This Brain-Inspired Chip Is A Hit
Top image credit: The new chip “makes associations between what it has experienced and what it experiences." Image credit: imec.
What is it? Researchers working at imec, a Belgium-based innovation hub focusing on nanoelectronics and digital technologies, have developed a “brain-inspired,” or neuromorphic, self-learning chip that composes music.
How does it work? Imec said the chip “makes associations between what it has experienced and what it experiences. The more it experiences, the stronger the connections will be. The chip presented today has learned to compose new music and the rules for the composition are learnt on the fly.”
Who cares? The brain uses only a tiny bit of energy to perform powerful computations, and neuromorphic chips are seeking the same Holy Grail. This meshes well with what the hub’s ultimate mission: “to design the process technology and building blocks to make artificial intelligence to be energy efficient so that that it can be integrated into sensors.” It says these intelligent sensors will fuel the internet of things’ evolution. “This would not only allow machine learning to be present in all sensors but also allow on-field learning capability to further improve the learning.”
What is it? Scientists at the Weill Cornell Medical College in New York found a way to make “an unlimited supply of healthy blood cells” in the lab from “readily available cells that line blood vessels.”
What’s the big deal? The breakthrough “might allow us to provide healthy stems cells to patients who need bone marrow donors but have no genetic match,” said Weill Cornell’s Joseph Scandura, co-senior author of the paper, which was published in the journal Nature. “It could lead to new ways to cure leukemia and myeloproliferative neoplasms (a type of blood cancer), and may help us correct genetic defects that cause blood diseases like sickle cell anemia.”
How did they do it? The team was able “efficiently convert” cells that line blood vessels into “abundant, fully functioning” hematopoietic stem cells (HSCs). These stems cells can turn into red and white blood cells and platelets as well as generate more HCSs. HSCs help the body replenish mature blood cells in the blood stream after they die. “This property allows just a few thousand HSCs to produce all of the blood cells a person has throughout his or her life,” Weill Cornell said in a press release.
Routine Blood Test Could Unmask Cancer
What is it? Researchers at the University of Exeter in England have found that a condition called thrombocytosis that indicates high platelet count in the blood was “a strong predictor of cancer.”
Why does matter? Thrombocytosis can be diagnosed with a routine blood test. “Our research suggests that substantial numbers of people could have their cancer diagnosed up to three months earlier if thrombocytosis prompted investigation for cancer,” wrote Exeter’s Sarah Bailey, lead author of the study. “This time could make a vital difference in achieving earlier diagnosis.” The university reported that “lung and colorectal cancer were more commonly diagnosed with thrombocytosis. One-third of patients with thrombocytosis and lung or colorectal cancer had no other symptoms that would indicate to their [general practitioner] that they had cancer.”
How did they do it? Bailey and her colleagues looked at 40,000 patient records in the U.K. They found that “more than 11 percent of men and 6 percent of women over the age of 40 with thrombocytosis went on to be diagnosed with cancer within a year. This rose to 18 percent of men and 10 percent of women if a second raised platelet count was recorded within six months.” The results were published in the British Journal of General Practice.
A Robot Inspired By A Carnivorous Plant
What is it? Researchers at Tampere University of Technology in Finland developed an “optical gripper” that resembles the jaws of the carnivorous plant Venus flytrap.
Why does it matter? The team calls their light-controlled, autonomous soft robot the first of its kind that can recognize objects. “The Venus flytrap plant stands still with its leaves wide open, waiting for an insect to land there,” says Arri Priimägi, leader of the university’s Smart Photonic Materials group. “It knows whether the thing sitting on its surface is an insect or a speck of dust, and only closes when it is an insect. That’s what we are trying to do with our ‘flytrap.’ ” He says that “the artificial flytrap could potentially be applied in ‘quality control’, i.e., it could automatically identify and pick out small components with defects from a production line.”
How does it work? The researchers built the soft robot, which is less than 1 centimeter long, from a pliant plastic strip that can be shaped with light. “When an object in the field of view of the flytrap reflects light onto the elastomer surface, the strip bends itself around the object, capturing it like a flytrap,” according to the university. “This allows different objects to be recognized based on their light reflection, and to move them in a controlled way. The flytrap’s lifting power is hundreds of times its body weight.” The research appeared in the journal Nature Communications.
Making Wi-Fi See Through Walls
What is it? Engineers at the Technical University of Munich have found a way to process signals from an ordinary Wi-Fi hotspot to image its surroundings. “They found that the technique could potentially allow users to peer through walls and could provide images 10 times per second,” according to the journal APS Physics.
How did they do it? The team reported that Wi-Fi signals traveling through a room form a hologram — “two-dimensional wave front encoding a three-dimensional view of all objects traversed by the light beam.” They recovered 3D images by “feeding the resulting data into digital reconstruction algorithms.”
Why does it matter? Wi-Fi imaging has been around for a while, but researchers struggled to assemble the signal into a coherent image. Privacy issues aside, the Munich team wrote that one day the technology “may aid in the recovery of victims buried under an avalanche or a collapsed building. While conventional methods only allow point localization of victims, holographic signal processing could provide a spatial representation of destroyed structures, allowing first responders to navigate around heavy objects and use cavities in the rubble to systematically elucidate the easiest approach to quickly reach victims.”