BBC "Science in Action" interview where I will talk about my latest publication on synthetic retinas
Time: 8:30 UK / 7:30 GMT
ORIGIN: BBC Two
jueves, 20 de abril de 2017
domingo, 16 de abril de 2017
With this new system, scientists never have to write a grant application again
 |
| Johan Bollen (left) and Marten Scheffer (right) say scientists should give each other money instead of writing and reviewing grants. Ingrid van de Leemput |
AMSTERDAM—Almost every scientist agrees: Applying for research funding is a drag. Writing a good proposal can take months, and the chances of getting funded are often slim. Funding agencies, meanwhile, spend more and more time and money reviewing growing stacks of applications.
That’s why two researchers are proposing a radically different system that would do away with applications and reviews; instead scientists would just give each other money. “Self-organized fund allocation” (SOFA), as it’s called, was developed by computer scientist Johan Bollen at Indiana University in Bloomington. When he first published about the idea in 2014, many people were skeptical. But interest appears to be growing, and thanks to the work of an enthusiastic advocate, ecologist Marten Scheffer of Wageningen University in the Netherlands, the Dutch parliament adopted a motion last year asking the country’s main funding agency, the Netherlands Organization for Scientific Research (NWO), to set up a SOFA pilot project.
Competition for funding has become too intense, especially for young scientists, Scheffer and Bollen say, and the current peer-review system is inefficient. It’s also unfair, they argue, because a few scientists get lots of grants—Scheffer is one of them—whereas many others get few or nothing. But when Scheffer explained his idea at an NWO workshop about “application pressure” here last week, the agency didn’t appear sold yet.
The duo says the numbers speak for themselves. At the U.S. National Institutes of Health, the overall success rate for grants applications has dropped from 30% in 2003 to 19.1% in 2016. In the latest round of European Research Council Starting Grants, the rate was a paltry 11.3%. At NWO, the success rate for grants for young scientists has dropped to 14%. A 2013 study estimated that writing and reviewing applications for €40 million worth of these grants costs €9.5 million annually.
In Bollen’s system, scientists no longer have to apply; instead, they all receive an equal share of the funding budget annually—some €30,000 in the Netherlands, and $100,000 in the United States—but they have to donate a fixed percentage to other scientists whose work they respect and find important. “Our system is not based on committees’ judgments, but on the wisdom of the crowd,” Scheffer told the meeting.
Bollen and his colleagues have tested their idea in computer simulations. If scientists allocated 50% of their money to colleagues they cite in their papers, research funds would roughly be distributed the way funding agencies currently do, they showed in a paper last year—but at much lower overhead costs.
Not everybody is convinced. At the meeting, some worried that scientists might give money mostly to their friends. Scheffer said an algorithm would prevent that, for instance by banning donations to people you have published with, but he acknowledged it would be a challenge in small research communities. SOFA might also result in a mismatch between what scientists need and what their colleagues donate, and a competition for donations could lead to a time-consuming and costly circus, comparable to an election campaign.
The way to find out, Scheffer and Bollen say, is a real-world test, and they say the Netherlands, a small country with short lines of communication between scientists, politicians, and funding agencies, is a good place for one. Last year, Scheffer convinced Eppo Bruins, a member of the Dutch House of Representatives, to submit a motion calling for a pilot program at NWO, which the parliament approved in June 2016. The money could be taken from a €150 million NWO pot currently distributed among consortia of innovative Dutch scientists, Bruins suggested.
But NWO is not obliged to carry out the proposal, and so far has shown little enthusiasm. “NWO is willing to explore together with scientists and other stakeholders how to improve allocation rates, but is still considering practicality and support” for SOFA, a spokesperson tells ScienceInsider. At last week’s meeting, NWO President Stan Gielen said the funds Bruins has in mind are distributed by NWO but are earmarked by the Ministry of Education, Culture and Science, which would have to give permission. Gielen added that any experiment should not come at the expense of existing funding.
Scheffer says he’s not giving up. It’s not a risky experiment, he says: “The money would not be wasted, after all, but just be given to other scientists.” But he says he understands why NWO is not thrilled: If applied universally, the novel system would make the agency redundant. Perhaps it’s telling, Scheffer says, that he has not been invited to an international conference on applications and peer review that NWO is organizing in June.
DOI: 10.1126/science.aal1055
ORIGINAL: Science Magazine
Apr. 13, 2017 , 3:00 PM
sábado, 15 de abril de 2017
Spectacular Visualizations of Brain Scans Enhanced with 1,750 Pieces of Gold Leaf
 |
| Self Reflected, 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The entire Self Reflected microetching under violet and white light. (photo by Greg Dunn and Will Drinker) |
Anyone who thinks that scientists can't be artists need look no further than Dr. Greg Dunn and Dr. Brian Edwards. The neuroscientist and applied physicist have paired together to create an artistic series of images that the artists describe as “the most fundamental self-portrait ever created.” Literally going inside, the pair has blown up a thin slice of the brain 22 times in a series called Self-Reflected.
