Vascular Segmentation of Functional Ultrasound Images using Deep Learning

2024-10-28

Comput. Biol. Medicine

AI Summary

This study addresses the challenge of accurate arterial/venous differentiation and non-invasive dynamic cerebral blood volume (CBV) quantification in functional ultrasound (fUS) imaging. We propose the first deep learning–based fUS vascular semantic segmentation framework, using flow direction (ascending/descending) as a surrogate biomarker for automatic artery–vein classification. To overcome the lack of pixel-level ground truth, we innovatively leverage ultrafast ultrasound localization microscopy (ULM) to generate weakly supervised annotations, integrated with a direction-aware training strategy and temporal modeling of fUS time series. Employing a U-Net architecture, our method achieves 90% classification accuracy, 71% F1-score, and 0.59 IoU using only 100 frames. It generalizes robustly across physiological states (e.g., rest to visual stimulation), and predicted CBV signals exhibit strong linear correlation (r > 0.9) with ground-truth vascular distributions—surpassing the invasiveness constraints and limited dynamic quantification capability of conventional ULM.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Brain-Sensing and Analysis

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Segmentation of medical images is a fundamental task with numerous applications. While MRI, CT, and PET modalities have significantly benefited from deep learning segmentation techniques, more recent modalities, like functional ultrasound (fUS), have seen limited progress. fUS is a non invasive imaging method that measures changes in cerebral blood volume (CBV) with high spatio-temporal resolution. However, distinguishing arterioles from venules in fUS is challenging due to opposing blood flow directions within the same pixel. Ultrasound localization microscopy (ULM) can enhance resolution by tracking microbubble contrast agents but is invasive, and lacks dynamic CBV quantification. In this paper, we introduce the first deep learning-based application for fUS image segmentation, capable of differentiating signals based on vertical flow direction (upward vs. downward), using ULM-based automatic annotation, and enabling dynamic CBV quantification. In the cortical vasculature, this distinction in flow direction provides a proxy for differentiating arteries from veins. We evaluate various UNet architectures on fUS images of rat brains, achieving competitive segmentation performance, with 90% accuracy, a 71% F1 score, and an IoU of 0.59, using only 100 temporal frames from a fUS stack. These results are comparable to those from tubular structure segmentation in other imaging modalities. Additionally, models trained on resting-state data generalize well to images captured during visual stimulation, highlighting robustness. Although it does not reach the full granularity of ULM, the proposed method provides a practical, non-invasive and cost-effective solution for inferring flow direction-particularly valuable in scenarios where ULM is not available or feasible. Our pipeline shows high linear correlation coefficients between signals from predicted and actual compartments, showcasing its ability to accurately capture blood flow dynamics.

Problem

• Segmentation of functional ultrasound images using deep learning
• Differentiating arterioles from venules in fUS imaging
• Enabling dynamic cerebral blood volume quantification non-invasively

Innovation

• Deep learning segments fUS images non-invasively
• ULM annotation enables dynamic CBV quantification
• UNet achieves 90% accuracy with minimal frames