An emotional question by a mother whose child has food allergies led me to discover the fascinating world of food scanners. Will we ever really know what’s on our plates? And if we knew, what could we do with the answer?




I often blog about trends that promise to have the biggest influence on medicine in the coming years. These posts receive a lot of feedback from people around the world. After I published predictions for how medical technology will advance in 2015, a mother left an emotional comment about her daughter of 8 years who was allergic to most foods.

She described how hard it was to find food her daughter can eat. She had to check labels diligently for allergic ingredients, and she had to cook very carefully. She was pregnant at the time she wrote the comment, and worried. Were there any labeling advances or technological devices that could make her life easier? Was it possible to scan foods for their ingredients?

Food scanners and printers create food on demand

The problem with our foods

Extrapolating from European statistics, 220 to 250 million people may suffer from food allergies worldwide. In the US one out of three people has a food allergy or modifies the family diet because of a suspected food allergy in a family member. Approximately 5% of children and 4% of teens and adults have clinically proven allergic reactions to foods.

Globally, about 35 % of adults are overweight, with half a billion being obese with a body mass index of over 30. According to The International Diabetes Federation, around half a billion people will have diabetes by 2035; half of them will be undiagnosed. Food allergies, obesity and diabetes are only the most striking examples where knowing calorie intake and ingredients are pivotal in managing medical issues day and day out.

Given these statistics, the European Commission launched a €1 million prize challenge to develop an affordable, mobile, non–invasive solution that enables users to measure and analyze their food intake. The contest runs until the end of 2016. Hundreds of other similar initiatives are needed to help people deal with diabetes or obesity.

We really have no idea what we eat now

We can only guess. Ingredients are listed on most of the products we buy, but every meal and every plate is different. One remedy would be having a list of exactly what ingredients and how many calories a meal contains, and what allergens and toxins might be in it. Not just the kind of meal we eat but the actual food on our plate and its specific amount. Several companies have been trying to address this.

Canadian TellSpec had a successful crowdfunding campaign in 2013 that raised more than $380,000. Now it aims to develop a hand–held food scanner that can inform users as about specific ingredients and macronutrients. Through spectroscopy, an analysis of how matter interacts with different wavelengths of light, it can quickly determine the chemical compounds in a given food. The company founder, Isabel Hoffmann, had a story similar to that of the mother who commented on my blog. Hoffman’s daughter got sick after the family moved to a new flat in Toronto. Months later the girl was found to have a variety of food allergies.

As a result, Hoffmann recruited a team to design a device that could tell what was in a particular food. That way she and others could know what to avoid. They sent out the first devices to beta testers in mid-2015. Several critics have raised questions about the device and whether the company can deliver what they promised. Backers in the crowdfunding campaign seem willing to wait and see. The first commercially available models are launching in 2016.

The promise and problems of food scanners

Another device, SCiO, from Israel, was founded by people with optical engineering backgrounds. They have raised over $2.7 million on Kickstarter in 2014. It uses a technology similar to TellSpec’s but is designed to identify the molecular content of foods, medicines, and even plants. It illuminates an object; optical sensors detect the reflected light; and the device analyzes it using an algorithm and a cloud–based database that is constantly updated. The company promises that in milliseconds the ingredients and molecular make–up of the foodstuff will appear on the user’s smartphone. The device was scheduled for shipment in 2015, but disappointed backers, but still hasn’t shipped. On top of the delays, experts in applied science criticized both SCiO and TellSpec for overstating what their inventions can do.

Scio food scanner scans cheese for ingredients


There are two major issues. One is size, because the device must be hand–held to become popular. With current technology this means engineers have to sacrifice sensitivity and accuracy in order to achieve a convenient size. The other issue is the algorithm. SCiO sends data to the cloud which then sends its calculation back to the device. But to simplify what the algorithm has to do, users need to tell the scanner specifics – like whether the sample is a solid food, a liquid, or vegetable. These inconveniences are the price of keeping the scanner small.

There aren’t any promising hand–held food scanners on the horizon besides these, but there is no reason to believe a solution will not arise in the coming years. The challenge is not when a workable device comes along but what we will do with the large amount of data it generates.


Big data and the Internet of things will improve nutrition

Let’s say a scanner tells me how many grams of sugar my fruit contains, or what the alcohol percentage of a drink is. So what? It won’t change my behavior and dietary habits unless I’m a dietitian and understand what the data means, and how it can be acted upon. Food scanners will need to progress similarly to wearable health trackers – move from raw data to automated analysis and smart suggestions to the user.

A good food scanner should accurately determine ingredients, and compare the data to my lifestyle, dietary choices, and my genomic background. Given how different we all are genetically, two people might digest the same food at a different pace. One might be allergic to an ingredient while the other is not. So far, pure luck and experience have alerted us to these differences. It should not work like that. Eating should be a conscious process where we know what we eat, and know what we should eat for optimum health. A food scanner, supported by a smart application could fill this place.

But let’s not leave out an interesting side note here, namely, incorporating genetic information into food scanners. I already have the data of my complete DNA sequence at home in a digital file. Literally thousands of studies speak to the genetic aspects of nutrition, a field called nutrigenomics. I should be able to learn what foods and individual ingredients are bad for me. Genetic tests showed me that I’m sensitive to caffeine and process alcohol more thoroughly than most people (I’m Hungarian after all).

Nutrigenomics tries to understand how nutrition affects our metabolic pathways, and what we can do to get the most out of nutrition in a personalized way. If I’ll have the opportunity to choose another type of meat or cheese as a smartphone app suggests based on my DNA, I will enjoy the meal more and take better care of my body in the long run. With access to such data, a scanner or app could tell us what products not to buy at the grocery store, what type of food makes us more productive, sleep better, or just feel healthy. Right now we’re depending on blind luck.

Some people wonder if this wouldn’t be an overly technological world where devices, scanners, and apps tell us what to eat and do. I prefer to look at it from a different angle, from the benefits of finally knowing what we eat and what ingredients lead to positive and negative consequences. I see customization to my specific genetic background as another benefit, too.

Diabetes patients would know how many carbohydrates their food contains. But knowledge doesn’t change behavior alone, otherwise nobody would smoke by now. Knowledge supported by gaming or technologies revealing our lifestyle choices to our family members or caregivers might do. Patients with rare genetic metabolic disorders such as phenylketonuria would know what to avoid at all cost. People with allergies could avoid dangerous meals. Having a good diet would not rely on the experience we bring with us from childhood and what we have learned since then. Instead, it could be based on informed decisions. If it means a food scanner should become a commodity in my life for this, count me in.


[The Medical Futurist]