Traveling across 500,000 neurons, the images took two years to complete, as Dunn and Edwards developed special technology for the project. Using a technique they've called reflective microetching, they microscopically manipulated the reflectivity of the brain's surface. Different regions of the brain were hand painted and digitized, later using a computer program created by Edwards to show the complex choreography our mind undergoes as it processes information.
After printing the designs onto transparencies, the duo added 1,750 gold leaf sheets to increase the art's reflectivity. The astounding results are images that demonstrate the delicate flow and balance of our brain's activity. “Self Reflected was created to remind us that the most marvelous machine in the known universe is at the core of our being and is the root of our shared humanity,” the artists share.
Self Reflected fine art prints and microetchings are available for purchase via Dunn's website.
Self Reflected is an unprecedented look inside the brain.
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The parietal gyrus where movement and vision are integrated. (photo by Greg Dunn and Will Drinker) |
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The brainstem and cerebellum, regions that control basic body and motor functions. (photo by Greg Dunn and Will Drinker) |
An astounding achievement in scientific art, the artists applied 1,750 leaves of gold to the final microetchings.
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The laminar structure of the cerebellum, a region involved in movement and proprioception (calculating where your body is in space). |
 |
Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The pons, a region involved in movement and implicated in consciousness. (photo by Greg Dunn and Will Drinker)
|
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. Raw colorized microetching data from the reticular formation. |
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The visual cortex, the region located at the back of the brain that processes visual information. |
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The thalamus and basal ganglia, sorting senses, initiating movement, and making decisions. (photo by Greg Dunn and Will Drinker) |
 |
| Self Reflected, 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The entire Self Reflected microetching under white light. (photo by Greg Dunn and Will Drinker) |
 |
| Self Reflected (detail), 22K gilded microetching, 96″ X 130″, 2014-2016, Greg Dunn and Brian Edwards. The midbrain, an area that carries out diverse functions in reward, eye movement, hearing, attention, and movement. (photo by Greg Dunn and Will Drinker) |
This video shows how the etched neurons twinkle as a light source is moved.
ORIGINAL: My MET
By Jessica Stewart
April 12, 2017
viernes, 7 de abril de 2017
Scientists Have Created an Artificial Synapse That Can Learn Autonomously
 |
| Sergey Tarasov/Shutterstock |
Developments and advances in artificial intelligence (AI) have been due in large part to technologies that mimic how the human brain works. In the world of information technology, such AI systems are called neural networks.
These contain algorithms that can be trained, among other things, to imitate how the brain recognises speech and images. However, running an Artificial Neural Network consumes a lot of time and energy.
Now, researchers from the National Centre for Scientific Research (CNRS) in Thales, the University of Bordeaux in Paris-Sud, and Evry have developed an artificial synapse called a memristor directly on a chip.
It paves the way for intelligent systems that required less time and energy to learn, and it can learn autonomously.
In the human brain, synapses work as connections between neurons. The connections are reinforced and learning is improved the more these synapses are stimulated.
The memristor works in a similar fashion. It's made up of a thin ferroelectric layer (which can be spontaneously polarised) that is enclosed between two electrodes.
Using voltage pulses, their resistance can be adjusted, like biological neurons. The synaptic connection will be strong when resistance is low, and vice-versa.
 |
(a) Sketch of pre- and post-neurons connected by a synapse. The synaptic transmission is modulated by the causality (Δt) of neuron spikes. (b) Sketch of the ferroelectric memristor where a ferroelectric tunnel barrier of BiFeO3 (BFO) is sandwiched between a bottom electrode of (Ca,Ce)MnO3 (CCMO) and a top submicron pillar of Pt/Co. YAO stands for YAlO3. (c) Single-pulse hysteresis loop of the ferroelectric memristor displaying clear voltage thresholds ( and  ). (d) Measurements of STDP in the ferroelectric memristor. Modulation of the device conductance (ΔG) as a function of the delay (Δt) between pre- and post-synaptic spikes. Seven data sets were collected on the same device showing the reproducibility of the effect. The total length of each pre- and post-synaptic spike is 600 ns. Source: Nature Communications |
The memristor's capacity for learning is based on this adjustable resistance.
AI systems have developed considerably in the past couple of years. Neural networks built with learning algorithms are now capable of performing tasks which synthetic systems previously could not do.
For instance, intelligent systems can now compose music, play games and beat human players, or do your taxes. Some can even identify suicidal behaviour, or differentiate between what is lawful and what isn't.
This is all thanks to AI's capacity to learn, the only limitation of which is the amount of time and effort it takes to consume the data that serve as its springboard.
With the memristor, this learning process can be greatly improved. Work continues on the memristor, particularly on exploring ways to optimise its function.
For starters, the researchers have successfully built a physical model to help predict how it functions.
Their work is published in the journal Nature Communications.
ORIGINAL: ScienceAlert
DOM GALEON, FUTURISM
7 APR 2017
Suscribirse a:
Entradas (Atom)