Archives: News

news custom post type.

How to identify hit songs with 97% accuracy, using machine learning

Every day, tens of thousands of songs are released. This constant stream of options makes it difficult for streaming services and radio stations to choose which songs to add to playlists, says Paul Zak, a professor at Claremont Graduate University.

Zak is senior author of a study published in Frontiers in Artificial Intelligence, where he offers a solution: “neuroforecasting”—applying machine learning to neurophysiologic data.

Neural data

In the study, participants listened to a set of 24 songs, and were asked about their preferences and some demographic data.

During the experiment, the scientists measured participants’ neurophysiologic responses to the song. A wearable watch provided heart rate data, which was used to infer neural states from activity of the cranial nerves (based on downstream effects of dopamine and oxytocin), as noted in JeŽová et al., 1985; Zak, 2012; Barraza et al., 2015. 

Machine learning

“The brain signals we’ve collected reflect activity of a brain network associated with mood and energy levels,” Zak said. This allowed the researchers to predict market outcomes, including the number of streams of a song.

The researchers found that a linear statistical model identified hit songs at a success rate of 69%. But when they applied machine learning to the data they collected, the rate of correctly identified hit songs jumped to 97%.

“That the neural activity of 33 people can predict if millions of others listened to new songs is quite amazing. Nothing close to this accuracy has ever been shown before,” Zak said.

“If in the future, wearable neuroscience technologies, like the ones we used for this study, become commonplace, the right entertainment could be sent to audiences based on their neurophysiology, ” Zak said.

But the researchers pointed to some limitations. For example, they used relatively few songs in their analysis. And the demographics of the study participants did not include members of “certain ethnic and age groups.”

Citation: Merritt, S. H., Gaffuri, K., & Zak, P. J. (2023). Accurately predicting hit songs using neurophysiology and machine learning. Frontiers in Artificial Intelligence, 6, 1154663. https://doi.org/10.3389/frai.2023.1154663 (open-access)

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

Extreme DNA precision: Researchers slow down and scan individual DNA molecules multiple times

École polytechnique fédérale de Lausanne (EPFL) researchers have achieved near-perfect control over the manipulation of individual molecules, allowing them to be identified and characterized with unprecedented precision

Background: A nanopore is a nanometer-sized hole formed in a synthetic membrane. It can be used for experimental direct sequencing of a single DNA molecule. As the DNA molecule passes through the nanopore, the passage causes changes in the ion current. Those changes can be used to determine the desired sequence of nucleotides (which encode genetic information) by analyzing how each nucleotide perturbs this current as it passes through. 

However, the passage of molecules through a nanopore and the timing of their analysis are influenced by random physical forces, making it difficult to achieve high analytical accuracy.

Advanced sensing precision

So Aleksandra Radenovic, head of the EPFL Laboratory of Nanoscale Biology in the School of Engineering, has “combined the sensitivity of nanopores with the precision of a scanning ion conductance microscopy (SICM) device.”

This innovation allows for controlling molecule transit speed through the nanopore, allowing thousands of consecutive readings to be taken of the same molecule, and even of different locations on the molecule, she noted.

It can also average multiple readings of the same molecule, which has resulted in an increase in signal-to-noise ratio of two orders of magnitude, compared to conventional methods, the scientists report.

Opens up use in peptide sequencing

“This exquisite control could help fill a big gap in the field,” said Radenovic. “This precision and versatility also mean that the approach could be applied to molecules beyond DNA, such as protein building blocks called peptides, which could help advance proteomics as well as biomedical and clinical research.

“Finding a solution for sequencing peptides has been a significant challenge due to the complexity of their ‘license plates,’ which are made up of 20 characters (amino acids) as opposed to DNA’s four nucleotides,” says Radenovic.”For me, the most exciting hope is that this new control might open an easier path ahead to peptide sequencing.”

Citation: Leitao, S. M., Navikas, V., Miljkovic, H., Drake, B., Marion, S., Pistoletti Blanchet, G., Chen, K., Mayer, S. F., Keyser, U. F., Kuhn, A., Fantner, G. E., & Radenovic, A. (2023). Spatially multiplexed single-molecule translocations through a nanopore at controlled speeds. Nature Nanotechnology, 1-7. https://doi.org/10.1038/s41565-023-01412-4

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

Small neuromorphic device ‘sees,’ makes ultra-fast decisions and creates memories

Researchers in Australia have created a small neuromorphic (brain-like) device that “sees” and stores memories in a similar way to human brains. It’s a step towards future devices that can make rapid, complex decisions — in self-driving cars, for example.

This invention is based on a single chip enabled by a sensing “doped indium oxide” element, which is thousands of times thinner than a human hair and requires no external parts to operate.

