Biosci Eng 14:143–147

    Article  MathSciNet  Google Scholar 

  12. Friedman J (2016) Greedy function approximation: a gradient boosting machine. arXiv, pp 1–36

  13. Gedeon T (1997) Data mining of inputs: analysing magnitude and functional measures. Int J Neural Syst 8(2):209–218

    Article  Google Scholar 

  14. Gee M, Mukesh G, Harisinghani M (2011) MRI in patients with inflammatory bowel disease. J Magn Reson Imaging 33:527–34. https://doi.org/10.1002/jmri.22504

    Article  Google Scholar 

  15. Giuffrida P, Pinzani M, Corazza G, Di Sabatino A (2016) Biomarkers of intestinal fibrosis-one step towards clinical trials for structuring inflammatory bowel disease. United Eur Gastroenterol J 4:523–530

    Article  Google Scholar 

  16. Goldstein A, Kapelner A, Bleich J, Pitkin E (2014) Peeking inside the black box: visualizing statistical learning with plots of individual conditional expectation. arXiv abs/1212.5701. arXiv:1309.6392v2

  17. Goodfellow I, Bengio Y, Courville A (2016) Deep learning. MIT Press, Cambridge

    MATH  Google Scholar 

  18. Gordon R (1994) The integrals of Lebesgue, Denjoy, Perron and Henstock, vol 4. American Mathematical Society, Providence

    MATH  Google Scholar 

  19. Grossman J, Grossman M, Katz R (1980) The first systems of weighted differential and integral calculus. Archimedes Foundation, Rockport

    MATH  Google Scholar 

  20. Haas K, Rubesova E, Bass D (2016) Role of imaging in the evaluation of inflammatory bowel disease: How much is too much? World J Radiol 8(2):124–131. https://doi.org/10.4329/wjr.v8.i2.124

    Article  Google Scholar 

  21. Higgins P, Fletcher J (2015) Characterization of inflammation and fibrosis in crohn’s disease lesions by magnetic resonance imaging. Am J Gastroenterol 110:441–443

    Article  Google Scholar 

  22. Huang W, Hong H, Bian K, Zhou X, Song G, Xie K (2015) Improving deep neural network ensembles using reconstruction error. In: 2015 international joint conference on neural networks (IJCNN), pp 1–7. https://doi.org/10.1109/IJCNN.2015.7280524

  23. Lattouf R, Younes R, Lutomski D, Naaman N, Godeau G, Senni K, Changotade S (2014) Picrosirius red staining: a useful tool to appraise collagen networks in normal and pathological tissues. J Histochem Cytochem 62(10):751–758. https://doi.org/10.1369/0022155414545787

    Article  Google Scholar 

  24. Melkumova L, Shatskikh S (2017) Comparing ridge and LASSO estimators for data analysis. In: 3rd international conference on information technology and nanotechnology, vol 201, pp 746–755

  25. Morilla I, Doblas S, Garteiser P, Zappa M, Ogier-Denis E (2017) Scores of intestinal fibrosis from wavelet-based magnetic resonance imaging models. In: Rojas I, Ortuño F (eds) Bioinform Biomed Eng. Springer, Cham, pp 569–578

    Chapter  Google Scholar 

  26. Morilla I, Uzzan M, Laharie D, Casals-Hatem D, Denost Q, Daniel F, Belleannee G, Bouhnik Y, Wainrib G, Panis Y, Ogier-Denis E, Treton X (2019) Colonic microrna profiles, identified by a deep learning algorithm, that predict responses to therapy of patients with acute severe ulcerative colitis. Clin Gastroenterol Hepatol 17:905–913

    Article  Google Scholar 

  27. Murphy K (2012) Machine learning: a probabilistic perspective. MIT Press, Cambridge

    MATH  Google Scholar 

  28. Panes J, Bouhnik Y, Reinisch W et al (2013) Imaging techniques for assessment of inflammatory bowel disease: joint ECCO and ESGAR evidence-based consensus guidelines. J Crohn’s Colitis 7:556–585

    Article  Google Scholar 

  29. Rimola J, Planell N, Rodriguez S, Delgado S, Ordas I, Ramirez-Morros A, Ayuso C, Aceituno M, Ricart E, Jauregui-Amezaga A, Panes J, Cuatrecasas M (2015) Characterization of inflammation and fibrosis in Crohn’s disease lesions by magnetic resonance imaging. Am J Gastroenterol 3:432–440

