Welcome to the June 23, 2021 edition of ACM TechNews, providing timely information for IT professionals three times a week.
ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
|
Algorithm That Predicts Deadly Infections Is Often Flawed
Wired
Tom Simonite
June 21, 2021
An algorithm designed by U.S. electronic health record provider Epic Systems to forecast sepsis infections is significantly lacking in accuracy, according to an analysis of data on about 30,000 patients in University of Michigan (U-M) hospitals. U-M researchers said the program overlooked two-thirds of the approximately 2,500 sepsis cases in the data, rarely detected cases missed by medical staff, and was prone to false alarms. The researchers said Epic tells customers its sepsis alert system can correctly differentiate two patients with and without sepsis with at least 76% accuracy, but they determined it was only 63% accurate. U-M's Karandeep Singh said the study highlights wider shortcomings with proprietary algorithms increasingly used in healthcare, noting that the lack of published science on these models is "shocking."
|
Fusion of AI, Nanopore Technology: Passing the COVID Test in Just 5 Minutes
SciTechDaily
June 21, 2021
A SARS-CoV-2 virus test developed by scientists at Japan's Osaka University combines artificial intelligence and nanopore technology to differentiate similarly-sized coronaviruses that underpin different respiratory ailments. The platform demonstrated 90% sensitivity and 96% specificity for SARS-CoV-2 detection from clinical saliva samples in just five minutes. The researchers bored 300-nanometer-diameter nanopores into a silicon nitride membrane; partial blockage of the opening when a virus was pulled through a nanopore by electrophoretic force caused a temporary ionic flow reduction inside the nanopore, which registered as a change in electrical current. Current as a function of time yielded data on volume, structure, and surface charge of the analyzed target, and the algorithm was trained on 40 polymerase chain reaction (PCR)-positive and 40 PCR-negative saliva samples. Said Osaka's Masateru Taniguchi, "We expect that this research will enable rapid point-of-care and screening tests for SARS-CoV-2 without the need for RNA extraction."
|
Stanford Researchers Develop Software for Designing Sustainable Cities
Stanford University
Sarah Cafasso
June 21, 2021
Researchers in Stanford University's Natural Capital Project have developed software to facilitate the design of sustainable cities. The free, open source Urban InVEST creates maps for city planners and developers to visualize where investments in nature could benefit human well-being. Urban InVEST software has been used to determine how natural infrastructure could reduce damages from a once-in-100-years storm in Shenzhen, China; to identify neighborhoods with the least access to green space in Paris, France; and to decide how to repurpose underused golf course land in Minneapolis, MN. Stanford's Perrine Hamel said, "We're answering three crucial questions with this software: where in a city is nature providing what benefits to people, how much of each benefit is it providing, and who is receiving those benefits?"
|
EU Data Protection Authorities Call for Ban on Facial Recognition
Politico Europe
Melissa Heikkila
June 21, 2021
The European Data Protection Supervisor (EDPS) and the European Data Protection Board (EDPB) have urged a ban on the use of artificial intelligence (AI)-driven facial recognition technology in public places. The European Commission's AI bill limits its use in public places by law enforcement, without prohibiting it outright. In a joint statement, EDPB chair Andrea Jeline and EDPS Wojciech Wiewiorowski said, "A general ban on the use of facial recognition in publicly accessible areas is the necessary starting point if we want to preserve our freedoms and create a human-centric legal framework for AI." They also urged a ban on logging gait, fingerprints, DNA, voice, keystrokes, and other biometric data, as well as on AI systems that biometrically distinguish ethnicity, gender, and political or sexual orientation.
|
Tesla Backs Vision-Only Approach to Autonomy Using Supercomputer
Tech Crunch
Rebecca Bellan
June 21, 2021
Tesla's Andrej Karpathy unveiled a new supercomputer at the 2021 Conference on Computer Vision and Pattern Recognition, a prototype of what the automaker's neural network training computer, Dojo, ultimately will become. Tesla CEO Elon Musk has said Dojo will make vision-only autonomous driving a reality. With this new supercomputer, Tesla can replace radar and LiDAR sensors on self-driving cars with high-quality optical cameras. It boasts 10 petabytes of "hot tier" NVME storage and runs at 1.6 terabytes per second. Karpathy said that with 1.8 EFLOPS, Dojo could be the fifth most-powerful supercomputer in the world. Tesla used the supercomputer to gather about 1 million videos of about 10 seconds each and label 6 billion objects within those videos with regard to depth, velocity, and acceleration.
