Confounder-Aware Visualization of ConvNets

Zhao, Qingyu; Adeli, Ehsan; Pfefferbaum, Adolf; Sullivan, Edith V.; Pohl, Kilian M.

doi:10.1007/978-3-030-32692-0_38

Cited by 12 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Next, we visualized all functional connections (off-diagonal entries) with importance value higher than 0.3. For sex prediction, the most important ROI identified by ST-GCN was the inferior temporal lobe, which echoed findings from other resting-state studies [4] and from a structural MRI analysis on the NCANDA cohort [27]. We also identified a significant effect in the frontal-posterior-cingulate (PCC) connection (red in Fig.…”

Section: Results and Analysissupporting

confidence: 84%

See 1 more Smart Citation

Spatio-Temporal Graph Convolution for Resting-State fMRI Analysis

Gadgil¹,

Zhao²,

Pfefferbaum³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

The BOLD signal of resting-state fMRI (rs-fMRI) records the functional brain connectivity in a rich dynamic spatio-temporal setting. However, existing methods applied to rs-fMRI often fail to consider both spatial and temporal characteristics of the data. They either neglect the functional dependency between different brain regions in a network or discard the information in the temporal dynamics of brain activity. To overcome those shortcomings, we propose to formulate functional connectivity networks within the context of spatio-temporal graphs. We then train a spatio-temporal graph convolutional network (ST-GCN) on short sub-sequences of the BOLD time series to model the non-stationary nature of functional connectivity. We simultaneously learn the graph edge importance within ST-GCN to enable interpretation of functional connectivities contributing to the prediction model. In analyzing the rs-fMRI of the Human Connectome Project (HCP, N =1,091) and the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA, N =773), ST-GCN is significantly more accurate than common approaches in predicting gender and age based on BOLD signals. The matrix recording edge importance localizes brain regions and functional connections with significant aging and sex effects, which are verified by the neuroscience literature.

show abstract

Section: Results and Analysissupporting

confidence: 84%

“…HCP. The Human Connectome Project (HCP) S1200 [20] contains the rs-fMRI data for 1096 young adults (ages [22][23][24][25][26][27][28][29][30][31][32][33][34][35]. We used the first session (15 min, T = 1200 frames, TR=0.72s) for each subject and excluded 5 rs-fMRIs with less than 1200 frames, resulting in the data of 498 females and 593 males.…”

Section: Methodsmentioning

confidence: 99%

Spatio-Temporal Graph Convolution for Resting-State fMRI Analysis

Gadgil¹,

Zhao²,

Pfefferbaum³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…With respect to the network architecture used in the experiments, we followed the design of FE in refs. 69,64 that contained 4 stacks of 2 × 2 × 2 3D convolution/ ReLu/batch-normalization/max-pooling layers, yielding 4096 intermediate features.…”

mentioning

confidence: 99%

Training confounder-free deep learning models for medical applications

2020

Self Cite

View full text Add to dashboard Cite

The presence of confounding effects (or biases) is one of the most critical challenges in using deep learning to advance discovery in medical imaging studies. Confounders affect the relationship between input data (e.g., brain MRIs) and output variables (e.g., diagnosis). Improper modeling of those relationships often results in spurious and biased associations. Traditional machine learning and statistical models minimize the impact of confounders by, for example, matching data sets, stratifying data, or residualizing imaging measurements. Alternative strategies are needed for state-of-the-art deep learning models that use end-to-end training to automatically extract informative features from large set of images. In this article, we introduce an end-to-end approach for deriving features invariant to confounding factors while accounting for intrinsic correlations between the confounder(s) and prediction outcome. The method does so by exploiting concepts from traditional statistical methods and recent fair machine learning schemes. We evaluate the method on predicting the diagnosis of HIV solely from Magnetic Resonance Images (MRIs), identifying morphological sex differences in adolescence from those of the National Consortium on Alcohol and Neurodevelopment in Adolescence (NCANDA), and determining the bone age from X-ray images of children. The results show that our method can accurately predict while reducing biases associated with confounders. The code is available at https://github.com/qingyuzhao/br-net.

show abstract

“…We evaluate the proposed model on two longitudinal neuroimaging datasets: the Alzheimer's Disease Neuroimaging Initiative (ADNI1) and a data set on Alcohol Use Dependence (AUD). All longitudinal MRIs in the following experiments were first preprocessed by a pipeline composed of denoising, bias field correction, skull striping, affine registration to a template, re-scaling to a 64 × 64 × 64 volume, transforming image intensities within the brainmask to z-scores [46]. We design an encoder composed of 4 stacks of 3 × 3 × 3 convolution/ReLU/max-pooling layers with dimension (16,32,64,16).…”

Section: Methodsmentioning

confidence: 99%

Longitudinal Self-Supervised Learning

Zhao¹,

Liu²,

Adeli³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Longitudinal neuroimaging or biomedical studies often acquire multiple observations from each individual over time, which entails repeated measures with highly interdependent variables. In this paper, we discuss the implication of repeated measures design on unsupervised learning by showing its tight conceptual connection to self-supervised learning and factor disentanglement. Leveraging the ability for 'self-comparison' through repeated measures, we explicitly separate the definition of the factor space and the representation space enabling an exact disentanglement of time-related factors from the representations of the images. By formulating deterministic multivariate mapping functions between the two spaces, our model, named Longitudinal Self-Supervised Learning (LSSL), uses a standard autoencoding structure with a cosine loss to estimate the direction linked to the disentangled factor. We apply LSSL to two longitudinal neuroimaging studies to show its unique advantage in extracting the 'brain-age' information from the data and in revealing informative characteristics associated with neurodegenerative and neuropsychological disorders. For a downstream task of supervised diagnosis classification, the representations learned by LSSL permit faster convergence and higher (or similar) prediction accuracy compared to several other representation learning techniques.

show abstract

Confounder-Aware Visualization of ConvNets

Cited by 12 publications

References 12 publications

Spatio-Temporal Graph Convolution for Resting-State fMRI Analysis

Spatio-Temporal Graph Convolution for Resting-State fMRI Analysis

Training confounder-free deep learning models for medical applications

Longitudinal Self-Supervised Learning

Contact Info

Product

Resources

About