The ever-growing digitalization within the healthcare sector in Germany has catalyzed the rapid growth of electronic health records (EHRs), promising significant enhancements in patient care through the development of data-driven insights. Despite the high potential of EHRs to improve healthcare practices, their effective utilization and interpretation remain major challenges, especially in medical settings where prompt and informed decision-making is crucial. This complexity and lack of transparency in data handling often hinder medical practitioners' ability to make quick, informed, and reliable decisions. This master thesis addresses these challenges by conceptualizing and developing a Visual Analytics system tailored for medical use. The proposed system integrates interactive visualizations with advanced pre-trained transformer technologies to aid medical doctors by aggregating and presenting complex patient information on a unified visual dashboard. The core innovation of this thesis lies in the seamless integration of model and visualization frameworks. It carefully models the transformation process from digitized textual medical documents to transformer-based named entity recognition, resulting in interactive visual interfaces. The objective is to create a tool that not only simplifies the interpretation of dense medical data but also enhances the operational efficiency of healthcare providers by facilitating quicker and more accurate decision-making. Through this integration, the thesis aims to forge a pathway for more intuitive, efficient, and effective use of EHRs in everyday medical practice, thereby significantly improving the quality of patient care. This research contributes to a practical solution for digital data management in healthcare by demonstrating the value of combining cutting-edge technologies in real-world applications. It opens the way for future advancements in Visual Analytics.
Visual Analytics for Medical Data. Conceptualisation and implementation of transformer models for named entity recognition of medical reports
MIR, SEYED MOSTAFA
2023/2024
Abstract
The ever-growing digitalization within the healthcare sector in Germany has catalyzed the rapid growth of electronic health records (EHRs), promising significant enhancements in patient care through the development of data-driven insights. Despite the high potential of EHRs to improve healthcare practices, their effective utilization and interpretation remain major challenges, especially in medical settings where prompt and informed decision-making is crucial. This complexity and lack of transparency in data handling often hinder medical practitioners' ability to make quick, informed, and reliable decisions. This master thesis addresses these challenges by conceptualizing and developing a Visual Analytics system tailored for medical use. The proposed system integrates interactive visualizations with advanced pre-trained transformer technologies to aid medical doctors by aggregating and presenting complex patient information on a unified visual dashboard. The core innovation of this thesis lies in the seamless integration of model and visualization frameworks. It carefully models the transformation process from digitized textual medical documents to transformer-based named entity recognition, resulting in interactive visual interfaces. The objective is to create a tool that not only simplifies the interpretation of dense medical data but also enhances the operational efficiency of healthcare providers by facilitating quicker and more accurate decision-making. Through this integration, the thesis aims to forge a pathway for more intuitive, efficient, and effective use of EHRs in everyday medical practice, thereby significantly improving the quality of patient care. This research contributes to a practical solution for digital data management in healthcare by demonstrating the value of combining cutting-edge technologies in real-world applications. It opens the way for future advancements in Visual Analytics.File | Dimensione | Formato | |
---|---|---|---|
Mir_SeyedMostafa.pdf
accesso riservato
Dimensione
1.83 MB
Formato
Adobe PDF
|
1.83 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/78081