@phdthesis{Muensterer2022,
  author    = {M{\"u}nsterer, Sascha Ottmar},
  title     = {Prognostische Wertigkeit der Herzfrequenz in Abh{\"a}ngigkeit von implantierten Devices bei akuter Herzinsuffizienz: Ergebnisse des prospektiven AHF-Registers W{\"u}rzburg},
  doi       = {10.25972/OPUS-28775},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-287755},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2022},
  abstract  = {Aims This study investigated, whether an activated R-mode, a surrogate of chronotropic incompetence in patients carrying a cardiovascular implantable electronic device (CIED), is associated with worse prognosis during and after an episode of acutely decompensated heart failure (AHF). Methods and Results Six hundred and twenty-three patients participating in an ongoing prospective cohort study that phenotypes and follows patients admitted for AHF were studied. We compared CIED carriers with R-mode stimulation (n=37) to CIED carriers not in R-mode (n=64) and patients without CIEDs (n=511). Mean heart rate on admission was significantly lower in R-mode patients vs. patients with CIED but without R-mode or patients withour CIED. In-hospital mortality was similar across groups, but age- and sex-adjusted 12-month mortality risk was higher in R-mode group. These effects persisted after multivariable adjustment for comorbidity burden. Conclusion In patients admitted for AHF, R-mode stimulation was associated with a significantly increased 12-month mortality risk. Our findings suggest that chronotropic incompetence per se mediates an adverse outcome and may not be adequately treated through accelerometer-based R-mode stimulation during and after an episode of AHF.},
  language  = {de}
}
@phdthesis{Muensterer2022,
  author    = {M{\"u}nsterer, Sascha Ottmar},
  title     = {Prognostische Wertigkeit der Herzfrequenz in Abh{\"a}ngigkeit von implantierten Devices bei akuter Herzinsuffizienz: Ergebnisse des prospektiven AHF-Registers W{\"u}rzburg},
  doi       = {10.25972/OPUS-33029},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-330293},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2022},
  abstract  = {Aims This study investigated, whether an activated R-mode, a surrogate of chronotropic incompetence in patients carrying a cardiovascular implantable electronic device (CIED), is associated with worse prognosis during and after an episode of acutely decompensated heart failure (AHF). Methods and Results Six hundred and twenty-three patients participating in an ongoing prospective cohort study that phenotypes and follows patients admitted for AHF were studied. We compared CIED carriers with R-mode stimulation (n=37) to CIED carriers not in R-mode (n=64) and patients without CIEDs (n=511). Mean heart rate on admission was significantly lower in R-mode patients vs. patients with CIED but without R-mode or patients withour CIED. In-hospital mortality was similar across groups, but age- and sex-adjusted 12-month mortality risk was higher in R-mode group. These effects persisted after multivariable adjustment for comorbidity burden. Conclusion In patients admitted for AHF, R-mode stimulation was associated with a significantly increased 12-month mortality risk. Our findings suggest that chronotropic incompetence per se mediates an adverse outcome and may not be adequately treated through accelerometer-based R-mode stimulation during and after an episode of AHF.},
  subject      = {Herzschrittmacher},
  language  = {de}
}
@phdthesis{Seer2024,
  author    = {Seer, Nadja},
  title     = {Prognostische Relevanz des Komorbidit{\"a}tenprofils bei Patienten mit akut dekompensierter Herzinsuffizienz und erhaltener Pumpfunktion (HFpEF)},
  doi       = {10.25972/OPUS-34796},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-347969},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Begleitend zu einer Herzinsuffizienz vorliegende Komorbidit{\"a}ten haben sowohl auf den Krankheitsverlauf als auch auf die Behandlung und Prognose solcher Patienten einen entscheidenden Einfluss. Ziel der vorliegenden Arbeit war es, Patienten mit Herzinsuffizienz mit erhaltener Pumpfunktion (HFpEF) anhand von elf begleitend zur Herzinsuffizienz vorliegenden Komorbidit{\"a}ten einer von sechs Ph{\"a}nogruppen zuzuteilen und diese Ph{\"a}nogruppen prognostisch einzusch{\"a}tzen. Dies wurde nach Vorlage der