Rapid decision-making

The prototype device captures visual information, pre-packages and transmits it (like an optical nerve) for storage, and classifies it, using a memory system similar to a human brain’s.

Collectively, these functions could enable ultra-fast live decision making, the team says.

RMIT University engineers in Australia led the work, with contributions from researchers at Deakin University and the University of Melbourne.

Team leader Professor Sumeet Walia, from RMIT’s School of Engineering, said the new device can perform all necessary functions: sensing, creating and processing information, and retaining memories, rather than relying on external energy-intensive computation, which prevents real-time decision-making.

“We’ve made real-time decision-making a possibility with our invention. It doesn’t need to process large amounts of irrelevant data and it’s not being slowed down by data transfer to separate processors,” said Walia.

Their findings and analysis are published in the journal Advanced Functional Materials.

Potential applications

The team’s device mimicks the retina’s capabilities to identify objects, colors and other visual features and to store and process visual information, Walia explained.

The team used ultraviolet light as part of their experiments, and is are now working to expand this technology to visible and infrared light, with many possible applications, such as autonomous operations in dangerous environments, shelf-life assessments of food and advanced forensics.

“Imagine a self-driving car that can see and recognize objects on the road in the same way that a human driver can. Or being able to able to rapidly detect and track space junk.”

Walia said neuromorphic systems could also adapt to new situations over time, becoming more efficient with more experience. “Traditional computer vision systems are typically programmed with specific rules and can’t adapt as easily,” he said.

“Neuromorphic robots have the potential to run autonomously for long periods, in dangerous situations where workers are exposed to possible cave-ins, explosions and toxic air.”

Citation: Mazumder, A., Nguyen, C. K., Aung, T., Low, M. X., Rahman, M. A., Russo, S. P., Tawfik, S. A., Wang, S., Bullock, J., Krishnamurthi, V., Syed, N., Ranjan, A., Zavabeti, A., Abidi, I. H., Guo, X., Li, Y., Ahmed, T., Daeneke, T., Al-Hourani, A., . . . Walia, S. Long Duration Persistent Photocurrent in 3 nm Thin Doped Indium Oxide for Integrated Light Sensing and In-Sensor Neuromorphic Computation. Advanced Functional Materials, 2303641. https://doi.org/10.1002/adfm.20

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

First human ‘synthetic embryos’ created

Magdalena Zernicka-Goetz, Professor of Development and Stem Cells at the University of Cambridge, announced June 14 at the International Society for Stem Cell Research annual meeting in Boston that her team had grown the first human “synthetic embryos” (made from stem cells) *

She also allowed them to grow to a stage equivalent to just past 14 days old (an internationally recognized ethical limit called the “14-day rule,” based on ethical issues regarding the possibility of growing into a human fetus). “In real embryos this [is equivalent to] a stage between day 7/8 and day 14,” notes Żernicka-Goetz.

Moral quandaries

“Just as there are real possibilities for gaining knowledge from synthetic human-like embryos, there are also real moral quandaries,” says Kathryn MacKay, Senior Lecturer in Bioethics, University of Sydney, writing in The Conversation.

“One of these quandaries arises around whether their creation really gets us away from the use of human embryos.

“Robin Lovell-Badge, the head of stem cell biology and developmental genetics at the Francis Crick Institute in London UK, reportedly said that if these human-like embryos can really model human development in the early stages of pregnancy, then we will not have to use human embryos for research.

“At the moment, it is unclear if this is the case for two reasons.

“First, the embryos were created from human embryonic stem cells, so it seems they do still need human embryos for their creation. Perhaps more light will be shed on this when Żernicka-Goetz’s research is published.

“Second, there are questions about the extent to which these human-like embryos really can model human development.

“At the moment, animal models of similar synthetic embryos suggest they are not capable of developing into a full living being. Studies in mice and monkeys have so far shown that the synthetic embryos die a short while after being implanted into a female’s womb, which means they are not viable.

“There could be significant limits to the usefulness of these synthetic embryos for learning about human developmental issues, if human-like synthetic embryos aren’t capable of developing into full human babies and do not form important body structures like a beating heart and a brain.

“One of the reasons researchers want to use these embryos is for research into miscarriage and developmental anomalies. This is very important, but will these synthetic embryos be “close enough” to real human embryos to reveal useful answers?

“Scientists may still rely on the use of human embryos if we do need human embryos for the creation of these models, or there are research questions that these synthetic embryos can’t address.

*What are synthetic embryos, exactly?

“The term is somewhat misleading as these structures aren’t really synthetic, nor are they exactly the same as embryos,” observes Clare Wilson in New Scientist. “They are similar to early embryos, a tiny ball of cells arising from a sperm fertilising an egg, but created from stem cells grown in the lab.”