    Article  Google Scholar 

  30. Singh G, Sachan M (2014) Multi-layer perceptron (MLP) neural network technique for offline handwritten Gurmukhi character recognition. In: 2014 IEEE international conference on computational intelligence and computing research, pp 1–5. https://doi.org/10.1109/ICCIC.2014.7238334

  31. Wei R, Wang J, Wang X, Xie G, Wang Y, Zhang H, Peng C, Rajani C, Kwee S, Liu P, Jia W (2018) Clinical prediction of hbv and hcv related hepatic fibrosis using machine learning. EBioMedicine 35:124–132

    Article  Google Scholar 

  32. Xue-hua L, Ren M, Si-yun H, Can-hui S, Qing-hua C, Zhuang-nian F, Zhong-wei Z, Li H, Jin-jiang L, Yu-jun C, Rimola J, Rieder F, Min-hu C, Shi-ting F, Zi-ping L (2018) Characterization of degree of intestinal fibrosis in patients with Crohn disease by using magnetization transfer MR imaging. Radiology 287(2):494–503. https://doi.org/10.1148/radiol.2017171221 PMID: 29357272

    Article  Google Scholar 

  33. Yu Y, Wang J, Ng C, Ma Y, Mo S, Fong E, Song Z, Xie Y, Si K, Wee A, Welsch R, So P, Yu H (2018) Deep learning enables automated scoring of liver fibrosis stages. Sci Rep 8:2045–2322

    Article  Google Scholar 

  34. Zappa M (2017) Intérêt de l’IRM pour l’évaluation de la fibrose intestinale dans un modèle murin de colite radio-induite. Ph.D. thesis, Physiologie et physiopathologie Sorbonne Paris Cité 2017. http://www.theses.fr/2017USPCC248/document

  35. Zappa M, Doblas S, Cazals-Hatem D, Fabien M, Lavigne J, Fanny D, Jallane A, Philippe G, Vilgrain V, Ogier-Denis E, Van Beers B (2017) Quantitative MRI in murine radiation-induced rectocolitis: comparison with histopathological inflammation score. NMR in Biomed 31(4):e3897. https://doi.org/10.1002/nbm.3897

    Article  Google Scholar 

  36. Zappa M, Stefanescu C, Cazals-Hatem D, Bretagnol F, Deschamps L, Attar A, Larroque B, Treton X, Panis Y, Vilgrain V, Bouhnik Y (2011) Which magnetic resonance imaging findings accurately evaluate inflammation in small bowel Crohn’s disease? A retrospective comparison with surgical pathologic analysis. Inflamm Bowel Dis 4:984–993

    Article  Google Scholar 

  37. Zeiler M (2012) ADADELTA: an adaptive learning rate method. CoRR. arXiv:1212.5701

Download references

Acknowledgements

We would like to thank Magaly Zappa and Eric Ogier-Denis for providing me with the raw MRI sequences and for their valuable discussions in the manuscript writing. We acknowledge the financial support by Institut National de la Santé et de la Recherche Medicale (INSERM), Inception IBD, Inserm-Transfert, Association Franois Aupetit (AFA), Université Diderot Paris 7, and the Investissements d’Avenir programme ANR-11-IDEX-0005-02 and 10-LABX-0017, Sorbonne Paris Cité, Laboratoire d’excellence INFLAMEX.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Ian Morilla.

Ethics declarations

Conflict of interest

The author declares that he has no conflict of interest.

Research involving human participants and/or animals

The studies the data derive from were conducted in compliance with the French regulations for animals experimentation (Ministry of Agriculture, Act 87-848, 19 Oct 1987) and approved by both the Institute for Radiological Protection and Nuclear Safety Ethics Committee and the Ministry for Higher Education and Scientific Research (Protocol P15-04).

Additional information

Publisher's Note

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

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 2629 KB)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Morilla, I. A deep learning approach to evaluate intestinal fibrosis in magnetic resonance imaging models. Neural Comput & Applic 32, 14865–14874 (2020). https://doi.org/10.1007/s00521-020-04838-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00521-020-04838-2

Keywords

Navigation