Martin Baumgartner

Mr. Martin Baumgartner

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 August 2023 in Journal of Healthcare Informatics Research
Verified
Journal Article
Open access
PDF file
MBMartin Baumgartner
Martin Baumgartner
SVSai Pavan Kumar Veeranki
Sai Pavan Kumar Veeranki
DHDieter Hayn
Dieter Hayn
GSGuenter Schreier
Guenter Schreier
Artificial intelligence and machine learning have led to prominent and spectacular innovations in various scenarios. Application in medicine, however, can be challenging due to privacy concerns and strict legal regulations. Methods that centralize knowledge instead of data could address this issue. In this work, 6 different decentralized machine learning algorithms are applied to 12-lead ECG classification and compared to conventional, centralized machine learning. The results show that state-of-the-art federated learning leads to reasonable losses of classification performance compared to a standard, central model (−0.054 AUROC) while providing a significantly higher level of privacy. A proposed weighted variant of federated learning (−0.049 AUROC) and an ensemble (−0.035 AUROC) outperformed the standard federated learning algorithm. Overall, considering multiple metrics, the novel batch-wise sequential learning scheme performed best (−0.036 AUROC to baseline). Although, the technical aspects of implementing them in a real-world application are to be carefully considered, the described algorithms constitute a way forward towards preserving-preserving AI in medicine.

10 April 2024 in Frontiers in Medicine
Verified
Journal Article
Open access
MBMartin Baumgartner
Martin Baumgartner
KKKarl Kreiner
Karl Kreiner
ALAaron Lauschensky
Aaron Lauschensky
BJBernhard Jammerbund
Bernhard Jammerbund
KDKlaus Donsa
Klaus Donsa
GSGünter Schreier
Günter Schreier
Introduction: The potential for secondary use of health data to improve healthcare is currently not fully exploited. Health data is largely kept in isolated data silos and key infrastructure to aggregate these silos into standardized bodies of knowledge is underdeveloped. We describe the development, implementation, and evaluation of a federated infrastructure to facilitate versatile secondary use of health data based on Health Data Space nodes.Materials and methods: Our proposed nodes are self-contained units that digest data through an extract-transform-load framework that pseudonymizes and links data with privacy-preserving record linkage and harmonizes into a common data model (OMOP CDM). To support collaborative analyses a multi-level feature store is also implemented. A feasibility experiment was conducted to test the infrastructures potential for machine learning operations and deployment of other apps (e.g., visualization). Nodes can be operated in a network at different levels of sharing according to the level of trust within the network.Results: In a proof-of-concept study, a privacy-preserving registry for heart failure patients has been implemented as a real-world showcase for Health Data Space nodes at the highest trust level, linking multiple data sources including (a) electronical medical records from hospitals, (b) patient data from a telemonitoring system, and (c) data from Austria’s national register of deaths. The registry is deployed at the tirol kliniken, a hospital carrier in the Austrian state of Tyrol, and currently includes 5,004 patients, with over 2.9 million measurements, over 574,000 observations, more than 63,000 clinical free text notes, and in total over 5.2 million data points. Data curation and harmonization processes are executed semi-automatically at each individual node according to data sharing policies to ensure data sovereignty, scalability, and privacy. As a feasibility test, a natural language processing model for classification of clinical notes was deployed and tested.Discussion: The presented Health Data Space node infrastructure has proven to be practicable in a real-world implementation in a live and productive registry for heart failure. The present work was inspired by the European Health Data Space initiative and its spirit to interconnect health data silos for versatile secondary use of health data.

8 October 2024 in IEEE Transactions on Biomedical Engineering
Verified
Journal Article
Closed access
AAAhmadreza Argha
Ahmadreza Argha
HRHamid Rokny
Hamid Rokny
MBMartin Baumgartner
Martin Baumgartner
GSGuenter Schreier
Guenter Schreier
BCBranko G. Celler
Branko G. Celler
NLNigel Lovell
Nigel Lovell
Objective: Telehealth paradigms are essential for remotely managing patients with chronic conditions. To assist clinicians in handling the large volumes of data collected through these systems, clinical decision support systems (CDSSs) have been developed. However, the effectiveness of CDSSs depends on the quality of remotely recorded physiological data and the reliability of the algorithms used for processing this data. This study aims to reliably detect atrial fibrillation (AF) from short-term single-lead (STSL) electrocardiogram (ECG) recordings obtained in unsupervised telehealth environments. Methods: A novel deep ensemble-based method was developed for detecting AF from STSL ECG recordings. Following this, a postprocessing algorithm was created to assess uncertainty in classified STSL ECGs and to refrain from interpretation when confidence is low. The proposed method was validated through a 5-fold cross-validation on the Cardiology Challenge 2017 (CinC2017) dataset. Results: The deep ensemble method achieved 83.5 $\pm$ 1.5% sensitivity, 98.4 $\pm$ 0.2% specificity, and an F$_{1}$-score of 0.847 $\pm$ 0.016in AF detection. Implementing the selective classification algorithm resulted in significant improvements, with sensitivity increasing to 92.8 $\pm$ 2.2%, specificity to 99.7 $\pm$ 0.0%, and an F$_{1}$-score of 0.919 $\pm$ 0.016. Conclusion: The proposed method demonstrates the feasibility of accurately detecting AF from STSL ECG recordings. The selective classification approach offers a substantial enhancement to automated ECG interpretation algorithms in telehealth solutions. Significance: These findings highlight the potential for improving the utility of telehealth systems by integrating advanced CDSSs capable of managing uncertainty and ensuring higher accuracy, thereby improving patient outcomes in remote healthcare settings.

