Machine learning-based modeling of physical systems has attracted significant interest in recent years. Based solely on the underlying physical equations and initial and boundary conditions, these new approaches allow to approximate, for example, the complex flow of blood in the case of fluid dynamics. Physics-informed neural networks offer certain advantages compared to conventional computational fluid dynamics methods as they avoid the need for discretized meshes and allow to readily solve inverse problems and integrate additional data into the algorithms. Today, the majority of published reports on learning-based flow modeling relies on fully-connected neural networks. However, many different network architectures are introduced into deep learning each year, each with specific benefits for certain applications. In this paper, we present the first comprehensive comparison of various state-of-the-art networks and evaluate their performance in terms of computational cost and accuracy relative to numerical references. We found that while fully-connected networks offer an attractive balance between training time and accuracy, more elaborate architectures (e.g., Deep Galerkin Method) generated superior results. Moreover, we observed high accuracy in simple cylindrical geometries, but slightly poorer estimates in complex aneurysms. This paper provides quantitative guidance for practitioners interested in complex flow modeling using physics-based deep learning.

In burn medicine, the percentage of the burned body surface area (TBSA-B) to the total body surface area (TBSA) is a crucial parameter to ensure adequate treatment and therapy. Inaccurate estimations of the burn extent can lead to wrong medical decisions resulting in considerable consequences for patients. These include, for instance, over-resuscitation, complications due to fluid aggregation from burn edema, or non-optimal distribution of patients. Due to the frequent inaccurate TBSA-B estimation in practice, objective methods allowing for precise assessments are required. Over time, various methods have been established whose development has been influenced by contemporary technical standards. This article provides an overview of the history of burn size estimation and describes existing methods with a critical view of their benefits and limitations. Traditional methods that are still of great practical relevance were developed from the middle of the 20th century. These include the “Lund Browder Chart”, the “Rule of Nines”, and the “Rule of Palms”. These methods have in common that they assume specific values for different body parts’ surface as a proportion of the TBSA. Due to the missing consideration of differences regarding sex, age, weight, height, and body shape, these methods have practical limitations. Due to intensive medical research, it has been possible to develop three-dimensional computer-based systems that consider patients’ body characteristics and allow a very realistic burn size assessment. To ensure high-quality burn treatment, comprehensive documentation of the treatment process, and wound healing is essential. Although traditional paper-based documentation is still used in practice, it no longer meets modern requirements. Instead, adequate documentation is ensured by electronic documentation systems. An illustrative software already being used worldwide is “BurnCase 3D”. It allows for an accurate burn size assessment and a complete medical documentation.

Electronic health records (EHRs) have been successfully used in data science and machine learning projects. However, most of these data are collected for clinical use rather than for retrospective analysis. This means that researchers typically face many different issues when attempting to access and prepare the data for secondary use. We aimed to investigate how raw EHRs can be accessed and prepared in retrospective data science projects in a disciplined, effective, and efficient way. We report our experience and findings from a large-scale data science project analyzing routinely acquired retrospective data from the Kepler University Hospital in Linz, Austria. The project involved data collection from more than 150,000 patients over a period of 10 years. It included diverse data modalities, such as static demographic data, irregularly acquired laboratory test results, regularly sampled vital signs, and high-frequency physiological waveform signals. Raw medical data can be corrupted in many unexpected ways that demand thorough manual inspection and highly individualized data cleaning solutions. We present a general data preparation workflow, which was shaped in the course of our project and consists of the following 7 steps: obtain a rough overview of the available EHR data, define clinically meaningful labels for supervised learning, extract relevant data from the hospital’s data warehouses, match data extracted from different sources, deidentify them, detect errors and inconsistencies therein through a careful exploratory analysis, and implement a suitable data processing pipeline in actual code. Only few of the data preparation issues encountered in our project were addressed by generic medical data preprocessing tools that have been proposed recently. Instead, highly individualized solutions for the specific data used in one’s own research seem inevitable. We believe that the proposed workflow can serve as a guidance for practitioners, helping them to identify and address potential problems early and avoid some common pitfalls.

Burn care is a highly relevant medical specialty in every part of the world. Different infrastructure, healthcare systems and access to medical supplies lead to different needs, treatment strategies and outcomes. A fundamental tool in a burn care provider's armamentarium is the use of different dressings. Several studies have investigated the question of the ideal burn dressing, but none could achieve a proper global perspective. With advanced dressings being on the rise, we conducted this study to get a global understanding of the actual use and idea of the ideal burn dressing. The objective of this study was to investigate the understanding of an 'ideal burn dressing' on a global scale. A questionnaire about burn care and the ideal burn dressing has been created and translated to five of the most spoken languages world-wide (English, Spanish, French, Chinese, Indonesian). It has been uploaded to an online survey platform and sent out to burn experts worldwide. The voluntary participation was possible for a period of four weeks. In total, 196 respondents from 49 countries participated in the study, yielding a response rate of 24.5%. The most important burn dressing characteristics in a cumulative ranking were (1) lack of adhesion (80.54%), (2) pain-free dressing change (79.87%), (3) requirement of fewer dressing changes, while in a linear ranking they were (1) anti-infective (35.14% 1st), pain-reduction (24.14% 2nd), and high absorbency (23.49% 3rd). Silver-based dressings are the most used dressings for superficial (45.21%) and deep (52.78%). 94.81% believe that the choice of burn dressing affects the outcome. This investigation has delivered valuable insights into the global perspective of the ideal burn dressing. Yet, the question of the ideal burn dressing is still inconclusive. Wound dressing research is of fundamental interest for patients, healthcare providers and healthcare systems.