Engineering

Evaluating how anesthesia trainees identify and treat malignant hyperthermia in the team-based simulation training with Epistemic Network Analysis

Technological advances in society have given the opportunity to modify and refine the education and experiences that occur within medical schools and the training provided to healthcare professionals. Clinical simulations provide team-based training to students and healthcare providers in ways that they otherwise may not receive through lectures and real-life settings. This study aims to analyze team communication during team-based training simulations on malignant hyperthermia (MH). The simulations were made available for analysis and evaluation through about 25 video recordings that were collected by the Clinical Simulation Center at the University of Michigan. MH is a rare but severe reaction to certain anesthetic drugs which requires the teams to work together to identify the symptoms (increasing end-tidal carbon dioxide, heart rate muscle rigidity, and temperature), properly treat the patient, and recognize the patient’s response to appropriate and inappropriate management interventions. The videos consist of different teams doing the same training on MH which may give insight as to why some teams are more successful than others. The analysis will also focus on how trainees allocate tasks efficiently within a group and develop individual/team skills in treating MH. The ability to automate computer assessments will also be evaluated to identify technical and nontechnical skills in team-based simulation training while treating MH. With improved training in a simulation environment, healthcare professionals will be able to improve their ability to work together in real environments where patient lives may be at risk.

Evaluating how anesthesia trainees identify and treat malignant hyperthermia in the team-based simulation training with Epistemic Network Analysis Read More »

Development of protein property prediction methods from sequence based on deep learning

harryjy

Although proteins have become increasingly easier to sequence, experimental determination of a protein’s structure remains difficult and time-consuming. Therefore, the prediction of a protein’s structure and properties based on its sequence is a key challenge in making better use of the vast amount of sequencing data. Our project seeks to develop a deep-learning-based method that uses a protein’s sequence to predict its properties, such as phi/psi angles and solvent accessibility. The initial goal of the project was to design and write the deep-learning program using the PyTorch library. After completion of the program, we assembled a training and testing set based on existing data from the Protein Data Bank and used the training data to train the model. We then ran the testing dataset and analyzed the results by comparing the predicted properties to the experimentally determined ones. While we do not have any results yet, we hope to be able to make conclusions about the relative effectiveness of the model we design compared to existing models for prediction. The results we obtain could help us determine which prediction techniques or algorithms are well-suited to this task, or which ones lead to errors and thus may need to be avoided in future research. The results could also contribute to improving the accuracy and efficiency of computational protein structure prediction, allowing scientists to make better use of the available sequencing data without the difficulties of experimental determination.

Development of protein property prediction methods from sequence based on deep learning Read More »

Building A Simulated World

conleth

Due to the high demand for motor vehicles, one of the key concerns for the automotive industry is to make vehicles accessible and safe. Both real test vehicles and driving simulators are used to assess the quality and performance of vehicles. However, many driving simulators are too expensive to purchase, too complex to use, take too long to run the software, and sometimes lack the desired functional characteristics. The goal of the research is to build a virtual world and an easy-to-use virtual driving simulator platform through the creative use of free software like CARLA and RoadRunner. This driving simulator will be suitable to support research on driver distraction, driver workload, and driver interfaces for partially automated vehicles. This will also inspire qualified people to use the simulation and work on safety on roads.

Building A Simulated World Read More »

Application for Cardiac Signal Visualization and Annotation

stefanhg

Massive amounts of data are being generated at a high velocity in the medical industry, at a rate and dimension that humans cannot catch up understanding them. Machine learning (ML) has the potential to analyze the medical time series data to support physicians in clinical decision making. However, in order to build useful machine learning (ML) algorithms, it is important to have annotated medical data set that can be used to train ML models. This project at the start is about building a viewer for surface ECG and intracardiac electrogram signals acquired during cardiac ablation procedures. Building the framework at early stages to support basic functionality is engineering-oriented with design decisions. We are currently developing a web application in Python. Specifications were created based upon the needed functionality of the application after talking with multiple cardiologists. The next step would be inviting medical students and cardiologists to use the app and gather feedback. Eventually, we want to build a platform/infrastructure that allows for development and evaluation of ML algorithms, and to improve cardiovascular disease management and treatment. By working with cardiologists and intelligent algorithms, we want to build a tool that can quickly review, search, annotate and analyze signals at high throughput. The end goal is that researchers across the country can use our tool to review their own data, and also deploy a real-time graphical decision support tool to assist cardiac procedures.

