Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Tailoring Graph Neural Network-based Flow-guided Localization to Individual Bloodstreams and Activities

Pablo Galván, Filip Lemic, Gerard Calvo Bartra, Sergi Abadal, Xavier Costa Pérez

TL;DR

This work tackles variability in patient bloodstreams and activities that challenge flow-guided nanoscale localization. It extends a state-of-the-art GNN approach by incorporating physiological indicators—height, weight, and heart rate—into extended graph designs (including a master node and probability-edge schemes) and by transforming raw data to reflect different profiles. Hyperparameter tuning shows that the combined extended design substantially improves region-level localization accuracy across most profiles, enabling more reliable, individualized monitoring beyond a resting bloodstream. The findings highlight the importance of tailoring localization systems to individual physiology to enable continuous, real-time medical monitoring with in-body nanodevices.

Abstract

Flow-guided localization using in-body nanodevices in the bloodstream is expected to be beneficial for early disease detection, continuous monitoring of biological conditions, and targeted treatment. The nanodevices face size and power constraints that produce erroneous raw data for localization purposes. On-body anchors receive this data, and use it to derive the locations of diagnostic events of interest. Different Machine Learning (ML) approaches have been recently proposed for this task, yet they are currently restricted to a reference bloodstream of a resting patient. As such, they are unable to deal with the physical diversity of patients' bloodstreams and cannot provide continuous monitoring due to changes in individual patient's activities. Toward addressing these issues for the current State-of-the-Art (SotA) flow-guided localization approach based on Graph Neural Networks (GNNs), we propose a pipeline for GNN adaptation based on individual physiological indicators including height, weight, and heart rate. Our results indicate that the proposed adaptions are beneficial in reconciling the individual differences between bloodstreams and activities.

Tailoring Graph Neural Network-based Flow-guided Localization to Individual Bloodstreams and Activities

TL;DR

This work tackles variability in patient bloodstreams and activities that challenge flow-guided nanoscale localization. It extends a state-of-the-art GNN approach by incorporating physiological indicators—height, weight, and heart rate—into extended graph designs (including a master node and probability-edge schemes) and by transforming raw data to reflect different profiles. Hyperparameter tuning shows that the combined extended design substantially improves region-level localization accuracy across most profiles, enabling more reliable, individualized monitoring beyond a resting bloodstream. The findings highlight the importance of tailoring localization systems to individual physiology to enable continuous, real-time medical monitoring with in-body nanodevices.

Abstract

Flow-guided localization using in-body nanodevices in the bloodstream is expected to be beneficial for early disease detection, continuous monitoring of biological conditions, and targeted treatment. The nanodevices face size and power constraints that produce erroneous raw data for localization purposes. On-body anchors receive this data, and use it to derive the locations of diagnostic events of interest. Different Machine Learning (ML) approaches have been recently proposed for this task, yet they are currently restricted to a reference bloodstream of a resting patient. As such, they are unable to deal with the physical diversity of patients' bloodstreams and cannot provide continuous monitoring due to changes in individual patient's activities. Toward addressing these issues for the current State-of-the-Art (SotA) flow-guided localization approach based on Graph Neural Networks (GNNs), we propose a pipeline for GNN adaptation based on individual physiological indicators including height, weight, and heart rate. Our results indicate that the proposed adaptions are beneficial in reconciling the individual differences between bloodstreams and activities.
Paper Structure (13 sections, 9 figures, 2 tables)

This paper contains 13 sections, 9 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Considered GNN Model
  4. GNN-enabled Flow-guided Localization
  5. GNN Architecture
  6. GNN-based Localization Tailored to Individual Profiles
  7. Extended Graph Designs
  8. Raw Data Transformation
  9. Evaluation Methodology and Results
  10. Evaluation Methodology
  11. Hyperparameter Tuning
  12. Evaluation Results
  13. Conclusion

Figures (9)

  • Figure 1: High-level overview of flow-guided nanoscale localization tailored to individual profiles
  • Figure 2: Overview of the utilized GNN-based flow-guided nanoscale localization approach bartra2023graph
  • Figure 3: Overview of the reference GNN architecture bartra2023graph
  • Figure 4: Overview of the extended graph designs, showing the master node, its potential features, and the probability edges between the heart nodes and the other regions
  • Figure 5: Cardiovascular system for different patient profiles and their settings, with the size differences emphasized for visualization purposes
  • ...and 4 more figures