Combining geometric and probabilistic reasoning for computer-based penetrating-trauma assessment

Digitalization in orthopaedics: a narrative review

Advances in technology and digital tools like the Internet of Things (IoT), artificial intelligence (AI), and sensors are shaping the field of orthopaedic surgery on all levels, from patient care to research and facilitation of logistic processes. Especially the COVID-19 pandemic, with the associated contact restrictions was an accelerator for the development and introduction of telemedical applications and digital alternatives to classical in-person patient care. Digital applications already used in orthopaedic surgery include telemedical support, online video consultations, monitoring of patients using wearables, smart devices, surgical navigation, robotic-assisted surgery, and applications of artificial intelligence in forms of medical image processing, three-dimensional (3D)-modelling, and simulations. In addition to that immersive technologies like virtual, augmented, and mixed reality are increasingly used in training but also rehabilitative and surgical settings. Digital advances can therefore increase the accessibility, efficiency and capabilities of orthopaedic services and facilitate more data-driven, personalized patient care, strengthening the self-responsibility of patients and supporting interdisciplinary healthcare providers to offer for the optimal care for their patients.

Decision support by machine learning systems for acute management of severely injured patients: A systematic review

Introduction

Treating severely injured patients requires numerous critical decisions within short intervals in a highly complex situation. The coordination of a trauma team in this setting has been shown to be associated with multiple procedural errors, even of experienced care teams. Machine learning (ML) is an approach that estimates outcomes based on past experiences and data patterns using a computer-generated algorithm. This systematic review aimed to summarize the existing literature on the value of ML for the initial management of severely injured patients.

Methods

We conducted a systematic review of the literature with the goal of finding all articles describing the use of ML systems in the context of acute management of severely injured patients. MESH search of Pubmed/Medline and Web of Science was conducted. Studies including fewer than 10 patients were excluded. Studies were divided into the following main prediction groups: (1) injury pattern, (2) hemorrhage/need for transfusion, (3) emergency intervention, (4) ICU/length of hospital stay, and (5) mortality.

Results

Thirty-six articles met the inclusion criteria; among these were two prospective and thirty-four retrospective case series. Publication dates ranged from 2000 to 2020 and included 32 different first authors. A total of 18,586,929 patients were included in the prediction models. Mortality was the most represented main prediction group (n = 19). ML models used were artificial neural network ( n = 15), singular vector machine (n = 3), Bayesian network (n = 7), random forest (n = 6), natural language processing (n = 2), stacked ensemble classifier [SuperLearner (SL), n = 3], k-nearest neighbor (n = 1), belief system (n = 1), and sequential minimal optimization (n = 2) models. Thirty articles assessed results as positive, five showed moderate results, and one article described negative results to their implementation of the respective prediction model.

Conclusions

While the majority of articles show a generally positive result with high accuracy and precision, there are several requirements that need to be met to make the implementation of such models in daily clinical work possible. Furthermore, experience in dealing with on-site implementation and more clinical trials are necessary before the implementation of ML techniques in clinical care can become a reality.

[Computer-assisted decision-making for trauma patients]

BACKGROUND

In the management of trauma patients in the resuscitation room many time-pressured and critical decisions must continuously be made in complex situations. Even experienced teams frequently make errors in this context. Computer-assisted decision-making systems can predict critical situations based on patient data continuously acquired online. Based on the calculated predictions these systems can suggest the next steps in managing the patient. This review summarizes the current literature on computer-assisted decision-making in the management of trauma patients.

OBJECTIVE

A literature review summarizing existing concepts and applications of computer-assisted decision-making support for the management of trauma patients.

METHODS

Narrative review article based on an analysis of literature in the German and English languages from the last 10 years.

RESULTS

There exist numerous computer-assisted decision-making systems in the field of trauma care. Several studies could show that computer-assisted decision-making can improve the outcome in the preclinical setting, the resuscitation room and in the intensive care unit. For further validation and implementation of these systems, information technological barriers have to be overcome, existing systems need to be adapted to current data protection regulations and large multicenter studies are necessary.

CONCLUSION

Computer-assisted decision-making can help to improve the management of trauma patients; however, before a ubiquitous implementation a number of technological and legislative barriers have to be overcome.

