Multi-scanner Harmonization of Paired Neuroimaging Data via Structure Preserving Embedding Learning

Torbati, Mahbaneh Eshaghzadeh; Tudorascu, Dana L.; Minhas, Davneet; Maillard, Pauline; DeCarli, Charles; Hwang, Seong Jae

doi:10.1109/iccvw54120.2021.00367

Cited by 10 publications

(11 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Scanner harmonization is an important step of brain MRI pre-processing to reduce noise and potential biases. Statistical methods such as ComBat (Fortin et al, 2018) harmonize on the image-derived feature level, while some deep learning harmonization methods adjust the image directly (Liu et al, 2021;Modanwal et al, 2020;Torbati et al, 2021b). Deep learningbased approaches are increasingly adopted for various applications in the medical imaging domain, including harmonization.…”

Section: Discussionmentioning

confidence: 99%

“…Perfect harmonization therefore should remove any difference between these imaging pairs. Following previous work (Torbati et al, 2021b(Torbati et al, , 2021a, we compare differences between the paired images on unharmonized, EB harmonized, and FB harmonized datasets for 22 imaging features that have previously been selected as regions of interest for studying AD (Pölsterl and Wachinger, 2020). The mean difference between paired 3T and 1.5T features (bias) and root mean squared deviation (RMSD), a measurement of variance, are computed in the three datasets.…”

Section: Test-retest Using Paired Scan Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

ComBat Harmonization: Empirical Bayes versus Fully Bayes Approaches

Reynolds

Chaudhary

Torbati

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Studying small effects or subtle neuroanatomical variation requires large-scale sample size data. As a result, combining neuroimaging data from multiple datasets is necessary. Variation in acquisition protocols, magnetic field strength, scanner build, and many other non-biologically related factors can introduce undesirable bias into studies. Hence, harmonization is required to remove the bias-inducing factors from the data. ComBat, introduced by (Johnson et al., 2007), is one of the most common methods applied to features from structural images. ComBat models the data using a hierarchical Bayesian model and uses the empirical Bayes approach to infer the distribution of the unknown factors. The empirical Bayes harmonization method is computationally efficient and provides valid point estimates. However, it tends to underestimate uncertainty. This paper investigates a new approach, fully Bayesian ComBat, where Monte Carlo Sampling is used for statistical inference. Our experiments show that our new fully Bayesian approach offers more accurate harmonization, unconstrained posterior distributions, and representative uncertainty quantification at the expense of higher computation costs for the inference. This fully Bayesian approach generates a rich posterior distribution, which is also useful for generating simulated imaging features for improving classifier performance in a limited data setting. We show the generative capacity of our model for augmenting and improving the detection of patients with Alzheimer's disease. Posterior distributions for harmonized imaging measures can also be used for brain-wide uncertainty comparison and more principled downstream statistical analysis. Code for our new fully Bayesian ComBat extension is available at https://github.com/batmanlab/BayesComBat.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Test-retest Using Paired Scan Evaluationmentioning

confidence: 99%

ComBat Harmonization: Empirical Bayes versus Fully Bayes Approaches

Reynolds

Chaudhary

Torbati

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…In order to achieve these goals, we designed a two-step training framework for MISPEL which consists of units of encoder and decoder modules for each of the scanners (Figure 2). More detail on MISPEL were provided in (Torbati et al, 2021b) and the code is publicly available 1 .…”

Section: Methodsmentioning

confidence: 99%

MISPEL: A supervised deep learning harmonization method for multi-scanner neuroimaging data

Torbati

Minhas

Laymon

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Large-scale data obtained from the aggregation of already collected multi-site neuroimaging datasets has brought benefits such as higher statistical power, reliability, and robustness to the studies. Despite these promises from growth in sample size, substantial technical variability stemming from differences in scanner specifications exists in the aggregated data and could inadvertently bias any downstream analyses on it. Such a challenge calls for data normalization and/or harmonization frameworks, in addition to comprehensive criteria to estimate the scanner-related variability and evaluate the harmonization frameworks. In this study, we propose MISPEL (Multi-scanner Image harmonization via Structure Preserving Embedding Learning), a supervised multi-scanner harmonization method. Unlike existing techniques, MISPEL does not assume a perfect coregistration across the scans, and the framework is naturally extendable to more than two scanners. We also designed a set of comprehensive criteria to investigate the scanner-related technical variability and evaluate the harmonization techniques. As an essential requirement of our criteria, we introduced a multi-scanner matched dataset of four 3T MRI T1 images, which, to the best of our knowledge is the first dataset of its kind. We also investigated our evaluations using two popular segmentation frameworks: FSL and segmentation in statistical parametric mapping (SPM). Lastly, we compared MISPEL to popular methods of normalization and harmonization, namely White Stripe, RAVEL, and CALAMITI. MISPEL outperformed these methods and is promising for many other neuroimaging modalities.

show abstract

“…• MRI harmonization via MISPEL (Multi-scanner Image harmonization via Structure Preserving Embedding Learning) when more than two scanners are used (Torbati et al, 2021b) • DeepHarmony addresses MRI contrast differences across two scanners (Dewey et al, 2019) • mica addresses MRI contrast differences across more than two scanners (Wrobel et al, 2020) • PET harmonization…”

Section: Image Processing Skillsmentioning

confidence: 99%

A guide for researchers seeking training in retrospective data harmonization for population neuroscience studies of Alzheimer's disease and related dementias

Shaaban¹,

Tudorascu²,

Glymour³

et al. 2022

Front. Neuroimaging

Self Cite

View full text Add to dashboard Cite

Due to needs surrounding rigor and reproducibility, subgroup specific disease knowledge, and questions of external validity, data harmonization is an essential tool in population neuroscience of Alzheimer's disease and related dementias (ADRD). Systematic harmonization of data elements is necessary to pool information from heterogeneous samples, and such pooling allows more expansive evaluations of health disparities, more precise effect estimates, and more opportunities to discover effective prevention or treatment strategies. The key goal of this Tutorial in Population Neuroimaging Curriculum, Instruction, and Pedagogy article is to guide researchers in creating a customized population neuroscience of ADRD harmonization training plan to fit their needs or those of their mentees. We provide brief guidance for retrospective data harmonization of multiple data types in this area, including: (1) clinical and demographic, (2) neuropsychological, and (3) neuroimaging data. Core competencies and skills are reviewed, and resources are provided to fill gaps in training as well as data needs. We close with an example study in which harmonization is a critical tool. While several aspects of this tutorial focus specifically on ADRD, the concepts and resources are likely to benefit population neuroscientists working in a range of research areas.

show abstract

Multi-scanner Harmonization of Paired Neuroimaging Data via Structure Preserving Embedding Learning

Cited by 10 publications

References 22 publications

ComBat Harmonization: Empirical Bayes versus Fully Bayes Approaches

ComBat Harmonization: Empirical Bayes versus Fully Bayes Approaches

MISPEL: A supervised deep learning harmonization method for multi-scanner neuroimaging data

A guide for researchers seeking training in retrospective data harmonization for population neuroscience studies of Alzheimer's disease and related dementias

Contact Info

Product

Resources

About