Kalakshar US

Synex Medical, a Toronto-based biotech research firm backed by Sam Altman (the CEO of OpenAI), has developed a tool that can measure your blood glucose levels without a finger prick. It uses a combination of low-field magnets and low-frequency radio waves to directly measure blood sugar levels non-invasively when a user inserts a finger into the device.

The tool uses magnetic resonance spectroscopy (MRS), which is similar to an MRI. Jamie Near, an Associate Professor at the University of Toronto who specializes in the research of MRS technology told Engadget that, “[an] MRI uses magnetic fields to make images of the distribution of hydrogen protons in water that is abundant in our body tissues. In MRS, the same basic principles are used to detect other chemicals that contain hydrogen.” When a user’s fingertip is placed inside the magnetic field, the frequency of a specific molecule, in this case glucose, is measured in parts per million. While the focus was on glucose for this project, MRS could be used to measure metabolites, according to the Synex, including lactate, ketones and amino acids.

Matthew Rosen, a Harvard physicist whose research spans from fundamental physics to bioimaging in the field of MRI, told Engadget that he thinks the device is “clever” and “a great idea.” Magnetic resonance technology is a common technique used for chemical analysis of compounds, however, traditional resonance technologies operate at high magnetic fields and they’re very expensive.

Synex found a way to get clear readings from low magnetic fields. “They’ve overcome the challenges really by developing a method that has high sensitivity and high specificity,” Rosen says. “Honestly, I have been doing magnetic resonance for thirty years. I never thought people could do glucose with a benchtop machine… you could do it with a big machine no problem.”

Professor Andre Simpson, a researcher and center director at the University of Toronto also told Engadget that he thinks Synex’s device is the “real deal.” “MRI machines can fit an entire human body and have been used to target molecule concentrations in the brain through localized spectroscopy,” he explained. “Synex has shrunk this technology to measure concentrations in a finger. I have reviewed their white paper and seen the instrument work.” Simpson said Synex’s ability to retrofit MRS technology into a small box is an engineering feat.

As of now, there are no commercially available devices that can measure blood glucose non-invasively. While there are continuous glucose monitors on the market that use microneedles, which are minimally invasive, there is still a risk of infection.

But there is competition in the space for no-prick diagnostics tools. Know Labs is trying to get approval for a portable glucose monitor that relies on a custom-made Bio-RFID sensing technology, which uses radio waves to detect blood glucose levels in the palm of your hand. When the Know Labs device was tested up against a Dexcom G6 continuous glucose monitor in a study, readings of blood glucose levels using its palm sensor technology were “within threshold” only 46 percent of the time. While the readings are technically in accordance with FDA accuracy limits for a new blood glucose monitor, Know Labs is still working out kinks through scientific research before it can begin FDA clinical trials.

Another start-up, German company DiaMonTech, is currently developing a pocket-sized diagnostic device that is still being tested and fine-tuned to measure glucose through “photothermal detection.” It uses mid-infrared lasers that essentially scan the tissue fluid at the fingertip to detect glucose molecules. CNBC and Bloomberg reported that even Apple has been “quietly developing” a sensor that can check your blood sugar levels through its wearables, though the company never confirmed. A scientific director at Synex, Mohana Ray, told Engadget that eventually, the company would like to develop a wearable. But further miniaturization was needed before they could bring a commercial product to market.

Rosen says he isn’t sure how the sensor technology can be retrofitted for smartwatches or wearables just yet. But he can imagine a world where these tools complement blood-based diagnostics. “Is it good enough for clinical use? I have to leave that for what clinicians have to say.”

This article originally appeared on Engadget at https://www.engadget.com/researchers-use-magnetic-fields-for-non-invasive-blood-glucose-monitoring-215052628.html?src=rss 

A set of features rolling out to Google Photos today will make it much easier to declutter your photo library, the company announced in a blog post. Google Photos will now automatically identify similar photos that you took in rapid succession – helpful for those times when you clicked 50 shots of that gorgeous sunset to get the one perfect frame you will never look at again – and group them in a single “stack” to clean up your library.

The service will select a top pick that best represents the moment, but you can manually choose an image you want too. If you prefer to have multiple sunsets littering your library, you can turn off stacking.

Photos will also automatically organize your pictures, separating IDs, receipts, and tickets into different albums, a feature that seems like it should have been there ages ago given how good Google Photos is at recognizing what’s in your images. You can also add an event to your calendar directly from a screenshot or a photo of a ticket, which seems really useful.

Unsurprisingly, Google says that all these features are powered by AI, something that the company has lately been cramming into all its products in general. Google Photos, however, has always been powered by AI. For years, the service sucked up all your photos and stored them in exchange for training its machine learning algorithms on them to be able to recognize their contents. That’s why you can search Google Photos for pictures of dogs or the beach, for instance, so easily (Google stopped providing free storage for Photos a couple of years ago).

Earlier this year, Google added another AI-powered feature to Photos that creates personalized scrapbook montages. Google’s algorithms sort your photos into relevant categories and create titles that you can modify if you want. Google also lets people use AI to easily make edits, such as removing unwanted people or objects from photos.

This article originally appeared on Engadget at https://www.engadget.com/google-photos-has-new-ai-powered-features-to-clean-up-your-library-204512642.html?src=rss