Hacker News Reader

Comment by Nk26 | Hacker News Reader

Does the photo recognition attempt to carb count what it sees? Is that even possible? My son is a T1D and he still struggles with carb counting.

View on HN

It is absolutely not possible to carb-count through photo recognition in a way that is reliable enough for a diabetic to safely use to make treatment decisions.

I meant more in a general way, like a piece of pizza is usually around X carbs. We have apps that make the guess a bit easier but it's almost always a guess. I was thinking could this look at a photo and know there's a sweet potato, a piece of chicken and some corn and give a basic idea.

The answer is still Absolutely Not, especially since all food can involve a treatment decision for people with type 1 diabetes.

Pizza is a good example of why not. Slices come in very different sizes, sauces have very different carb content, so do crusts, and toppings.

Edit: for example this pizza(1) is 31g per slice and this pizza(2) is 73g per slice. The difference is very meaningful and the “general idea” given by photo recognition would likely be wrong to the point of dangerous for a diabetic in both cases.

If you’re looking for software that can make a guess simply for the sake of generating a number to write down and not be used in any way, a random number generator would be safer since the risk of output being misconstrued as actual information is much lower.

1 https://www.costcobusinessdelivery.com/kirkland-signature-ca...

2 https://sbarro.is/product/bbq/

What do you use then to make these decisions? If you use your eyes, app, nutrition label or Chatgpt, you would still have the same variables. You're still making the decision based on averages, and best guesses.

I use nutrition labels. I have absolutely no idea whatsoever why anyone would lump nutrition labels in with your eyes or chatgpt.

The people that make the label make the food. They know what they put in it. Because they made it. They wrote down what they put in it for you to read and make decisions off of. The difference is categorical.

I cook myself and i know which and how much ingredients i use and how much carbs they contain. Either from a food label or in general (like 100g of cooked potatoes contain about 16g carbs).

Then I calculate how much my serving contains.

Depending on what you eat, what type of diabetes you have and how it’s treated you may have to consider the amount of protein and fat as well (they slow digestion and cause a delayed rise in blood sugar levels). If you have an insulin pump you may want to program a delayed insulin dose to handle that.

Sounds complicated? It is, but only during the first weeks. You quickly learn the carbs content of the food you frequently eat and learn to estimate how much is on your plate. Like, two units for a bun. There are also great nutrition apps out there that help a lot.

Personally, I take a representative sample and then use a calorimeter to test it. Anyone who doesn't do this is being grossly irresponsible and will only have themselves to blame when they eat so dangerously. I recommend a CK 5E-C5808J but you have to ensure a trained professional is helping you. Otherwise, you might as well not eat at all.

Yep. And the issue with pizza is the amount of fat that comes with the carbs. This quite often (depending on the position of the moon) gives you some of the carbs when you eat it to your blood, and the rest will come after several hours. What you want to do is to inject a bit of insulin before eating, then after two or three hours more while measuring your glucose levels.

Of course if you eat a Neapolitan pizza with not that much of cheese everything changes again. And YMMV, I'm just talking about my experiences.

Not only fat plays a role with pizza, but also the amount of protein in it. When having pizza we usually add protein to the carbs. 50% immediately bolus. Other 50% spread over 3-4 hours, and let AAPS dose the insulin.

Yeah, you can try this on the ChatGPT app. Take a picture and ask ChatGPT to give you the nutrition info, then do your own calculations based on weight and the USDA database and see how it compares.

Similarly, chatgpt can run a mile for you if you ask it to and then get up and run a mile.

However, if you dont have carb info, the alternative is to judge yourself. Your own model may be better than gpts model, though. I would use GPTs output and at least look at it on a case by case basis

Indeed, around 2019 I was reading many computer vision papers for volume estimation and came across a few that tried to estimate the weight of the meals from pictures using the size of known objects (cutlery beside the plate). The idea was good but they were very far from accurate and not robust at all, and that was just for the weight, not even carb counting. I know CV is a fast moving field but I wouldn't bet that the tech has improved enough to be anywhere near medically safe.

It's probably slightly worse than an educated guess?

>Is that even possible?

No.

Reply on news.ycombinator.com