Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Cross-domain Fiber Cluster Shape Analysis for Language Performance Cognitive Score Prediction

Yui Lo, Yuqian Chen, Dongnan Liu, Wan Liu, Leo Zekelman, Fan Zhang, Yogesh Rathi, Nikos Makris, Alexandra J. Golby, Weidong Cai, Lauren J. O'Donnell

TL;DR

This work addresses whether the shape of white matter fiber clusters, derived from diffusion MRI tractography, can predict individual language performance. It introduces SFFormer, an encoder-only transformer with a multi-head cross-attention fusion module that integrates shape, microstructure, and connectivity features across 953 fiber clusters per subject. Evaluated on 1065 healthy young adults from the HCP-YA dataset, SFFormer with domain fusion outperforms a CNN and a baseline transformer in predicting TPVT and TORRT language scores, highlighting the contribution of fiber-cluster shape information. The findings imply that the geometry of white matter connections provides meaningful signals for language function and that cross-domain fusion of shape with microstructural and connectivity features enhances cognitive performance prediction.

Abstract

Shape plays an important role in computer graphics, offering informative features to convey an object's morphology and functionality. Shape analysis in brain imaging can help interpret structural and functionality correlations of the human brain. In this work, we investigate the shape of the brain's 3D white matter connections and its potential predictive relationship to human cognitive function. We reconstruct brain connections as sequences of 3D points using diffusion magnetic resonance imaging (dMRI) tractography. To describe each connection, we extract 12 shape descriptors in addition to traditional dMRI connectivity and tissue microstructure features. We introduce a novel framework, Shape--fused Fiber Cluster Transformer (SFFormer), that leverages a multi-head cross-attention feature fusion module to predict subject-specific language performance based on dMRI tractography. We assess the performance of the method on a large dataset including 1065 healthy young adults. The results demonstrate that both the transformer-based SFFormer model and its inter/intra feature fusion with shape, microstructure, and connectivity are informative, and together, they improve the prediction of subject-specific language performance scores. Overall, our results indicate that the shape of the brain's connections is predictive of human language function.

Cross-domain Fiber Cluster Shape Analysis for Language Performance Cognitive Score Prediction

TL;DR

This work addresses whether the shape of white matter fiber clusters, derived from diffusion MRI tractography, can predict individual language performance. It introduces SFFormer, an encoder-only transformer with a multi-head cross-attention fusion module that integrates shape, microstructure, and connectivity features across 953 fiber clusters per subject. Evaluated on 1065 healthy young adults from the HCP-YA dataset, SFFormer with domain fusion outperforms a CNN and a baseline transformer in predicting TPVT and TORRT language scores, highlighting the contribution of fiber-cluster shape information. The findings imply that the geometry of white matter connections provides meaningful signals for language function and that cross-domain fusion of shape with microstructural and connectivity features enhances cognitive performance prediction.

Abstract

Shape plays an important role in computer graphics, offering informative features to convey an object's morphology and functionality. Shape analysis in brain imaging can help interpret structural and functionality correlations of the human brain. In this work, we investigate the shape of the brain's 3D white matter connections and its potential predictive relationship to human cognitive function. We reconstruct brain connections as sequences of 3D points using diffusion magnetic resonance imaging (dMRI) tractography. To describe each connection, we extract 12 shape descriptors in addition to traditional dMRI connectivity and tissue microstructure features. We introduce a novel framework, Shape--fused Fiber Cluster Transformer (SFFormer), that leverages a multi-head cross-attention feature fusion module to predict subject-specific language performance based on dMRI tractography. We assess the performance of the method on a large dataset including 1065 healthy young adults. The results demonstrate that both the transformer-based SFFormer model and its inter/intra feature fusion with shape, microstructure, and connectivity are informative, and together, they improve the prediction of subject-specific language performance scores. Overall, our results indicate that the shape of the brain's connections is predictive of human language function.
Paper Structure (14 sections, 2 figures, 2 tables)

This paper contains 14 sections, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. Tractography and Fiber Clustering
  5. Traditional and Shape Features
  6. Shape-Fused Fiber Cluster Transformer (SFFormer)
  7. Implementation Details
  8. EXPERIMENTS & RESULTS
  9. Language Assessments Scores
  10. Evaluation Metric
  11. Results and Discussions
  12. Conclusion
  13. COMPLIANCE WITH ETHICAL STANDARDS
  14. ACKNOWLEDGEMENTS

Figures (2)

  • Figure 1: Four example individual white matter connections (fiber clusters) extracted from the entire white matter of the human brain using a fiber clustering approach Zhang2018-iw. Example shape descriptors are extracted for the blue fiber cluster.
  • Figure 2: Overview of the SFFormer framework. HCP dMRI data undergoes whole brain tractography to obtain 953 fiber clusters. The microstructure, connectivity and shape features of the fiber clusters are calculated and used as inputs to the SFFormer framework that leverages both a helper feature and a primary feature in the multi-head cross-attention module to output a language prediction score.