Information
- Publication Type: Master Thesis
- Workgroup(s)/Project(s): not specified
- Date: 2025
- TU Wien Library: AC17679434
- Open Access: yes
- First Supervisor: Renata Georgia Raidou
- Pages: 87
- Keywords: Visual Analysis Pipeline, NCG-TMS, Data-Users-Tasks Design Triangle, Data-Users-Workflow-Tasks Design Pyramid
Abstract
We designed a highly flexible notebook-based visual analysis pipeline to explore Transcranial Magnetic Stimulation (TMS) and heart rate (HR) in different subjects’ cognitive states. TMS is a promising treatment of major depressive disorder not responsive to pharmacological treatment. However, the mechanism of action is not yet fully understood. The research in the best acquiring settings, such as stimulation intensities and target sites, is emerging. Multimodal analysis pipeline integrating TMS, Functional Magnetic Resonance Imaging (fMRI), and HR could shed light on both understanding the neural pathways and increasing the efficiency of TMS. To extend the already available concurrent TMS-fMRI analysis pipeline towards multimodal concurrent TMS-fMRI-HR, exploring the effect of TMS on HR is the next step. To design the visual analysis pipeline, we introduce and apply an extension of Data–Users–Tasks design triangle [Miksch and Aigner, 2014] by integrating the previous data workflow approach in the designing process. When the data processing workflow in the domain is only evolving, integrating the previous workflow approach into the design process benefits by respecting the data legacy, supports users’ adaptability, and ensures tasks’ compatibility. We refer to this framework as the Data–Users–Workflow-Tasks design pyramid. We subsequently provide a visual analysis pipeline to support data exploration in the early stages of research. The interactive preprocessing pipeline involves extracting data, handling missing data, and reducing noise. To compare time series of HR with different properties, we visualize Dynamic Time Warping (DTW) similarity measurement, and heart rate variability (HRV) metric clustering. We quantitatively evaluate the preprocessing steps using simulated ECG data. The key result is that linear and polynomial interpolation with root mean squared error (RMSE) values as low as to the power of -3 and -5, respectively, are especially effective as imputation methods for ECG with 400 Hz sampling frequency. To further assess the values of the usage scenarios for TMS and ECG data exploration, we employ Qualitative Result Inspection (QRI). Our proposed visual analysis pipeline assembles the first steps towards integrating TMS-HR analysis into a trimodal concurrent TMS-fMRI-HR approach.
Additional Files and Images
Weblinks
BibTeX
@mastersthesis{grexova-2025-dva,
title = "Designing a visual analysis pipeline for exploring TMS
effects on heart rate",
author = "R\'{a}chel Grexov\'{a}",
year = "2025",
abstract = "We designed a highly flexible notebook-based visual analysis
pipeline to explore Transcranial Magnetic Stimulation (TMS)
and heart rate (HR) in different subjects’ cognitive
states. TMS is a promising treatment of major depressive
disorder not responsive to pharmacological treatment.
However, the mechanism of action is not yet fully
understood. The research in the best acquiring settings,
such as stimulation intensities and target sites, is
emerging. Multimodal analysis pipeline integrating TMS,
Functional Magnetic Resonance Imaging (fMRI), and HR could
shed light on both understanding the neural pathways and
increasing the efficiency of TMS. To extend the already
available concurrent TMS-fMRI analysis pipeline towards
multimodal concurrent TMS-fMRI-HR, exploring the effect of
TMS on HR is the next step. To design the visual analysis
pipeline, we introduce and apply an extension of
Data–Users–Tasks design triangle [Miksch and Aigner,
2014] by integrating the previous data workflow approach in
the designing process. When the data processing workflow in
the domain is only evolving, integrating the previous
workflow approach into the design process benefits by
respecting the data legacy, supports users’ adaptability,
and ensures tasks’ compatibility. We refer to this
framework as the Data–Users–Workflow-Tasks design
pyramid. We subsequently provide a visual analysis pipeline
to support data exploration in the early stages of research.
The interactive preprocessing pipeline involves extracting
data, handling missing data, and reducing noise. To compare
time series of HR with different properties, we visualize
Dynamic Time Warping (DTW) similarity measurement, and heart
rate variability (HRV) metric clustering. We quantitatively
evaluate the preprocessing steps using simulated ECG data.
The key result is that linear and polynomial interpolation
with root mean squared error (RMSE) values as low as to the
power of -3 and -5, respectively, are especially effective
as imputation methods for ECG with 400 Hz sampling
frequency. To further assess the values of the usage
scenarios for TMS and ECG data exploration, we employ
Qualitative Result Inspection (QRI). Our proposed visual
analysis pipeline assembles the first steps towards
integrating TMS-HR analysis into a trimodal concurrent
TMS-fMRI-HR approach.",
pages = "87",
address = "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
school = "Research Unit of Computer Graphics, Institute of Visual
Computing and Human-Centered Technology, Faculty of
Informatics, TU Wien",
keywords = "Visual Analysis Pipeline, NCG-TMS, Data-Users-Tasks Design
Triangle, Data-Users-Workflow-Tasks Design Pyramid",
URL = "https://www.cg.tuwien.ac.at/research/publications/2025/grexova-2025-dva/",
}
thesis