# What diseases can be identified by some defined algorithm?

I'm looking for some examples about diseases that can be identified by some equations? I mean something like, that if some person have this overweight, this age, these medical findings that means that in 70% he can suffer from heart disease.

Why I need that? I'm working on a project about homomorphic encryption - in summary, encryption that allows to make mathematical equations on encrypted data. I'm looking for a use-cases for this and have an idea about medical data that could be encrypted with this and patient could send his data totally secure to check if he can suffer from some illness. To present this I'm looking for some example diseases that can be detected just looking at the findings and making some mathematical equations on them.

Below quick diagram of this demo system.

I was looking for this from a long time but web searching doesn't quite understand my request and I cannot find data that interests me. If someone can advise where I can look to find such data I would be very thankful!

• Like risk of death from covid-19? Commented Apr 25, 2020 at 20:55
• If the data can contain lab results there are literally hundreds of trivially simple algorithms available. Using @GrahamChiu's COVID-19 example: dry cough + shortness of breath + positive SARS-CoV-2 test = COVID-19. If you simply google something like "diseases that can be diagnosed by urinalysis" you'll find plenty of examples. Commented Apr 25, 2020 at 21:16
• I'm looking more like for a numerical algorithms, on which I could make some mathematical operations. So as in example something like if patient have 50kg overweight, is 70 years old and other numerical attributes then... Commented Apr 25, 2020 at 22:08
• Point about urinalysis sounds more like it. I will look closer what can i find there. Thanks! Commented Apr 25, 2020 at 22:09
• If your application can't handle boolean data, which comprises a huge amount of medical data, and which is still actually mathematical (1 != 0), then your application is crippled entirely. Commented Apr 26, 2020 at 3:53

Some diseases have laboratory result-based definitions. For example, diabetes is defined by some organizations as a fasting blood sugar value over a certain level, or other similar laboratory test results. For a project related to encrypting health data you may wish to start with simpler examples like this to understand how you might protect the identities and the data of individuals while processing the algorithm on encrypted data.

https://www.diabetes.org/a1c/diagnosis

“Diabetes is diagnosed at fasting blood sugar of greater than or equal to 126 mg/dl”