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

AI algorithms find three drugs that could combat aging

Three “senolytics”—compounds in drugs that could help stave off the effects of aging (such as cancer, type-2 diabetes, osteoarthritis and viral infection)—have been discovered by researchers at the University of Edinburgh.

These algorithms were trained on already-published data, making them cost-effective, the researchers note.

Lab tests in human cells revealed that three of the compounds—ginkgetin, periplocin and oleandri—were able to remove senescent* cells without damaging healthy cells.

All three are natural products found in traditional herbal medicines, the team says. Oleandri was found to be more effective than the best-performing known senolytic drug of its kind.

*Cellular senescence is “a stress response involved in aging and diverse disease processes, including cancer, type-2 diabetes, osteoarthritis and viral infection,” the researchers explain.

Citation: Quintanilla, A., Elliott, R. J., Dawson, J. C., Sun, J., Campa, V. M., Carragher, N. O., Acosta, J. C., & Oyarzún, D. A. (2023). Discovery of senolytics using machine learning. Nature Communications, 14(1), 1-15. https://doi.org/10.1038/s41467-023-39120-1 (open-access)

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

Taurine may be a key to longer and healthier life

A deficiency of taurine—a nutrient produced in the body and found in many foods—is a driver of aging in animals. That’s a key finding of a new study led by Columbia University researchers, involving dozens of aging researchers around the world. It was published June 8 in the journal Science.

“This study suggests that taurine could be an elixir of life within us that helps us live longer and healthier lives,” says the study’s leader, Vijay Yadav, PhD, assistant professor of genetics & development at Columbia University Vagelos College of Physicians and Surgeons.

In mice and monkeys, supplementing with taurine increased lifespan

Experts found that at age 2 (60 in human years), mice supplemented with taurine for one year were healthier than untreated mice in almost every way.

They saw similar health effects in middle-aged rhesus monkeys, which were given daily taurine supplements for six months. Taurine prevented weight gain, reduced fasting blood glucose and markers of liver damage, increased bone density in the spine and legs, and improved the health of their immune systems.

Promising anti-aging strategy for humans

The researchers don’t know yet if taurine supplements will improve health or increase longevity in humans, but two experiments they conducted suggest taurine has potential. “Taurine abundance goes down with age, so restoring taurine to a youthful level in old age may be a promising anti-aging strategy.”

Other potential anti-aging drugs—including metformin, rapamycin, and NAD analogs—are also being considered for testing in clinical trials.

Foods high in taurine

According to WebMD, turkey (dark meat) has the highest taurine content of any animal meat. Other foods with high taurine include tuna (dark meat), tilapia fish (dark meat), octopus, chicken (dark meat), seaweed, and beef. Sources for this data include the American Heart Association, Harvard Medical School, and Frontiers in Physiology.

Citation: Singh, P., Gollapalli, K., Mangiola, S., Schranner, D., Yusuf, M. A., Chamoli, M., Shi, S. L., Bastos, B. L., Nair, T., Riermeier, A., Vayndorf, E. M., Wu, J. Z., Nilakhe, A., Nguyen, C. Q., Muir, M., Kiflezghi, M. G., Foulger, A., Junker, A., Devine, J., Yadav, V. K. (2023). Taurine deficiency as a driver of aging. Science. https://www.science.org/doi/10.1126/science.abn9257

What are your thoughts on taurine as a key to a longer and healthier life? Other ideas? Please comment below!

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

Chatbot designs robot, raises questions

Researchers from TU Delft and EPFL claim they’re the first to co-design a robot with ChapGPT.

“We wanted ChatGPT to design not just a robot, but one that is actually useful,” says TU Delft assistant professor Cosimo Della Santina. So with PhD student Francesco Stella and Josie Hughes from EPFL, and following ChatGPT’s advice, the team chose food supply as their challenge. Specifically, creating a mobile tomato-harvesting robot.

The researchers followed all of ChatGPT’s design decisions. “ChatGPT extends the designer’s knowledge to other areas of expertise,” said Della Santina. “For example, the chat taught us which crop would be most economically valuable to automate, to “make the gripper out of silicone or rubber to avoid crushing tomatoes” and recommended that “a Dynamixel motor is the best way to drive the robot.”

ChatGPT as researcher/designer — is that a good thing?

In a paper just published in Nature Machine Intelligence, the researchers explore the varying degrees of cooperation between humans and Large Language Models (LLM). In this case, the LLM acts as both researcher and engineer, while the human acts as manager, in charge of specifying the design objectives. So that raises the question: is that desirable?

