91av

Electromagnetic radiation leaking from a microwave oven can provide information for estimating the nutrients and calories of the food inside. That could enable smart microwave ovens to measure the nutritional value of both home-cooked casseroles and reheated restaurant leftovers.

“Even if we individually measure the ingredients’ calorie counts and the nutritional values and you try to actually cook them all together, it’s very hard to actually come up with a very good estimate,” says at The Ohio State University. “We have shown that this [method] is actually a more reliable way of counting caloric values and nutritional values in the food.”

Srinivasan and , also at The Ohio State University, have shown how off-the-shelf radio receivers – such as those found in smartphones and Wi-Fi networking devices – can be calibrated to detect specific food-related signals in electromagnetic radiation from microwave ovens. The signals indicate how different food molecules vibrate and heat up at different rates.

A prototype system using a small radio antenna receiver placed in front of a microwave and hooked up to software for analysis demonstrated 81 per cent accuracy in detecting the percentages of water, protein, carbohydrates and fat in more than 150 different foods. The food selection included both home-cooked and restaurant meals such as pizza, cooked salmon and vegetable biryani.

For best accuracy, the system must include the pre-calculated effect of different microwave containers made from plastic, glass and porcelain, as the container materials, surface area and orientation can all affect the microwave leakage pattern. It also needs to know the mass of the food in order to calculate the total calorie content. The researchers also haven’t yet trained it to detect fibre – a significant component of many foods.

Still, the researchers say it already outperforms previous attempts to calculate food nutrition and calories based on images. “Since our work is dealing with the intrinsic properties of the food, I think we are doing much better,” says Banerjee.

The concept makes scientific sense, but given the lack of published detail about it, the system remains a “black box operation”, says at McGill University in Canada. The researchers have applied for a patent.

“They can say that whatever is in the microwave is a biological material and it’s likely to have water, it’s likely to have protein, it’s likely to have carbs,” says Orsat. “The issue is in the precision level of that, and the level of training they would have to do in their device.”

The system would need to be calibrated each time it is moved to a new location relative to the microwave oven. That could be trickier for a smartphone app implementation, but less of a problem if installed with a fixed Wi-Fi access point or built into a future microwave oven. A microwave oven might someday use this method to help automatically determine the optimal cooking time and power level for a given meal, says Srinivasan.