Application for Cardiac Signal Visualization and Annotation Read More »

Synthesizing reactions between metal chlorides and hexamethyldisilazane

cmsaid

The reaction of metal chlorides with hexamethyldisilazane (HMDS) provides a novel set of syntheses generating metal-hexamethyldisilazide compounds (M-HMDS), which can be used as precursors for functional ceramic materials. The current methods to synthesize M-HMDS are expensive and require multi-step processes; for example, at the moment lithium aluminum hydride is needed to create Al-HMDS, which creates harmful hydrogen gas. Our method of synthesizing Al-HMDS is proving to be a much more efficient method that does not create hydrogen gas: it is easily controlled, low-cost, and scalable. Applications of Al-HMDS include the creation of aluminum nitride (AlN) and silicon carbide (SiC), which are semiconductors that can be used as films or coatings for solid-state batteries. The current methods to synthesize AlN and SiC require high temperatures, are prone to impurities that decrease conductivity, and must undergo multiple processes. It is expected that the novel method can allow for commercial use of AlN and SiC more accessible than it is currently, permitting solid-state batteries to be more viable.

Synthesizing reactions between metal chlorides and hexamethyldisilazane Read More »

Using bioinformatics to analyze the effects of COVID-19 on organs

whahg

The International Statistical Classification of Diseases and Related Health Problems, 10th and 9th revisions (ICD10 and ICD9), has been adopted worldwide to share insurance codes for diseases, symptoms, and findings. An effective ICD system is especially important currently with the COVID-19 pandemic, as the international Consortium for Clinical Characterization of COVID-19 by EHR (4CE) website stores diagnosis COVID-19 disease data using ICD10/ICD9 codes. Unfortunately, the ICD system is difficult to decode due to its’ many shortcomings. Our project aims to address these shortcomings through a new ICD system which uses ontology.

Using bioinformatics to analyze the effects of COVID-19 on organs Read More »

Team Mental Models

jesszhu

Jessica Zhu Pronouns: she, her, hers Research Mentor(s): Walter Lasecki, Assistant Professor Research Mentor School/College/Department: Computer Science and Engineering, College of Engineering Presentation Date: Thursday, April 22, 2021 Session: Session 4 (2pm-2:50pm) Breakout Room: Room 2 Presenter: 5 Event Link Abstract For privacy concerns this abstract cannot be published at this time. Authors: Jessica Zhu

Team Mental Models Read More »

Safety Applications from Connected Vehicle Trajectories

Sion Pizzi

Connected Vehicle (CV) technology allows road users and the transportation infrastructure to maintain constant communication. This technology is anticipated to increase safety, enhance mobility, and curb the environmental footprint of the transportation sector. However, for these benefits to be realized, data-driven applications need to be developed. This project focuses on using CV data to develop a safety application with a focus on pedestrians–the most vulnerable road users. The goal of this study is to predict a pedestrian’s trajectory so as to intervene and prevent dangerous scenarios that pose safety risk to pedestrians . In order to develop our application, we use data gathered in the form of multi-dimensional trajectories in a variety of contexts. We take into account several kinds of time-series data such as latitude, longitude, velocity, acceleration, rotation etc. This data is then processed to be used as input to a deep learning model that can predict the future trajectory of an agent according to its history trajectory. Deep learning is selected because it allows for predicting future trajectory without making modeling assumptions. To predict the future trajectory, we define a reachable set for each sub-trajectory. Next, we develop a framework called step attention composed of multiple deep-learning models, the output of which is the trajectory of an agent in the next few seconds. The ultimate goal of this application is to alert pedestrians or vehicles of imminent high-risk situations and possibly recommend actions on how to avoid such situations altogether, or reduce the severity of an upcoming incident if it cannot be avoided.

Safety Applications from Connected Vehicle Trajectories Read More »

Testing and Formulation of Icephobic Coatings

mzahran

The buildup of ice on surfaces can be detrimental to the performance of essential infrastructures, technologies, and transportation systems such as aircrafts, ships, turbines, powerlines, and more. Icephobic coatings can act as a shield against the accumulation of ice on surfaces, providing a cost-effective way of protecting the integrity and safety of systems and reducing or eliminating potential maintenance costs created by ice damage. There are numerous combinations of reagents, solvents, and catalysts that, when reacted together, form a product exhibiting icephobic properties. An ideal icephobic coating will have an ice adhesion strength, defined as the force required to debond a specified area of ice from a substrate, of less than 100 kPa. Additionally, it is desired for these coatings to be optically clear and to have fast curing, or drying, rates. The purpose of this research is to find a formula that provides low ice adhesion strength, optimal optical properties, and a fast curing rate. Low adhesion strength and clear coatings are relatively simple to produce, but these properties in conjunction with a fast curing rate (ideally of just a few minutes) poses a challenge. Making use of polyurethane as the base polymer for icephobicity, the amounts and choices of reagents, solvents, and catalysts were changed to produce different formulas. Ice adhesion numbers were tested for all samples, and the samples were compared to one another to determine which gave the most desired results. By changing the amounts of one variable (such as solvent, catalyst, oil, or diisocyanate amount) and keeping all others constant, the affect that each variable has on the icephobic properties of the formula could be studied.

Testing and Formulation of Icephobic Coatings Read More »