10 Bits: the Data News Hotlist
This week’s list of data news highlights covers April 2-8, 2016 and includes articles about a new smartphone app that can diagnose respiratory diseases and about how data analytics helped reporters investigate the Panama Papers.
The UK Parliament will soon introduce legislation to require private bus companies to publish timetables, delays, fares, and other transportation information as open data. By making this data publicly available, local transit authorities will be able to better negotiate partnerships with private operators and commuters will be able to better plan their trips. The bill will also lay the groundwork for all of the country’s transportation systems to use a common ticketing system.
Facebook has implemented a new feature called automatic alternative text that uses machine learning algorithms to recognize and caption the contents of photos, which will allow individuals with vision impairments to use text-to-speech software to better engage with content on the site. Automatic alternative text can identify and describe objects in images, as well as people and their expressions.
Massachusetts General Hospital has partnered with microchip manufacturer Nvidia to use the company’s new DGX-1 supercomputer, designed specifically for artificial intelligence applications, to advance research into disease diagnostics and treatment. The hospital’s Clinical Data Science Center will use the supercomputer to train an artificial neural network on the hospital’s database of 10 billion medical images. Initially, the center will focus on radiology and pathology applications, which rely heavily on image analysis. Eventually, it will also use the supercomputer to analyze genetic data and electronic health records.
The National Aeronautics and Space Administration (NASA) and the Japan Aerospace Exploration Agency have taken the first 3D images of precipitation from space. Unlike other weather monitoring imaging, 3D images can indicate both the size and distribution of water particles in clouds by altitude, which can help scientists make better predictions about the intensity of rain and snowfall. The images come from the Global Precipitation Measurement (GPM) Core Observatory, a satellite launched in 2014 by the United States and Japan with specialized radar and imaging sensors to improve forecasting for extreme weather events.
Researchers at Princess Margaret Hospital and Joondalup Health Campus in Perth have developed an app that can determine if a person has a respiratory disease.The app uses the microphone on a smartphone to listen to a user’s breathing and analyzes this data with machine-learning algorithms that can identify if he or she has a respiratory disease and differentiate between specific diseases, including asthma, pneumonia, and bronchiolitis. The app has an average accuracy of 89 percent, although in certain situations, it is almost 100 percent accurate. For example, it was 97 percent accurate at diagnosing patients with lower respiratory tract disease that were initially cleared by doctors but later correctly diagnosed after clinical testing.
A convoy of self-driving trucks from multiple manufacturers have successfully traversed large portions of Europe as part of the European Truck Platooning Challenge organized by the Dutch government. When self-driving trucks can communicate with each other, they can follow each other much more closely than human drivers can in a process known as “platooning.” This can reduce congestion, lower the risk of accidents caused by human error, and reduce fuel use by up to 15 percent. For example, two trucks platooning for 100,000 miles annually can save nearly $7,000 a year in fuel costs compared to driving on cruise control.
IBM and Pfizer have partnered to pilot a system of connected sensors and devices that can provide continuous disease monitoring services for patients with Parkinson’s disease and share this data with researchers. Treating Parkinson’s disease is complicated because patients can require frequent medication adjustments depending on their unique responses to the drugs and the rate of the disease’s progression. With a host of motion-detecting sensors and monitoring technology, doctors could monitor a patient’s symptoms in real time to make better treatment decisions and researchers can analyze more granular data about treatment effectiveness.
German newspaper Süddeutsche Zeitung (SZ) and the International Consortium of Investigative Journalists (ICIJ), which led the investigation into the Panama Papers, a massive leak of documents exposing decades of shady financial activity by politicians, celebrities, and other elites, used analytics technology developed by software company Nuix to reveal the massive network of tax evasion, sanctions violations, and other criminal activity. The Panama Papers consist of 2.6 terabytes of 11.5 million documents, many in difficult-to-analyze formats such as PDFs and images. SZ and ICIJ used Nuix’s algorithms to extract metadata from the documents and convert them into machine-readable data, which allowed them to search and index the data to advance their investigation much more quickly than if they had to manually analyze each document.
A team of data scientists, engineers, and art historians have created an original piece of artwork titled “the Next Rembrandt” in the style of the famous 17th-century Dutch painter Rembrandt by using analytics and 3D printing. The team used facial recognition algorithms, geometric analysis, and other analytic techniques on Rembrandt’s works over 18 months to generate a 148 million pixel portrait that accurately mimicked Rembrandt’s distinctive style.
Researchers at Toyota are developing an autonomous system they refer to as a “guardian angel” capable of taking control of a vehicle or making adjustments to avoid danger while a human is driving. The researchers will train the system by analyzing how human drivers respond to simulated crash scenarios in a virtual environment, such as how they turn and brake, so it can understand the best ways to respond to driver behavior.