Website under construction: the content is provisional and not legally binding. Learn more
DAISY brings together valuable clinical information from different systems and promptly alerts the care team to suspected sepsis and other critical conditions.
Participating institutions
Participating companies
Integrated with clinical systems from
Research
DAISY links laboratory results with clinical data and identifies indications of acute conditions. The research covers the development and validation of clinical algorithms and their integration into routine care.
| Priority | Description |
|---|---|
| Early detection of sepsis | Development and validation of a machine learning model that identifies an increased risk of sepsis from the complete blood count. |
| Detection of sample mix-ups | Identification of implausible result sequences that indicate a mix-up of blood samples (wrong blood in tube). |
| Integration into clinical routine | Delivery of alerts into the hospital’s existing information systems, without additional applications. |
| Acceptance in everyday clinical practice | Evaluation of alerting with regard to relevance, acceptance and the avoidance of alarm fatigue. |
| Open source | Release of selected components as open-source software to make methods transparent, traceable and reusable.DAISY on GitHub |
Data sources
Connection of further clinical systems, such as radiology.
DAISY core
Consolidation of relevant patient data from different clinical systems.
Analysis of the data using defined criteria and clinical algorithms.
Targeted delivery, escalation and acknowledgement of alerts.
Information and alerting
Colour-coded risk indicator in the ward overview.
Result as a separate value on the laboratory order.
Note on the printed report.
Further information on each alert.
The laboratory calls the ward for urgent results.
Users
Receive alerts within their existing workflow.
Validates critical alerts and agrees on further steps with the wards.
Feedback from clinical practice is continuously fed into further development
Roadmap
DAISY is being developed step by step: from the validation of individual algorithms through pilot operation to use as an approved medical device.
Development and validation of a machine learning model that detects an increased risk of sepsis from the complete blood count.
Use of the sepsis algorithm at the Detmold site of Klinikum Lippe, on ward 1E and in the emergency department.
Integration of further algorithms, including acute kidney injury, hypokalaemia, liver fibrosis (FIB-4) and the detection of sample mix-ups.
Studies on alert behaviour, acceptance and clinical benefit based on data collected in operation.
Integration of additional clinical data sources, such as radiology.
Preparation of the approval of DAISY as a medical device.
Publications
Peer-reviewed publications from the DAISY project.
| Year | Publication |
|---|---|
| 2025 | Analyte Importance Analysis in Machine Learning-Based Detection of Wrong-Blood-in-Tube Errors Using Complete Blood Count Data Barış Gün Sürmeli, René Staritzbichler, Clemens Ringel, Saleem Al-Dakkak, Helene Dörksen, Thorsten Kaiser Journal of Personalized Medicine 15(9):404 (2025) · DOI 10.3390/jpm15090404 |
Press
Media coverage of the DAISY project and the research at the University Institute for Laboratory Medicine. The articles are in German.
Klinikum Lippe ·
Prof. Dr. med. Thorsten Kaiser und Dr. René Staritzbichler erhalten den mit 40.000 Euro dotierten Jörg Schwarzbich Inventor Award für ihre Entwicklung in der Labordiagnostik.
Read article
Aktuell Uni Bielefeld ·
Der „Spectimprover“ verbindet Ramanspektroskopie mit Künstlicher Intelligenz und misst mehrere Biomarker gleichzeitig, ohne Reagenzien. Die Universität Bielefeld würdigt die Arbeit mit dem Inventor Award.
Read article
cibX Magazin ·
Der Beitrag ordnet Sepsis als weltweites Gesundheitsproblem ein und stellt DAISY vor: ein KI-Modell, das ein Sepsisrisiko anhand eines einfachen Blutbilds erkennt.
Read article
kma Online ·
Ein interdisziplinäres Forschungsteam hat ein KI-Modell zur Sepsis-Früherkennung entwickelt. Grundlage ist eine einfache Blutuntersuchung.
Read articleTeam and partners
Institute
University Institute for Laboratory Medicine, Microbiology and Clinical Pathobiochemistry
Hospital
Campus Klinikum Lippe, University Hospital OWL of Bielefeld University
Funding
To be announced
Univ.-Prof. Dr. René Staritzbichler · Saleem Al Dakkak · Univ.-Prof. Dr. Thorsten Kaiser · Dr. Clemens Friedrich Ringel
M. Struck · B. G. Sürmeli · M. Tashtiba · M. Sure · P. Kaftan
FAQ
DAISY (Digital Alert and Information System) is a clinical decision support system with a focus on laboratory medicine, developed at Klinikum Lippe. It evaluates incoming laboratory results automatically and informs the care team when there are indications of sepsis and other acute conditions.
No. An alert indicates a markedly increased risk and does not replace a diagnosis. The absence of an alert does not rule out a disease. The medical decision always rests with the care team.
DAISY only processes data that are already available as part of routine care: laboratory results from the laboratory information system (LIS), data from the hospital information system (HIS) and spectroscopy results. The sepsis algorithm only needs the complete blood count, age and sex; no additional blood sample is required.
The sepsis algorithm is in pilot operation at the Detmold site of Klinikum Lippe, on ward 1E and in the emergency department. It evaluates around 500 blood counts per day on average.
No. DAISY runs entirely within the IT infrastructure of Klinikum Lippe.
Contact
We welcome enquiries about partnerships, scientific exchange or joining the project.
Send emailUniversitätsinstitut für Laboratoriumsmedizin, Mikrobiologie und klinische Pathobiochemie
Univ.-Prof. Dr. med. Thorsten Kaiser
Klinikum Lippe, Röntgenstraße 18, 32756 Detmold
+49 5231 72-1231 (Office)