|
Digitizing Rural Land Records, 1 Drone at a Time
Bloomberg Quint
Pallavi Nahata
June 21, 2021
The Indian government's Svamitva (Survey of Villages and Mapping with Improvised Technology in Village Areas) project involves the use of drone technology to survey the inhabited areas of rural villages. The project aims to give villagers a "record of rights" that could be used as an asset and to handle property disputes. A pilot project in nine states was rolled out nationwide in April 2021, and aims to cover about 662,000 villages by 2024. The process involves marking property boundaries with limestone powder and using drones to collect high-resolution aerial images. The images are sent to the Survey of India to be transformed into maps, and villagers are given 15 days to verify the accuracy of their maps. Almost 46,000 villages had been surveyed as of early May.
|
Quantum Data Link Established Between 2 Distant Chinese Cities
New Scientist
Matthew Sparkes
June 21, 2021
Researchers at the University of Science and Technology of China have created a secure quantum link extending 511 kilometers (almost 320 miles) between two Chinese cities. The researchers strung a fiber-optic connection between Jinan and Qingdao, with a central receiver located between the two cities in Mazhan. Lasers at both ends of the cable send photons toward each other. The relay in the middle does not read the data, checking only whether the two signals matched. The researchers found the two ends could exchange a quantum key that could be used to encrypt data sent over traditional networks. University of Sussex's Peter Kruger said, "Single photons over hundreds of kilometers is quite remarkable."
*May Require Paid Registration
|
Researchers Create Brain Interface That Can Sing What a Bird's Thinking
The Next Web
Tristan Greene
June 18, 2021
A machine learning system developed by University of California San Diego researchers can predict what a bird is about to sing, a step on the road to training systems to translate human speech in real time. The researchers implanted electrodes in the brains of a dozen zebra finches and recorded neural activity as the birds sang. They trained another system to reduce the birds’ real-time songs to recognizable patterns. The researchers said, "Birdsong shares a number of unique similarities with human speech, and its study has yielded general insight into multiple mechanisms and circuits behind learning, execution, and maintenance of vocal motor skill."
|
Israeli Researchers Develop Electronic Nose to Detect Diseases, Poisons
The Jerusalem Post
June 19, 2021
An artificial nose developed by researchers at Israel's Ben-Gurion University of the Negev (BGU) can distinguish between different types of bacteria, viruses, and poisonous gases based on their "smell print." This smell print is produced by the absorption of gases using carbon nanoparticles and the electrical reaction caused by the particles as a result of the absorption. The researchers said they "were able to 'train' the electronic nose using machine learning techniques to detect different gas molecules, individually or in a mixture, with high accuracy." BGU's Raz Yelink said the low-cost technology could be used to warn cities about the presence of dangerous gases and air pollution, detect bacterial infections within an hour via a "throat swab" test, and warn of the presence of bacteria in food products.
|
Data Breaches: Most Victims Unaware When Shown Evidence of Multiple Compromised Accounts
University of Michigan News
June 21, 2021
Most data breach victims do not realize their personal data has been exposed in five breaches on average, according to a study by an international team of researchers. Investigators at the University of Michigan, George Washington University (GW), and Germany's Karlsruhe Institute of Technology showed 413 participants facts from up to three breaches involving their personal information, and 74% said they were unaware of the breaches. Most victims blamed their personal habits for the problem—like using the same password for multiple accounts—while just 14% blamed it on external factors. GW's Adam Aviv said, "The fault for breaches almost always lies with insufficient security practices by the affected company, not by the victims of the breach."
|
Microscopy Deep Learning Predicts Viral Infections
University of Zurich (Switzerland)
June 21, 2021
A deep learning artificial neural network can reliably identify cells infected by adenoviruses or herpes viruses from fluorescent images of live cells. The algorithm, developed by researchers at Switzerland's University of Zurich (UZH), is designed to detect changes in the arrangement of cellular nuclei that signal the presence of infection. The researchers trained the model to identify patterns typical of infected or uninfected cells using a large dataset of microscopy images. The algorithm also demonstrated the ability to identify acute and severe infections with 95% accuracy, up to 24 hours in advance, by distinguishing between two separate phases of lytic infections.
|
Algorithm Shows the Alcohol Burden on Ambulance Service in Scotland
University of Glasgow (U.K.)
June 21, 2021
Researchers at the U.K.'s University of Glasgow led a team that developed an algorithm to determine the number of alcohol-related ambulance callouts by the Scottish Ambulance Service (SAS). The study, the first to quantify the burden of alcohol on SAS, used 2019 SAS data to determine that 16.2% of ambulance callouts were alcohol-related, more than three times higher than reported previously. The algorithm searches paramedic notes in patient records for references to alcohol with 99% accuracy compared to reviews of records by an experienced paramedic. The researchers estimated the total cost of these callouts at about £31.5 million (US$44 million). SAS medical director Dr. Jim Ward said the study helps his organization “to better understand the impact alcohol has on the demand for ambulance response."
|
|