6 August 2024 in Future Internet
Verified
Journal Article
Open access
PDF file
MBMartin Baumgartner
Martin Baumgartner
KKKarl Kreiner
Karl Kreiner
FWFabian Wiesmüller
Fabian Wiesmüller
DHDieter Hayn
Dieter Hayn
CPChristian Puelacher
Christian Puelacher
GSGuenter Schreier
Guenter Schreier
Background: The recent rise of large language models has triggered renewed interest in medical free text data, which holds critical information about patients and diseases. However, medical free text is also highly sensitive. Therefore, de-identification is typically required but is complicated since medical free text is mostly unstructured. With the Masketeer algorithm, we present an effective tool to de-identify German medical text. Methods: We used an ensemble of different masking classes to remove references to identifiable data from over 35,000 clinical notes in accordance with the HIPAA Safe Harbor Guidelines. To retain additional context for readers, we implemented an entity recognition scheme and corpus-wide pseudonymization. Results: The algorithm performed with a sensitivity of 0.943 and specificity of 0.933. Further performance analyses showed linear runtime complexity (O(n)) with both increasing text length and corpus size. Conclusions: In the future, large language models will likely be able to de-identify medical free text more effectively and thoroughly than handcrafted rules. However, such gold-standard de-identification tools based on large language models are yet to emerge. In the current absence of such, we hope to provide best practices for a robust rule-based algorithm designed with expert domain knowledge.

29 July 2024 in Cancers
Verified
Journal Article
Open access
PDF file
DHDieter Hayn
Dieter Hayn
KKKarl Kreiner
Karl Kreiner
ESEmanuel Sandner
Emanuel Sandner
MBMartin Baumgartner
Martin Baumgartner
BJBernhard Jammerbund
Bernhard Jammerbund
GSGuenter Schreier
Guenter Schreier
Large datasets in paediatric oncology are inherently rare. Therefore, it is paramount to fully exploit all available data, which are distributed over several resources, including biomaterials, images, clinical trials, and registries. With privacy-preserving record linkage (PPRL), personalised or pseudonymised datasets can be merged, without disclosing the patients’ identities. Although PPRL is implemented in various settings, use case descriptions are currently fragmented and incomplete. The present paper provides a comprehensive overview of current and future use cases for PPRL in paediatric oncology. We analysed the literature, projects, and trial protocols, identified use cases along a hypothetical patient journey, and discussed use cases with paediatric oncology experts. To structure PPRL use cases, we defined six key dimensions: distributed personalised records, pseudonymisation, distributed pseudonymised records, record linkage, linked data, and data analysis. Selected use cases were described (a) per dimension and (b) on a multi-dimensional level. While focusing on paediatric oncology, most aspects are also applicable to other (particularly rare) diseases. We conclude that PPRL is a key concept in paediatric oncology. Therefore, PPRL strategies should already be considered when starting research projects, to avoid distributed data silos, to maximise the knowledge derived from collected data, and, ultimately, to improve outcomes for children with cancer.

10 April 2025 in JMIR Formative Research
Verified
Journal Article
Open access
FWFabian Wiesmüller
Fabian Wiesmüller
APAndreas Prenner
Andreas Prenner
AZAndreas Ziegl
Andreas Ziegl
GEGihan El-Moazen
Gihan El-Moazen
RMRobert Modre-Osprian
Robert Modre-Osprian
DHDieter Hayn
Dieter Hayn
Background: Telehealth has been effective in managing cardiovascular diseases like stroke and heart failure and has shown promising results in managing patients with peripheral arterial disease. However, more work is needed to fully understand the effect of telehealth-based predictive modeling on the physical fitness of patients with peripheral arterial disease. Objective: For this work, data from the Keep Pace study were analyzed in depth to gain insights on temporal developments of patients’ conditions and to develop models to predict the patients’ total walking distance at the study end. This could help to determine patients who are likely to benefit from the telehealth program and to continuously provide estimations to the patients as a motivating factor. Methods: This work analyzes continuous patient-reported telehealth data, in combination with in-clinic data from 19 Fontaine stage II patients with peripheral arterial disease who underwent a 12-week telehealth-based walking program. This analysis granted insights into the increase of the total walking distance of the 6-minute walk tests (6MWT) as a measure for physical fitness, the steady decrease in the patients’ pain, and the positive correlation between well-being and the total walking distance measured by the 6MWT. Results: This work analyzed trends of and correlations between continuous patient-generated data. Findings of this study include a significant decrease of the patients’ pain sensation over time (P=.006), a low but highly significant correlation between pain sensation and steps taken on the same day (r=−0.11; P<.001) and the walking distance of the independently performed 6MWTs (r=−0.39; P<.001). Despite the reported pain, adherence to the 6MWT measurement protocol was high (85.53%). Additionally, patients significantly improved their timed-up-and-go test times during the study (P=.002). Predicting the total walking distance at the study end measured by the 6MWT worked well at study baseline (root mean squared error of 30 meters; 7.04% of the mean total walking distance at the study end of 425 meters) and continuously improved by adding further telehealth data. Future work should validate these findings in a larger cohort and in a prospective setting based on a clinical outcome. Conclusions: We conclude that the prototypical trend estimation has great potential for an integration in the telehealth system to be used in future work to provide tailored patient-specific advice based on these predictions. Continuous data from the telehealth system grant a deeper insight and a better understanding of the patients’ status concerning well-being and level of pain as well as their current physical fitness level and the progress toward reaching set goals.