“In fact, LLM output can be misleading if it is not verified or validated,” Della Santina says. “AI bots are designed to generate the ‘most probable’ answer to a question, so there is a risk of misinformation and bias in the robotic field.” Working with LLMs also raises other important issues, such as plagiarism, traceability and intellectual property, she notes.

Della Santina, Stella and Hughes are continuing their study of LLMs to design new robots, including designing the robot’s own body. “Ultimately, an open question for the future of our field is how LLMs can be used to assist robot developers without limiting the creativity and innovation needed for robotics to rise to the challenges of the 21st century,” says Stella.

Citation: Stella, F., Della Santina, C., & Hughes, J. (2023). How can LLMs transform the robotic design process? Nature Machine Intelligence, 1-4. https://doi.org/10.1038/s42256-023-00669-7

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

Generative AI Model Radically Improves Electrocardiogram Diagnoses

Researchers at the Mount Sinai Health System have developed an innovative generative AI transformer model that improves the accuracy and effectiveness of ECG (electrocardiogram)-related diagnoses.

In a study published in the June 6 online issue of npj Digital Medicine, the team reported that its new deep-learning model, HeartBEiT, surpassed established commonly used AI methods for ECG analysis — convolutional neural networks (CNNs), used for computer-vision tasks.

Researchers pre-trained HeartBEiT on 8.5 million ECGs from 2.1 million patients collected over four decades from four hospitals within the Mount Sinai Health System. Then they tested the model’s performance against CNN architectures commonly used in cardiac diagnostic areas.

The study found that HeartBEiT had “significantly higher performance at lower sample sizes, along with better ‘explainability’ and can perform as well as, if not better than, these methods, using a tenth of the data,” according to Akhil Vaid, MD, Instructor of Data-Driven and Digital Medicine (D3M) at the Icahn School of Medicine at Mount Sinai.

The researchers tested the model on three tasks: learning if a patient is having a heart attack; if they have a genetic disorder called hypertrophic cardiomyopathy; and how effectively their heart is functioning. “In each case, our model performed better than all other tested baselines.”

This study was funded by the National Heart, Lung, and Blood Institute of the NIH and by the National Center for Advancing Translational Sciences of the NIH.

Citation: Vaid, A., Jiang, J., Sawant, A., Lerakis, S., Argulian, E., Ahuja, Y., Lampert, J., Charney, A., Greenspan, H., Narula, J., Glicksberg, B., & Nadkarni, G. N. (2023). A foundational vision transformer improves diagnostic performance for electrocardiograms. Npj Digital Medicine, 6(1), 1-8. https://doi.org/10.1038/s41746-023-00840-9 (open-access)

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

In a first, a space solar power system has wirelessly transmitted power in space

A Caltech space solar power prototype, launched into orbit in January, is now operational, and Caltech engineers have demonstrated its ability to beam detectable solar power to Earth for the first time.

A Caltech ground station on Earth confirmed the transmission, according to Ali Hajimiri, Bren Professor of Electrical Engineering and Medical Engineering and co-director of the Caltech Space Solar Power Project (SSPP).

“When fully realized, Caltech’s SSPP will deploy a constellation of modular spacecraft which will collect sunlight, transform it into electricity, then convert it to microwaves that will be transmitted wirelessly over long distances to wherever it is needed — including locations that currently have no access to reliable power,” he said.

The ultimate goal: “A world powered by uninterruptible renewable energy.”

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter. 

Accelerated wound-healing ink uses 3D-printing pen

Researchers in China and Singapore have developed a wound-healing ink that can actively encourage the body to heal. The ink can be spread into a cut of any shape using a standard 3D-printing pen. In mice, the technology nearly completely repaired wounds in just 12 days.

This method accelerates the natural healing process, in which extracellular vesicles (EVs) secreted from white blood cells play important roles in promoting blood vessel formation and reducing inflammation during healing.

PAINT treatment process

(credit: ACS Applied Materials & Interfaces)

The team developed a system called PAINT (portable bioactive ink for tissue healing) using EVs secreted from macrophages combined with sodium alginate. These components were combined in a 3D-printing pen, mixing at the pen’s tip and forming a sturdy gel at the site of injury within three minutes.

The EVs promoted blood vessel formation and reduced inflammatory markers in human epithelial cells, shifting them into the “proliferative” (growth) phase of healing.

Citation. Paintable Bioactive Extracellular Vesicle Ink for Wound Healing. ACS Appl. Mater. Interfaces 2023, 15, 21, 25427–25436. May 19, 2023. DOI: 10.1021/acsami.3c03630

Let us know your thoughts! Sign up for a Mindplex account now, join our Telegram, or follow us on Twitter.