This week’s list of top data news highlights covers April 9, 2022 to April 15, 2022 and includes articles on using virtual reality to provide remote medical care to Ukrainian refugees and detecting interstellar meteors with sensor technology.
1. Predicting Sudden Cardiac Arrests
Researchers at Johns Hopkins University have created an AI system that can predict a patient’s chance of experiencing sudden cardiac arrest within the next ten years. The team trained the system to detect scar patterns in cardiac images, improving upon current analysis methods that use only scar features like volume and mass. They created the system with images of cardiac scarring and demographic data from hundreds of patients.
2. Preventing COVID-19 Infection
Researchers at Northwestern University, the University of Washington, and Washington University at St. Louis have used a supercomputer to create a nasal spray that can protect individuals from COVID-19 variants. The spray uses protein binders to block the virus’s binding tools and prevent further infection. The spray recently advanced to Phase 1 human clinical trials.
Researchers at Stanford University have created an AI system that can better detect earthquakes by removing unrelated background noise. The team trained the system on 80,000 samples of urban seismic noise and 33,751 samples of earthquake seismic data. The team tested the system with data from an earthquake in La Habra, California and recorded four times more seismic detections than previously observed.
The Houston Astros have partnered with Amazon to use Just Walk Out technology in their stadium. The technology uses sensors, cameras, and an AI system to enable customers to purchase snacks or souvenirs without visiting a cashier. The Astros are the first Major League Baseball team to install Just Walk Out technology in their park.
Researchers at the Technology Innovation Institute, a government-funded technology research organization based in the United Arab Emirates, have created a natural language processing model for Arabic. The model is the biggest of its kind and was trained on 10 billion parameters, including news articles, poetry, and web data.
6. Modeling Prehistoric Human Evolution
Researchers at Pusan National University in South Korea have used a supercomputer to model the effect of prehistoric weather patterns on human evolution. The team created the model with synthetic data on paleoclimate records such as Antarctic temperatures and summer monsoons, and combined the data with fossil and archaeological records. They found that evolutionary responses to hotter and drier weather conditions in South Africa 700,000 years ago likely led to the rise of Homo sapiens.
Alkadur RobotSystems, a German robotic technology company, has created a robot that can autonomously slice kebabs and el trompo de pastor from meat on a spinning top. The robot contains sensors that can measure the amount of meat necessary for a specific meal and an AI system that instructs the robot to slice the meat.
Clinicians with Kohav Meir, an Israeli field hospital in western Ukraine, have partnered with Israeli telehealth companies Tytocare and Pulsenmore to provide medical care to Ukrainian refugees. The team uses virtual reality and sensor technology to connect patients with doctors in Israel and perform remote medical examinations.
9. Measuring Players’ Performances
Researchers at the Pennsylvania State University have created a machine learning model that can determine the impact of baseball players’ performances during games. The team trained the model with data on player records, season results, previously collected pitch data, including player positioning, base occupancy, and pitch velocity, and potential sequences of events that can occur when a pitch is thrown. The system uses this data and sequences of game plays to create numerical representations of players’ performances.
10. Finding Earth’s First Interstellar Meteor
The U.S. Space Command has confirmed that a meteor that traveled through Earth’s atmosphere in 2014 originated outside of our solar system. The Space Command used sensor technology to detect the meteor and estimate its velocity, and confirmed previous studies classifying the meteor as interstellar. The meteor is the first recorded interstellar object to enter Earth’s atmosphere.
Image credit: Flickr user Abbeenormal – Abbee Day