DREAM / The Prime Minister resigns and wakes up under the Parisian bridge!

 Kliko këtu për të tjera
Voxpress:

Life-saving, positive or futile efforts?

2022-06-05 08:56:27Kosova&Bota SHKRUAR NGA REDAKSIA VOX

It's really fantastic: a sky blue indicator on the screen that shows your blood sugar level drops sharply. The cause is obvious: the banana cakes you ate in the morning.

These changes are indicated by your glucose monitor, a device as large as a coin, which attaches to the skin and can even be displayed on the screen of your smartphone. The device lasts for two weeks and has a small needle inserted under the skin. Every few minutes, it measures the concentration of sugar in the fluid between the cells, which more or less reflects exactly what is happening in the circulatory system.

As soon as such glucose monitors began to replace fingerprint blood tests for diabetics in 2014, they also began to appear in the arms of non-diabetics in Silicon Valley.

These people were looking for ways to "hack" their metabolism, to boost more energy, or clarity of the brain. Their ad-hoc experiments were repeated by official researchers conducting more extensive metabolism studies. On the other hand, these initiatives have also developed the field of personalized nutrition.

Such metabolic studies have changed the scientific mindset on how a healthy diet should be. It has been found that many people who can look healthy often experience large increases in blood sugar after eating a meal, which are known as pre-diabetic symptoms.

If no medication, or preventative measures are taken, prediabetes turns into diabetes in 37-70% of cases, within four years. Immediate reductions in blood sugar levels, followed by immediate increases, have been found to be also problematic because they make people feel hungry. People who experience such marked drops in blood sugar consume about 300 calories more per day than those who do not have such symptoms.

In 2015, researchers in Israel created an Artificial Intelligence-based algorithm that could predict how different foods would affect blood sugar levels.

The algorithm data included blood tests, sleep, exercise, height, and weight, all of which affect daily metabolic variations. They also included the composition of the gut microbiome, which are the trillions of bacteria that live in the gut and process the foods we eat. Microbiome analysis is performed by genomic sequence taken from a stool sample.

In the last five years, startup companies in America, Europe and Asia, have created personalized food applications, based on Artificial Intelligence, that rely on these discoveries.

By knowing exactly what is in the foods we eat and by measuring the changes in blood sugar and fat levels that result from eating these foods, the company can create a predictive model of its customers's metabolism.

Her algorithm then creates a custom catalog of foods and meals. Tushar Vashisht, co-founder of HealthifyMe, an Indian startup company that offers digital weight loss training, says customer data collection is very useful. It can help build applications for people who do not have such devices.

For such systems to keep their metabolism in balance, they must be respected. Knowing what is going on inside your body is useless if you do nothing to change your behavior. Apps sold as a way to achieve health benefits, based on measurements made by "body-worn" devices, may not be effective if the user does not fundamentally change their lifestyle and regimen.

The novelty that such personalized diets bring makes them easier to maintain in the long run, because it gives people different opportunities to cook foods, which according to the algorithm are particularly bad, or less harmful.

Algorithms can give tips such as, sprinkling some nuts on ice cream, or a long walk after eating. January AI, another startup company on personalized food, says it has released the nutritional content of 16 million local grocery store items, recipes and menu items, making it easier for users to plan and track their meals .

Positive results

It is still early, but the results reported by users of such nutrition programs, look positive. Users say they are losing weight, have higher energy levels and are sleeping better. Some diabetic users no longer need medication. Some studies want to determine such benefits more accurately.

However, as is often the case with modeling new technology products, behavioral science is neglected. Natasha Schüll, cultural anthropologist from NYU, noted that it was very difficult to decipher logic, after many health technology products.

Then, she realized that in fact, there was no logic at all. It was "a combination of two or three psychological views that resembled a set of ad-hoc thoughts that were not studied or scientific." The model's typical strategy, she concluded, "was based on the idea of ​​making a lot of claims in the hope that something would stick."

This may be why many applications fail to become a regular part of people's lives. There are more than 400,000 health and wellness apps in the Apple and Google app stores and 250 new ones are added every day. Their appetite is healthy, with about 5 million app downloads per day. But 95% of them are deleted within 24 hours.

The problem is that people do not simply need a product that is well-designed. They need a product that is well-designed specifically for them, says Liz Ashall-Payne of Orcha, a British organization that values ​​the quality of healthcare applications.

As she points out, buying a pair of pants online is facilitated by filters for size, color and style, but such a system does not exist in app stores. A teenager seeking help for anxiety will need some sort of different application from his grandfather, though both may want the same thing.

Lots of wasted effort

Orcha has rated 7,000 health apps with three criteria: privacy, user experience and proof that they work.

Only about a quarter of them meet the minimum quality threshold in all three categories. Mental health applications are particularly poor. But the quality is improving, says Ms. Ashall-Payne. As the market matures, bad products will disappear, "but it will take time."

Encouraging people to adhere to healthy behaviors is probably the biggest challenge in public health. The latest innovations that increase compliance are therefore important. The technology developed by Sweetch, the Israeli startup company, is making the advice given by health applications more practical and personalized.

Sweetch's Artificial Intelligence algorithm is a hybrid model of a personal secretary and motivational trainer. It keeps track of everything the user has to do, e.g. perform a certain number of steps or weight control each week, and find the best time to advise these tasks.

The alert robot may notice, for example, that you have 20 minutes of free time and may advise you to go for a walk. It changes goals depending on how you act, to prevent discouragement. Requests are available in 33 different voice tones, with word combinations that can be friendly, demanding, commanding, etc.

Yoni Nevo, Sweetch chief executive, says the algorithm estimates about 700 million possible word combinations to say to an individual, on any given day. It takes about four to five weeks for the algorithm to adapt to your habits.

Device manufacturers, for their part, are beginning to realize that when it comes to giving people health information, moderation is the right way to go. Many people just want to know if something is good or bad, instead of being overwhelmed with data as early as morning.

Some smart scales now just buzz enthusiastically when you step on them, to confirm that your weight and body fat percentage are in the target range, and give no figures at all.

Devices like the Oura ring are now able to tell you not only what your heart rate, sleep or temperature trends are, but also what that means - and what changes can make things better. And when the data shows that things have gone wrong, new digital treatments can save you./ The Economist-monitor


Në përputhje me standarded JTI

Video

SHIKO KANALIN NE YOUTUBE
×