Robust machine learning segmentation for large-scale analysis of heterogeneous clinical brain MRI datasets

Abstract

Every year, millions of brain MRI scans are acquired in hospitals, which is a figure considerably larger than the size of any research dataset. Therefore, the ability to analyze such scans could transform neuroimaging research. Yet, their potential remains untapped since no automated algorithm is robust enough to cope with the high variability in clinical acquisitions (MR contrasts, resolutions, orientations, artifacts, and subject populations). Here, we present SynthSeg + , an AI segmentation suite that enables robust analysis of heterogeneous clinical datasets. In addition to whole-brain segmentation, SynthSeg + also performs cortical parcellation, intracranial volume estimation, and automated detection of faulty segmentations (mainly caused by scans of very low quality). We demonstrate SynthSeg + in seven experiments, including an aging study on 14,000 scans, where it accurately replicates atrophy patterns observed on data of much higher quality. SynthSeg + is publicly released as a ready-to-use tool to unlock the potential of quantitative morphometry.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 58

1. 1.

   J. Ashburner, K. Friston, Unified segmentation. NeuroImage 26, 39–51 (2005 ).

   Not in our library10.1016/j.neuroimage.2005.02.018

2. 2.

   M. Jenkinson, C. Beckmann, T. Behrens, M. Woolrich, S. Smith, FSL. Neuroimage 62, 782–790 (2012 ).

   Not in our library10.1016/j.neuroimage.2011.09.015

3. 3.

   B. Fischl, FreeSurfer. NeuroImage 62, 774–781 (2012 ).

   Not in our library10.1016/j.neuroimage.2012.01.021

4. 4.

   O. Oren, E. Kebebew, J. Ioannidis, Curbing unnecessary and wasted diagnostic imaging. JAMA 321, 245–246 (2019 ).

   Not in our library10.1001/jama.2018.20295

5. 5.

   D. Hibar, Common genetic variants influence human subcortical brain structures. Nature 520, 224–229 (2015 ).

   Not in our library10.1038/nature14101

6. 6.
   Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank · 2018

   No notice
7. 7.

   C. R. Jack, The Alzheimer’s Disease Neuroimaging Initiative (ADNI): MRI methods. J. Magn. Reson. Imaging: JMRI 27, 685–691 (2008 ).

   Not in our library10.1002/jmri.21049

8. 8.

   Alzheimer’s Disease Neuroimaging Initiative, ADNI participant demographic. https://adni.loni.usc.edu/data-samples/adni-participant-demographic/ (Accessed 1 February 2023).

   Not in our library

9. 9.
   Comparison of Sociodemographic and Health-Related Characteristics of UK Biobank Participants With Those of the General Population · 2017

   No notice
10. 10.

    O. Puonti, J. E. Iglesias, K. Van Leemput, Fast and sequence-adaptive whole-brain segmentation using parametric Bayesian modeling. NeuroImage 143, 235–249 (2016 ).

    Not in our library10.1016/j.neuroimage.2016.09.011

11. 11.

    H. Choi, D. Haynor, Y. Kim, Partial volume tissue classification of multichannel magnetic resonance images-a mixel model. IEEE Trans. Med. Imaging 10, 395–407 (1991 ).

    Not in our library10.1109/42.97590

12. 12.

    K. Van Leemput, F. Maes, D. Vandermeulen, P. Suetens, A unifying framework for partial volume segmentation of brain MR images. IEEE Trans. Med. Imaging 22, 105–119 (2003 ).

    Not in our library10.1109/tmi.2002.806587

13. 13.

    F. Milletari, N. Navab, S. Ahmadi, “V-Net: Fully convolutional neural networks for volumetric medical image segmentation” in International Conference on 3D Vision (2016), pp. 565–571.

    Not in our library

14. 14.

    O. Ronneberger, P. Fischer, T. Brox, “U-Net: Convolutional networks for biomedical image segmentation” in Medical Image Computing and Computer Assisted Intervention (2015), pp. 234–241.

    Not in our library

15. 15.

    S. Pan, Q. Yang, A survey on transfer learning. IEEE Trans. Knowl. Data Eng. 22, 45–59 (2010 ).

    Not in our library

16. 16.

    M. Ghafoorian et al., “Transfer learning for domain adaptation in MRI: Application in brain lesion segmentation” in Medical Image Computing and Computer Assisted Intervention (2017), pp. 516–524.

    Not in our library

17. 17.
    Deep Learning for Brain MRI Segmentation: State of the Art and Future Directions · 2017

    No notice
18. 18.

    N. Karani, E. Erdil, K. Chaitanya, E. Konukoglu, Test-time adaptable neural networks for robust medical image segmentation. Med. Image Anal. 68, 2–15 (2021 ).

    Not in our library10.1016/j.media.2020.101907

19. 19.

    L. Zhang, Generalizing deep learning for medical image segmentation to unseen domains via deep stacked transformation. IEEE Trans. Med. Imaging 39, 2531–2540 (2020 ).

