EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Integrating Anomaly Detection and LLM-Based Explanation Generation in Clinical Data Dashboards

Hamidreza Maharlou, Paula Fierek, Hannah Benedictine Maier, Steffen Oeltze-Jafra

Studies in Health Technology and Informatics · 21.05.2026

DOI: 10.3233/shti260124 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, erschienen am 21.05.2026. Verfasst von Hamidreza Maharlou und 3 weiteren Autor:innen. Beteiligte Einrichtung: Peter L. Reichertz Institute, Hannover Medical School (MHH), Germany. Hauptthemen laut MEDLINE: Large Language Models, Dashboard Systems, Natural Language Processing, Electronic Health Records. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents an integrated approach that combines unsupervised anomaly detection with large language model (LLM)-based explanation generation to enhance the interpretability of clinical study dashboards. Using data from the P4D (Personalized, Predictive, Precise, and Preventive Medicine for Major Depression) cohort study, an Isolation Forest algorithm was employed to identify atypical observations within multidimensional clinical data. A t-distributed Stochastic Neighbor Embedding (t-SNE) projection was used to assess the plausibility of detected outliers. To improve transparency for clinical stakeholders, several current-generation LLMs, including GPT-5, GPT-4, Gemini 2.5 Flash, and Llama 3, were evaluated for automatically generating concise, patient-specific explanations of anomalies. The evaluation followed a reasoning-based framework comparing model explanations with clinical assessments. Results show GPT-5 provided the most clinically helpful and format-consistent natural language outputs, while Llama 3 showed limitations in contextual precision. The integration of these components into the existing P4D Dashboard [1] for clinical study monitoring supports clinicians in identifying data inconsistencies and understanding their likely causes, thereby improving the usability and trustworthiness of automated clinical data analyses.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Large Language ModelsDashboard SystemsNatural Language ProcessingElectronic Health RecordsHumansAlgorithms

Stichwörter der Autor:innen

Anomaly DetectionClinical DashboardsExplainable AILLMs

Autor:innen und Institutionen

  • Peter L. Reichertz Institute, Hannover Medical School (MHH), Germany.
  • Peter L. Reichertz Institute, Technical University of Braunschweig, Germany.
  • Department of Psychiatry, Social Psychiatry and Psychotherapy, MHH, Germany.

Zitieren

APA

Maharlou, H., Fierek, P., Maier, H. B., & Oeltze-Jafra, S. (2026). Integrating Anomaly Detection and LLM-Based Explanation Generation in Clinical Data Dashboards. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti260124

Vancouver

Maharlou H, Fierek P, Maier HB, Oeltze-Jafra S. Integrating Anomaly Detection and LLM-Based Explanation Generation in Clinical Data Dashboards. Studies in Health Technology and Informatics. 2026. doi:10.3233/shti260124

BibTeX
@article{Maharlou2026,
  author = {Maharlou, Hamidreza and Fierek, Paula and Maier, Hannah Benedictine and Oeltze-Jafra, Steffen},
  title = {Integrating Anomaly Detection and LLM-Based Explanation Generation in Clinical Data Dashboards},
  journal = {Studies in Health Technology and Informatics},
  year = {2026},
  doi = {10.3233/shti260124}
}
Kostenlos starten →