Uncertainty modelling in deep learning for safer neuroimage enhancement: Demonstration in diffusion MRI

Tanno, Ryutaro; Worrall, Daniel E.; Kaden, Enrico; Ghosh, Aurobrata; Grussu, Francesco; Bizzi, Alberto; Sotiropoulos, Stamatios N.; Criminisi, Antonio; Alexander, Daniel C.

doi:10.1016/j.neuroimage.2020.117366

Cited by 89 publications

(83 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Alternatives or complements to the fully-connected DNN architecture should also be explored. Promising avenues include the use of dropout (Gal and Ghahramani, 2016;Tanno et al, 2021) or deep ensemble strategies (Lakshminarayanan et al, 2016;Qin et al, 2021) as a means to derive uncertainty metrics, the use of network structures inspired by non-learning-based iterative fitting frameworks (Ye, 2017), or use denoising networks (Fadnavis et al, 2020;Wang et al, 2019) to minimize the amount of noise present in the data that is supplied to the function-fitting DNN. However, the sensitivity analysis presented here should be applied to any new fitting strategy to test specificity to change in single model parameters.…”

Section: Discussionmentioning

confidence: 99%

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Martins

Nilsson

Lampinen

et al. 2021

Preprint

View full text Add to dashboard Cite

Specific features of white-matter microstructure can be investigated by using biophysical models to interpret relaxation-diffusion MRI brain data. Although more intricate models have the potential to reveal more details of the tissue, they also incur time-consuming parameter estimation that may converge to inaccurate solutions due to a prevalence of local minima in a degenerate fitting landscape. Machine-learning fitting algorithms have been proposed to accelerate the parameter estimation and increase the robustness of the attained estimates. So far, learning-based fitting approaches have been restricted to lower-dimensional microstructural models where dense sets of training data are easy to generate. Moreover, the degree to which machine learning can alleviate the degeneracy problem is poorly understood. For conventional least-squares solvers, it has been shown that degeneracy can be avoided by acquisition with optimized relaxation-diffusion-correlation protocols that include tensor-valued diffusion encoding; whether machine-learning techniques can offset these acquisition requirements remains to be tested. In this work, we employ deep neural networks to vastly accelerate the fitting of a recently introduced high-dimensional relaxation-diffusion model of tissue microstructure. We also develop strategies for assessing the accuracy and sensitivity of function fitting networks and use those strategies to explore the impact of acquisition protocol design on the performance of the network. The developed learning-based fitting pipelines were tested on relaxation-diffusion data acquired with optimized and sub-sampled acquisition protocols. We found no evidence that machine-learning algorithms can by themselves replace a careful design of the acquisition protocol or correct for a degenerate fitting landscape.

show abstract

Section: Discussionmentioning

confidence: 99%

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Martins

Nilsson

Lampinen

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Furthermore, our findings point out the importance of computing uncertainty, cf. [41], in MLbased estimation, particularly when ML is used to compensate for lower quality data.…”

Section: Summary Of Resultsmentioning

confidence: 99%

Training Data Distribution Significantly Impacts the Estimation of Tissue Microstructure with Machine Learning

Gyori

Palombo

Clark

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Purpose: Supervised machine learning (ML) provides a compelling alternative to traditional model fitting for parameter mapping in quantitative MRI. The aim of this work is to demonstrate and quantify the effect of different training strategies on the accuracy and precision of parameter estimates when supervised ML is used for fitting. Methods: We fit a two-compartment biophysical model to diffusion measurements from in-vivo human brain, as well as simulated diffusion data, using both traditional model fitting and supervised ML. For supervised ML, we train several artificial neural networks, as well as random forest regressors, on different distributions of ground truth parameters. We compare the accuracy and precision of parameter estimates obtained from the different estimation approaches using synthetic test data. Results: When the distribution of parameter combinations in the training set matches those observed in similar data sets, we observe high precision, but inaccurate estimates for atypical parameter combinations. In contrast, when training data is sampled uniformly from the entire plausible parameter space, estimates tend to be more accurate for atypical parameter combinations but may have lower precision for typical parameter combinations. Conclusion: This work highlights the need to consider the choice of training data when deploying supervised ML for estimating microstructural metrics, as performance depends strongly on the training-set distribution. We show that high precision obtained using ML may mask strong bias, and visual assessment of the parameter maps is not sufficient for evaluating the quality of the estimates.

show abstract

“…However, blindly trusting the result (output) of such a model risks different undetected failures (e.g., removal of structures and most important features) [25]. As expressed by Tanno et al [26] these predictive failures have two main reasons:…”

Section: Uncertainty Quantification In Deep Learningmentioning

confidence: 99%

Automated Detection of Presymptomatic Conditions in Spinocerebellar Ataxia Type 2 Using Monte Carlo Dropout and Deep Neural Network Techniques with Electrooculogram Signals

Stoean

Atencia

et al. 2020

Sensors

View full text Add to dashboard Cite

Application of deep learning (DL) to the field of healthcare is aiding clinicians to make an accurate diagnosis. DL provides reliable results for image processing and sensor interpretation problems most of the time. However, model uncertainty should also be thoroughly quantified. This paper therefore addresses the employment of Monte Carlo dropout within the DL structure to automatically discriminate presymptomatic signs of spinocerebellar ataxia type 2 in saccadic samples obtained from electrooculograms. The current work goes beyond the common incorporation of this special type of dropout into deep neural networks and uses the uncertainty derived from the validation samples to construct a decision tree at the register level of the patients. The decision tree built from the uncertainty estimates obtained a classification accuracy of 81.18% in automatically discriminating control, presymptomatic and sick classes. This paper proposes a novel method to address both uncertainty quantification and explainability to develop reliable healthcare support systems.

show abstract

Uncertainty modelling in deep learning for safer neuroimage enhancement: Demonstration in diffusion MRI

Cited by 89 publications

References 53 publications

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Training Data Distribution Significantly Impacts the Estimation of Tissue Microstructure with Machine Learning

Automated Detection of Presymptomatic Conditions in Spinocerebellar Ataxia Type 2 Using Monte Carlo Dropout and Deep Neural Network Techniques with Electrooculogram Signals

Contact Info

Product

Resources

About