    Not in our library10.1109/tmi.2020.2973595

20. 20.

    C. Chen, Q. Dou, H. Chen, J. Qin, P. A. Heng, Synergistic image and feature adaptation: Towards cross-modality domain adaptation for medical image segmentation. AAAI Conf. Artif. Intell. 33, 65–72 (2019 ).

    Not in our library

21. 21.

    B. Billot et al., SynthSeg: Domain Randomisation for Segmentation of Brain Scans of any Contrast and Resolution arXiv [Preprint] [cs] (2021). http://arxiv.org/abs/2107.09559 (Accessed 4 January 2023).

    Not in our library

22. 22.

    J. Tobin et al., “Domain randomization for transferring deep neural networks from simulation to the real world” in IEEE/RSJ International Conference on Intelligent Robots and Systems (2017), pp. 23–30.

    Not in our library

23. 23.

    F. Isensee, P. Jaeger, S. Kohl, J. Petersen, K. Maier-Hein, nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation. Nat. Methods 18, 203–211 (2021 ).

    Not in our library10.1038/s41592-020-01008-z

24. 24.

    H. R. Roth, An application of cascaded 3D fully convolutional networks for medical image segmentation. Comput. Med. Imaging Graphics 66, 90–99 (2018 ).

    Not in our library10.1016/j.compmedimag.2018.03.001

25. 25.

    M. Nosrati, G. Hamarneh, Incorporating prior knowledge in medical image segmentation: A survey arXiv [Preprint] [cs] (2016). http://arxiv.org/abs/1607.01092 (Accessed 15 July 2016).

    Not in our library

26. 26.

    O. Oktay, Anatomically constrained neural networks (ACNNs): Application to cardiac image enhancement and segmentation. IEEE Trans. Med. Imaging 37, 384–395 (2018 ).

    Not in our library10.1109/tmi.2017.2743464

27. 27.

    M. Sinclair, Atlas-ISTN: Joint segmentation, registration and atlas construction with image-and-spatial transformer networks. Med. Image Anal. 78, 102–123 (2022 ).

    Not in our library10.1016/j.media.2022.102383

28. 28.

    A. Larrazabal, C. Martinez, B. Glocker, E. Ferrante, Post-DAE. IEEE Trans. Med. Imaging 39, 3813–3820 (2020).

    Not in our library10.1109/tmi.2020.3005297

29. 29.

    V. Valindria, Reverse classification accuracy: Predicting segmentation performance in the absence of ground truth. IEEE Trans. Med. Imaging 36, 1597–1606 (2017 ).

    Not in our library10.1109/tmi.2017.2665165

30. 30.

    T. Kohlberger, V. Singh, C. Alvino, C. Bahlmann, L. Grady, “Evaluating segmentation error without ground truth” in Medical Image Computing and Computer Assisted Intervention (2019), pp. 528–536.

    Not in our library10.1007/978-3-642-33415-3_65

31. 31.

    F. Liu, Y. Xia, D. Yang, A. Yuille, D. Xu, “An alarm system for segmentation algorithm based on shape model” in ICCV (2019), pp. 10652–10661.

    Not in our library

32. 32.

    S. Wang et al., “Deep generative model-based quality control for cardiac MRI segmentation” in Medical Image Computing and Computer Assisted Intervention (2020), pp. 88–97.

    Not in our library

33. 33.

    B. Billot, C. Magdamo, S. E. Arnold, S. Das, J. E. Iglesias, “Robust segmentation of brain MRI in the wild with hierarchical CNNs and no retraining” in Medical Image Computing and Computer Assisted Intervention (2022), pp. 538–548.

    Not in our library

34. 34.

    E. DeLong, D. DeLong, D. Clarke-Pearson, Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics 44, 837–845 (1988 ).

    Not in our library

35. 35.

    A. P. Bradley, The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit. 30, 1145–1159 (1997 ).

    Not in our library

36. 36.
    Accurate automatic estimation of total intracranial volume: A nuisance variable with less nuisance · 2015

    No notice
37. 37.

    K. Gosche, J. Mortimer, C. Smith, W. Markesbery, D. Snowdon, Hippocampal volume as an index of Alzheimer neuropathology. Neurology 58, 1476–1482 (2002 ).

    Not in our library10.1212/wnl.58.10.1476

38. 38.

    J. Cohen, Statistical Power Analysis for the Behavioural Sciences (Routledge Academic, 1988).

    Not in our library

39. 39.
    Brain charts for the human lifespan · 2022

    Corrected
40. 40.

    P. Coupé, G. Catheline, E. Lanuza, J. Manjón, Towards a unified analysis of brain maturation and aging across the entire lifespan: A MRI analysis. Hum. Brain Mapp. 38, 5501–5518 (2017 ).

    Not in our library10.1002/hbm.23743

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.