A5030577380- Machine Learning in Healthcare
- Sepsis Diagnosis and Treatment
- Clinical Nutrition and Gastroenterology
- Bone health and osteoporosis research
- Bacterial Identification and Susceptibility Testing
- Topic Modeling
- Clinical practice guidelines implementation
- Bone health and treatments
- Hydrological Forecasting Using AI
- Blood Pressure and Hypertension Studies
- Bone and Joint Diseases
- Metabolomics and Mass Spectrometry Studies
- Pharmacovigilance and Adverse Drug Reactions
- Health Systems, Economic Evaluations, Quality of Life
- Dysphagia Assessment and Management
- Expert finding and Q&A systems
- Child Nutrition and Feeding Issues
Leipzig University
2022-2024
University Hospital Leipzig
2022-2024
Abstract Background Timely diagnosis is crucial for sepsis treatment. Current machine learning (ML) models suffer from high complexity and limited applicability. We therefore created an ML model using only complete blood count (CBC) diagnostics. Methods collected non-intensive care unit (non-ICU) data a German tertiary centre (January 2014 to December 2021). Using patient age, sex, CBC parameters (haemoglobin, platelets, mean corpuscular volume, white red cells), we trained boosted random...
Background: The refeeding syndrome (RFS) is an oftentimes-unrecognized complication of reintroducing nutrition in malnourished patients that can lead to fatal cardiovascular failure. We hypothesized a clinical decision support system (CDSS) improve RFS recognition and management. Methods: developed algorithm from current diagnostic criteria for detection, tested the on retrospective dataset combined final with therapy referral recommendations knowledge-based CDSS. CDSS integration into...
Abstract Background Delay in diagnosing sepsis results potentially preventable deaths. Mainly due to their complexity or limited applicability, machine learning (ML) models predict have not yet become part of clinical routines. For this reason, we created a ML model that only requires complete blood count (CBC) diagnostics. Methods Non-intensive care unit (non-ICU) data from German tertiary centre were collected January 2014 December 2021. Patient age, sex, and CBC parameters (haemoglobin,...
Abstract Machine learning is a powerful tool to develop algorithms for clinical diagnosis. However, standard machine are not perfectly suited data since the interconnected and may contain time series. As shown recommender systems molecular property predictions graph neural networks (GNNs) represent alternative. In this study, we evaluated performance consumption of GNNs compared state-of-the-art on classification sepsis from blood count as well importance slope each feature final...
Abstract Objectives Severe hypo- and hypercalcemia are common urgent treatment is recommended. Free calcium (fCa) the gold standard but needs blood gas tests with challenging preanalytics. Total (tCa) calculated adjusted (aCa) readily available, their interpretation hampered by identical tCa aCa cutoffs, laborious local calculation difficult comparability of biomarkers. Methods Laboratory results from University Medicine Leipzig were evaluated over a five-year period (236,274 patients). A...