Medical guidelines have a significant role in the field of evidence-based medical treatment. The content of a medical guideline is based on a systematic review of clinical evidence with instructions and recommendations that clinicians can refer to. Most of the guidelines are available in an unstructured text format. Hence, clinicians must take a considerable time to search and find relevant recommendations in their semantic context. Using Machine Learning algorithms, automatic information extraction from medical guidelines has recently become possible. We present a novel system for information extraction and a fuzzy rule database developed for clinical guidelines.
1031 – 1035
Date of Publication:
17 May 2019
In this work, a new hybrid algorithm for disease risk classification is proposed. The proposed methodology is based on Dynamic Time Warping (DTW). This methodology can be applied to time series from various domains such as vital sign time series available in medical big data. To validate our methodology, we applied it to risk classification for sepsis, which is one of the most challenging problems within the area of medical data analysis. In the first step the algorithm uses different statistical properties of time series features. Furthermore, using differently labeled training data sets, we created a DTW Barycenter Averaging (DBA) on each feature.
In the second step, validation data sets and DTW are used to validate the precision of classification and the final results are compared. The performance of our methodology is validated with real medical data and on six different criteria definitions for the sepsis diseases. Results show that our algorithm performed, in the best case, with precision and recall of 96,38% and 90,90%, respectively.
Lejla Begic Fazlic, Ahmed Hallawa, Matthias Dziubany, Marlies Morgen, Jens Schneider, Marvin Schacht, Anke Schmeink, Lukas Martin, Arne Peine, Thomas Vollmer, Stefan Winter and Guido Dartmann, 8th International Conference on Cyber – Physical Systems and Internet-of-Things (CPS & IoT‘ 2020), Budva, Montenegro, 2020.
Date of Conference:
8-11 June 2020
Decision support systems in intensive care units are developed with safety, efficiency, and effectiveness in mind. In contrast, user experience (UX) for decision support systems has received limited attention in practice and research. In this paper, we present an application of the Experience Design approach from Hassenzahl  for a clinical decision support system (CDSS) for volume therapy – the administration of intravenous fluids – in intensive care units. In semi-structured interviews with five nurses, we gathered nurses‘ work activities around volume therapy, which are perceived as particularly beneficial for their well-being. Lead by the narratives gathered in the interviews we designed positive experience through the CDSS. The resulting concept of a CDSS for nurses aims to create a positive experience during volume therapy, which is supported by addressing and fulfilling the needs for competence and popularity.
This paper presents an incremental learning approach for estimating the structural parameters in stochastic state-based models (SSMs). SSMs have proven to be useful for modelling biological and medical processes, as they can represent both time dependency and stochastic processes. A major challenge in modelling in bioinformatics is that learning processes usually rely on large publicly accessible databases. In this work ,a new approach is presented, where models are trained incrementally locally at different data sources, e.g., hospitals, without having to pass on sensitive data. After learning, only the parameters of the model are passed on, in this case the arc weights of stochastic Petri nets. As a result, data protection and privacy of patients in hospitals are respected and it is no longer necessary to rely on the existence of a suitable accessible database. Simulations are used to evaluate the performance of the algorithm for a gene regulatory network.
R. Lipp, G. Dartmann, L. Fazlic, T. Vollmer, S. Winter, A. Peine, L. Martin and A. Schmeink, „Incremental Parameter Estimation of Stochastic State-Based Models“, IEEE 19th World Symposium on Applied Machine Intelligence and Informatics (SAMI2021), Herl’any, Slovakia, 2021.
The aim of this work was to develop and evaluate the reinforcement learning algorithm VentAI, which is able to suggest a dynamically optimized mechanical ventilation regime for critically-ill patients. We built, validated and tested its performance on 11,943 events of volume-controlled mechanical ventilation derived from 61,532 distinct ICU admissions and tested it on an independent, secondary dataset (200,859 ICU stays; 25,086 mechanical ventilation events). A patient „data fingerprint“ of 44 features was extracted as multidimensional time series in 4-hour time steps. We used a Markov decision process, including a reward system and a Q-learning approach, to find the optimized settings for positive end-expiratory pressure (PEEP), fraction of inspired oxygen (FiO2) and ideal body weight-adjusted tidal volume (Vt). The observed outcome was in-hospital or 90-day mortality. VentAI reached a significantly increased estimated performance return of 83.3 (primary dataset) and 84.1 (secondary dataset) compared to physicians‘ standard clinical care (51.1). The number of recommended action changes per mechanically ventilated patient constantly exceeded those of the clinicians. VentAI chose 202.9% more frequently ventilation regimes with lower Vt (5-7.5 mL/kg), but 50.8% less for regimes with higher Vt (7.5-10 mL/kg). VentAI recommended 29.3% more frequently PEEP levels of 5-7 cm H2O and 53.6% more frequently PEEP levels of 7-9 cmH2O. VentAI avoided high (>55%) FiO2 values (59.8% decrease), while preferring the range of 50-55% (140.3% increase). In conclusion, VentAI provides reproducible high performance by dynamically choosing an optimized, individualized ventilation strategy and thus might be of benefit for critically ill patients.
Arne Peine, Ahmed Hallawa, Johannes Bickenbach, Guido Dartmann, Lejla Begic Fazlic, Anke Schmeink, Gerd Ascheid, Christoph Thiemermann, Andreas Schuppert, Ryan Kindle, Leo Celi, Gernot Marx & Lukas Martin, „Development and validation of a reinforcement learning algorithm to dynamically optimize mechanical ventilation in critical care“, npj Digit. Med. 4, 32 (2021).
In this paper, we propose a method we call MEMEography for HCI research to understand people and their interactional contexts from the remixed internet memes they post in internet communities. While memes might not be the most obvious choice of a research subject, they allow us to investigate unfamiliar domains even when access to the field is beyond reach. We describe an initial approach of data selection, collection, prioritization and analysis. In addition, we demonstrate the kinds of insights we can gain through MEMEographies by analyzing a corpus of memes in the intensive care unit (ICU) context posted 2020 on Instagram. ICU memes open up insights into the environment, work practices, challenges, emotions and familiarized us with ICU practitioners’ language, even though access to an actual ICU was completely impossible during 2020.
Annika Kaltenhauser, Nađa Terzimehić, Andreas Butz, „MEMEography: Understanding Users Through Internet Memes“, CHI EA ’21: Extended Abstracts of the 2021 CHI Conference on Human Factors in Computing Systems, Japan (remote), 2021.