polytomen latenten Klassenanalyse (poLCA) von David Kao et al., ver{\"o}ffentlicht im Jahr 2015 im European Journal of Heart Failure, durchgef{\"u}hrt. Mithilfe einer poLCA k{\"o}nnen innerhalb einer Population Subgruppen mit {\"a}hnlichen Merkmalsauspr{\"a}gungen identifiziert werden. Die Patienten der vorliegenden Arbeit stammten aus dem Kollektiv des AHF (Acute-Heart-Failure-) Registers der Universit{\"a}tsklinik W{\"u}rzburg. Zus{\"a}tzlich wurde mit denselben elf Variablen eine von der Vergleichspublikation unabh{\"a}ngige poLCA f{\"u}r die Patienten des AHF-Registers erstellt, sowie eine dritte poLCA, die zus{\"a}tzlich die H{\"o}he des NT-proBNP ber{\"u}cksichtigte. Die Ergebnisse der Arbeit zeigten, dass die poLCA von Kao et al. durchaus auf andere Studienpopulationen {\"u}bertragen werden kann, um Patienten mit HFpEF im klinischen Alltag mit wenig Aufwand prognostisch einsch{\"a}tzen zu k{\"o}nnen. Mehr statistisch signifikante Ergebnisse wurden allerdings bei Anwendung einer eigenen poLCA f{\"u}r das AHF-Register erzielt. Die H{\"o}he des NT-proBNP hatte signifikanten Einfluss auf die Prognose und Klassenzuteilung eines Patienten.},
  subject      = {Herzinsuffizienz},
  language  = {de}
}
@phdthesis{Allgaier2024,
  author    = {Allgaier, Johannes},
  title     = {Machine Learning Explainability on Multi-Modal Data using Ecological Momentary Assessments in the Medical Domain},
  doi       = {10.25972/OPUS-35118},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-351189},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Introduction. Mobile health (mHealth) integrates mobile devices into healthcare, enabling remote monitoring, data collection, and personalized interventions. Machine Learning (ML), a subfield of Artificial Intelligence (AI), can use mHealth data to confirm or extend domain knowledge by finding associations within the data, i.e., with the goal of improving healthcare decisions. In this work, two data collection techniques were used for mHealth data fed into ML systems: Mobile Crowdsensing (MCS), which is a collaborative data gathering approach, and Ecological Momentary Assessments (EMA), which capture real-time individual experiences within the individual's common environments using questionnaires and sensors. We collected EMA and MCS data on tinnitus and COVID-19. About 15 \% of the world's population suffers from tinnitus. Materials \& Methods. This thesis investigates the challenges of ML systems when using MCS and EMA data. It asks: How can ML confirm or broad domain knowledge? Domain knowledge refers to expertise and understanding in a specific field, gained through experience and education. Are ML systems always superior to simple heuristics and if yes, how can one reach explainable AI (XAI) in the presence of mHealth data? An XAI method enables a human to understand why a model makes certain predictions. Finally, which guidelines can be beneficial for the use of ML within the mHealth domain? In tinnitus research, ML discerns gender, temperature, and season-related variations among patients. In the realm of COVID-19, we collaboratively designed a COVID-19 check app for public education, incorporating EMA data to offer informative feedback on COVID-19-related matters. This thesis uses seven EMA datasets with more than 250,000 assessments. Our analyses revealed a set of challenges: App user over-representation, time gaps, identity ambiguity, and operating system specific rounding errors, among others. Our systematic review of 450 medical studies assessed prior utilization of XAI methods. Results. ML models predict gender and tinnitus perception, validating gender-linked tinnitus disparities. Using season and temperature to predict tinnitus shows the association of these variables with tinnitus. Multiple assessments of one app user can constitute a group. Neglecting these groups in data sets leads to model overfitting. In select instances, heuristics outperform ML models, highlighting the need for domain expert consultation to unveil hidden groups or find simple heuristics. Conclusion. This thesis suggests guidelines for mHealth related data analyses and improves estimates for ML performance. Close communication with medical domain experts to identify latent user subsets and incremental benefits of ML is essential.},
  subject      = {Maschinelles Lernen},
  language  = {en}
}