Prediction of anemia using facial images and deep learning technology in the emergency department

Deep learning allows computational models that are composed of multiple processing layers to learn representations of data with multiple levels of abstraction. These methods have dramatically improved the state-of-the-art in speech recognition, visual object recognition, object detection and many other domains such as drug discovery and genomics. Deep learning discovers intricate structure in large data sets by using the backpropagation algorithm to indicate how a machine should change its internal parameters that are used to compute the representation in each layer from the representation in the previous layer. Deep convolutional nets have brought about breakthroughs in processing images, video, speech and audio, whereas recurrent nets have shone light on sequential data such as text and speech.

Background Severe traumatic injury continues to present challenges to healthcare systems around the world, and post-traumatic bleeding remains a leading cause of potentially preventable death among injured patients. Now in its fifth edition, this document aims to provide guidance on the management of major bleeding and coagulopathy following traumatic injury and encourages adaptation of the guiding principles described here to individual institutional circumstances and resources. Methods The pan-European, multidisciplinary Task Force for Advanced Bleeding Care in Trauma was founded in 2004, and the current author group included representatives of six relevant European professional societies. The group applied a structured, evidence-based consensus approach to address scientific queries that served as the basis for each recommendation and supporting rationale. Expert opinion and current clinical practice were also considered, particularly in areas in which randomised clinical trials have not or cannot be performed. Existing recommendations were re-examined and revised based on scientific evidence that has emerged since the previous edition and observed shifts in clinical practice. New recommendations were formulated to reflect current clinical concerns and areas in which new research data have been generated. Results Advances in our understanding of the pathophysiology of post-traumatic coagulopathy have supported improved management strategies, including evidence that early, individualised goal-directed treatment improves the outcome of severely injured patients. The overall organisation of the current guideline has been designed to reflect the clinical decision-making process along the patient pathway in an approximate temporal sequence. Recommendations are grouped behind the rationale for key decision points, which are patient- or problem-oriented rather than related to specific treatment modalities. While these recommendations provide guidance for the diagnosis and treatment of major bleeding and coagulopathy, emerging evidence supports the author group’s belief that the greatest outcome improvement can be achieved through education and the establishment of and adherence to local clinical management algorithms. Conclusions A multidisciplinary approach and adherence to evidence-based guidance are key to improving patient outcomes. If incorporated into local practice, these clinical practice guidelines have the potential to ensure a uniform standard of care across Europe and beyond and better outcomes for the severely bleeding trauma patient. Electronic supplementary material The online version of this article (10.1186/s13054-019-2347-3) contains supplementary material, which is available to authorized users.