A critical review of anaesthetised animal models and alternatives for military research, testing and training, with a focus on blast damage, haemorrhage and resuscitation

Military research, testing, and surgical and resuscitation training, are aimed at mitigating the consequences of warfare and terrorism to armed forces and civilians. Traumatisation and tissue damage due to explosions, and acute loss of blood due to haemorrhage, remain crucial, potentially preventable, causes of battlefield casualties and mortalities. There is also the additional threat from inhalation of chemical and aerosolised biological weapons. The use of anaesthetised animal models, and their respective replacement alternatives, for military purposes -- particularly for blast injury, haemorrhaging and resuscitation training -- is critically reviewed. Scientific problems with the animal models include the use of crude, uncontrolled and non-standardised methods for traumatisation, an inability to model all key trauma mechanisms, and complex modulating effects of general anaesthesia on target organ physiology. Such effects depend on the anaesthetic and influence the cardiovascular system, respiration, breathing, cerebral haemodynamics, neuroprotection, and the integrity of the blood-brain barrier. Some anaesthetics also bind to the NMDA brain receptor with possible differential consequences in control and anaesthetised animals. There is also some evidence for gender-specific effects. Despite the fact that these issues are widely known, there is little published information on their potential, at best, to complicate data interpretation and, at worst, to invalidate animal models. There is also a paucity of detail on the anaesthesiology used in studies, and this can hinder correct data evaluation. Welfare issues relate mainly to the possibility of acute pain as a side-effect of traumatisation in recovered animals. Moreover, there is the increased potential for animals to suffer when anaesthesia is temporary, and the procedures invasive. These dilemmas can be addressed, however, as a diverse range of replacement approaches exist, including computer and mathematical dynamic modelling of the human body, cadavers, interactive human patient simulators for training, in vitro techniques involving organotypic cultures of target organs, and epidemiological and clinical studies. While the first four of these have long proven useful for developing protective measures and predicting the consequences of trauma, and although many phenomena and their sequelae arising from different forms of trauma in vivo can be induced and reproduced in vitro, non-animal approaches require further development, and their validation and use need to be coordinated and harmonised. Recommendations to these ends are proposed, and the scientific and welfare problems associated with animal models are addressed, with the future focus being on the use of batteries of complementary replacement methods deployed in integrated strategies, and on greater transparency and scientific cooperation.

An Approach for Incorporating Context in Building Probabilistic Predictive Models

With the increasing amount of information collected through clinical practice and scientific experimentation, a growing challenge is how to utilize available resources to construct predictive models to facilitate clinical decision making. Clinicians often have questions related to the treatment and outcome of a medical problem for individual patients; however, few tools exist that leverage the large collection of patient data and scientific knowledge to answer these questions. Without appropriate context, existing data that have been collected for a specific task may not be suitable for creating new models that answer different questions. This paper presents an approach that leverages available structured or unstructured data to build a probabilistic predictive model that assists physicians with answering clinical questions on individual patients. Various challenges related to transforming available data to an end-user application are addressed: problem decomposition, variable selection, context representation, automated extraction of information from unstructured data sources, model generation, and development of an intuitive application to query the model and present the results. We describe our efforts towards building a model that predicts the risk of vasospasm in aneurysm patients.

Predicting response to short-acting bronchodilator medication using Bayesian networks

AIMS

Bronchodilator response tests measure the effect of beta(2)-agonists, the most commonly used short-acting reliever drugs for asthma. We sought to relate candidate gene SNP data with bronchodilator response and measure the predictive accuracy of a model constructed with genetic variants.

MATERIALS & METHODS

Bayesian networks, multivariate models that are able to account for simultaneous associations and interactions among variables, were used to create a predictive model of bronchodilator response using candidate gene SNP data from 308 Childhood Asthma Management Program Caucasian subjects.

RESULTS

The model found that 15 SNPs in 15 genes predict bronchodilator response with fair accuracy, as established by a fivefold cross-validation area under the receiver-operating characteristic curve of 0.75 (standard error: 0.03).

CONCLUSION

Bayesian networks are an attractive approach to analyze large-scale pharmacogenetic SNP data because of their ability to automatically learn complex models that can be used for the prediction and discovery of novel biological hypotheses.

Prediction of chronic obstructive pulmonary disease (COPD) in asthma patients using electronic medical records

OBJECTIVE

Identify clinical factors that modulate the risk of progression to COPD among asthma patients using data extracted from electronic medical records.

DESIGN

Demographic information and comorbidities from adult asthma patients who were observed for at least 5 years with initial observation dates between 1988 and 1998, were extracted from electronic medical records of the Partners Healthcare System using tools of the National Center for Biomedical Computing "Informatics for Integrating Biology to the Bedside" (i2b2).

MEASUREMENTS

A predictive model of COPD was constructed from a set of 9,349 patients (843 cases, 8,506 controls) using Bayesian networks. The model's predictive accuracy was tested using it to predict COPD in a future independent set of asthma patients (992 patients; 46 cases, 946 controls), who had initial observation dates between 1999 and 2002.

RESULTS

A Bayesian network model composed of age, sex, race, smoking history, and 8 comorbidity variables is able to predict COPD in the independent set of patients with an accuracy of 83.3%, computed as the area under the Receiver Operating Characteristic curve (AUROC).

CONCLUSIONS

Our results demonstrate that data extracted from electronic medical records can be used to create predictive models. With improvements in data extraction and inclusion of more variables, such models may prove to be clinically useful.

