Leveraging Large Language Models for Efficient Labeling of Ballistocardiogram (BCG) Data

• Typ der Arbeit: Masterarbeit
• Status der Arbeit: offen
• Projekte: AuRelia
• Betreuer: Kazi Mohammad Abidur Rahman

Motivation

Ballistocardiography (BCG) is a promising non-invasive technique for monitoring cardiovascular health, but its widespread adoption in machine learning applications is hindered by the lack of high-quality labeled datasets. Manual annotation remains a major bottleneck due to its reliance on clinical expertise, susceptibility to variability, and lack of scalability. This thesis should explore the integration of Large Language Models (LLMs) into the BCG data annotation pipeline. By combining time-series signal processing techniques with context-aware prompt engineering, the goal is to enable LLMs to assist in semi-automated labeling of BCG events and morphological features. Students participating in this thesis will gain experience in biomedical signal processing, machine learning, and LLMs, contributing to the development of a scalable framework that improves annotation efficiency and consistency for BCG datasets.

Who can apply

We welcome applications from highly motivated students with a strong background in Computer Science & Engineering, Data Science, or Signal Processing, who are capable of working independently and have a keen interest in exploring the applications of Large Language Models (LLMs).

Tasks

TBD

Further Reading

TBD