Joint and individual analysis of breast cancer histologic images and genomic covariates

Carmichael, Iain; Calhoun, Benjamin C.; Hoadley, Katherine A.; Troester, Melissa A.; Geradts, Joseph; Couture, Heather D.; Olsson, Linnea T.; Perou, Charles M.; Niethammer, Marc; Hannig, Jan; Marron, J. S.

doi:10.1214/20-aoas1433

Cited by 7 publications

(3 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We project the patches onto a given projection vector and visualize the patches along the spectrum. A similar interpretability approach was taken in Carmichael et al (2021). Following (3), for a projection vector v and the the jth instance, we define the signed attention weighted squared residuals as…”

Section: Interpretability Via Attention Scoresmentioning

confidence: 99%

Incorporating intratumoral heterogeneity into weakly-supervised deep learning models via variance pooling

Carmichael¹,

Song²,

Chen³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Supervised learning tasks such as cancer survival prediction from gigapixel whole slide images (WSIs) are a critical challenge in computational pathology that requires modeling complex features of the tumor microenvironment. These learning tasks are often solved with deep multi-instance learning (MIL) models that do not explicitly capture intratumoral heterogeneity. We develop a novel variance pooling architecture that enables a MIL model to incorporate intratumoral heterogeneity into its predictions. Two interpretability tools based on "representative patches" are illustrated to probe the biological signals captured by these models. An empirical study with 4,479 gigapixel WSIs from the Cancer Genome Atlas shows that adding variance pooling onto MIL frameworks improves survival prediction performance for five cancer types.

show abstract

Section: Interpretability Via Attention Scoresmentioning

confidence: 99%

Incorporating intratumoral heterogeneity into weakly-supervised deep learning models via variance pooling

Carmichael¹,

Song²,

Chen³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…If at this stage we were interested in understanding co-variation between brain shape and connectivity, we could perform CCA on the scores matrices A S and A C , or alternatively, an AJIVE analysis on the tangent vectors (Feng et al, 2018;Carmichael et al, 2019). Nevertheless, the estimated joint variation components would be an average of the co-variation that is due to genetic and environmental factors.…”

Section: A :"mentioning

confidence: 99%

“…The proposed model can be regarded as an extension of the classical single-trait and bivariate-trait variance component models in Amos (1994); Almasy et al (1997) and is formulated as a multivariate mixed effects model on the Karhunen-Loève basis coefficients of the tangent space coordinates. If instead, the aim of the analysis was to estimate joint modes of variation in shape and connectivity, regardless of their sources, a Canonical Correlation Analysis (CCA) (Hotelling, 1936;Johnson and Wichern, 2002) or an Angle-based Joint and Individual Variation Explained (AJIVE) analysis could have been performed on the tangent space coordinates (Lock et al, 2013;Feng et al, 2018;Carmichael et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Functional random effects modeling of brain shape and connectivity

Lila¹,

Aston²

2020

Preprint

View full text Add to dashboard Cite

We present a statistical framework that jointly models brain shape and functional connectivity, which are two complex aspects of the brain that have been classically studied independently. We adopt a Riemannian modeling approach to account for the non-Euclidean geometry of the space of shapes and the space of connectivity that constrains trajectories of co-variation to be valid statistical estimates. In order to disentangle genetic sources of variability from those driven by unique environmental factors, we embed a functional random effects model in the Riemannian framework. We apply the proposed model to the Human Connectome Project dataset to explore spontaneous co-variation between brain shape and connectivity in young healthy individuals.

show abstract

Incorporating Intratumoral Heterogeneity into Weakly-Supervised Deep Learning Models via Variance Pooling

Carmichael

Song

Chen

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Joint and individual analysis of breast cancer histologic images and genomic covariates

Cited by 7 publications

References 58 publications

Incorporating intratumoral heterogeneity into weakly-supervised deep learning models via variance pooling

Incorporating intratumoral heterogeneity into weakly-supervised deep learning models via variance pooling

Functional random effects modeling of brain shape and connectivity

Incorporating Intratumoral Heterogeneity into Weakly-Supervised Deep Learning Models via Variance Pooling

Contact Info

Product

Resources

About