A comparison of methods for assessing penetrating trauma on retrospective multi-center data

OBJECTIVE

TraumaSCAN-Web (TSW) is a computerized decision support system for assessing chest and abdominal penetrating trauma which utilizes 3D geometric reasoning and a Bayesian network with subjective probabilities obtained from an expert. The goal of the present study is to determine whether a trauma risk prediction approach using a Bayesian network with a predefined structure and probabilities learned from penetrating trauma data is comparable in diagnostic accuracy to TSW.

METHODS

Parameters for two Bayesian networks with expert-defined structures were learned from 637 gunshot and stab wound cases from three hospitals, and diagnostic accuracy was assessed using 10-fold cross-validation. The first network included information on external wound locations, while the second network did not. Diagnostic accuracy of learned networks was compared to that of TSW on 194 previously evaluated cases.

RESULTS

For 23 of the 24 conditions modeled by TraumaSCAN-Web, 16 conditions had Areas Under the ROC Curve (AUCs) greater than 0.90 while 21 conditions had AUCs greater than 0.75 for the first network. For the second network, 16 and 20 conditions had AUCs greater than 0.90 and 0.75, respectively. AUC results for learned networks on 194 previously evaluated cases were better than or equal to AUC results for TSW for all diagnoses evaluated except diaphragm and heart injuries.

CONCLUSIONS

For 23 of the 24 penetrating trauma conditions studied, a trauma diagnosis approach using Bayesian networks with predefined structure and probabilities learned from penetrating trauma data was better than or equal in diagnostic accuracy to TSW. In many cases, information on wound location in the first network did not significantly add to predictive accuracy. The study suggests that a decision support approach that uses parameter-learned Bayesian networks may be sufficient for assessing some penetrating trauma conditions.

Using ontologies linked with geometric models to reason about penetrating injuries

OBJECTIVE

Medical assessment of penetrating injuries is a difficult and knowledge-intensive task, and rapid determination of the extent of internal injuries is vital for triage and for determining the appropriate treatment. Physical examination and computed tomographic (CT) imaging data must be combined with detailed anatomic, physiologic, and biomechanical knowledge to assess the injured subject. We are developing a methodology to automate reasoning about penetrating injuries using canonical knowledge combined with specific subject image data.

METHODS AND MATERIAL

In our approach, we build a three-dimensional geometric model of a subject from segmented images. We link regions in this model to entities in two knowledge sources: (1) a comprehensive ontology of anatomy containing organ identities, adjacencies, and other information useful for anatomic reasoning and (2) an ontology of regional perfusion containing formal definitions of arterial anatomy and corresponding regions of perfusion. We created computer reasoning services ("problem solvers") that use the ontologies to evaluate the geometric model of the subject and deduce the consequences of penetrating injuries.

RESULTS

We developed and tested our methods using data from the Visible Human. Our problem solvers can determine the organs that are injured given particular trajectories of projectiles, whether vital structures--such as a coronary artery--are injured, and they can predict the propagation of injury ensuing after vital structures are injured.

CONCLUSION

We have demonstrated the capability of using ontologies with medical images to support computer reasoning about injury based on those images. Our methodology demonstrates an approach to creating intelligent computer applications that reason with image data, and it may have value in helping practitioners in the assessment of penetrating injury.

Methods for reasoning from geometry about anatomic structures injured by penetrating trauma

This paper presents the methods used for three-dimensional (3D) reasoning about anatomic structures affected by penetrating trauma in TraumaSCAN-Web, a platform-independent decision support system for evaluating the effects of penetrating trauma to the chest and abdomen. In assessing outcomes for an injured patient, TraumaSCAN-Web utilizes 3D models of anatomic structures and 3D models of the regions of damage associated with stab and gunshot wounds to determine the probability of injury to anatomic structures. Probabilities estimated from 3D reasoning about affected anatomic structures serve as input to a Bayesian network which calculates posterior probabilities of injury based on these initial probabilities together with available information about patient signs, symptoms and test results. In addition to displaying textual descriptions of conditions arising from penetrating trauma to a patient, TraumaSCAN-Web allows users to visualize the anatomy suspected of being injured in 3D, in this way providing a guide to its reasoning process.

Evaluating the discriminatory power of a computer-based system for assessing penetrating trauma on retrospective multi-center data

OBJECTIVE

To evaluate the discriminatory power of TraumaSCAN-Web, a system for assessing penetrating trauma, using retrospective multi-center case data for gunshot and stab wounds to the thorax and abdomen.

METHODS

80 gunshot and 114 stab cases were evaluated using TraumaSCAN-Web. Areas under the Receiver Operator Characteristic Curves (AUC) were calculated for each condition modeled in TraumaSCAN-Web.

RESULTS

Of the 23 conditions modeled by TraumaSCAN-Web, 19 were present in either the gunshot or stab case data. The gunshot AUCs ranged from 0.519 (pericardial tamponade) to 0.975 (right renal injury). The stab AUCs ranged from 0.701 (intestinal injury) to 1.000 (tracheal injury).