-
-
Ibuprofen safe dosage range
-
Amoxicillin safe dosage range
What does the project do? Our project creates healthcare accessibility for prescription dosages. Rather than searching through medical websites, any person can access our website and check if their dosage prescription falls within the recommended dosage range for their age and weight. The website increases awareness for patients about the medicine they are taking. Moreover, healthcare providers can use the website as a fast and easy double-checker before giving prescriptions to patients.
Our frontend generates our website, which accepts inputs from a user. The information is then sent to our backend and model, where the input is compared with data from our Kaggle dataset, weighing the recommended, maximum, and minimum dosages with the input. The backend returns whether the input is within the recommended dosage range.
Because most datasets relating to this issue require training certificates in healthcare, our selection of datasets were very limited. The one we found only had 1000 data points, so our model currently has a 32% percent accuracy.
Although we ran out of time, we were planning to add features like assessing drug interactions and possible side effects.
The healthcare problem it solves: Incorrect dosage is one of the most common and dangerous medication mistakes, accounting for over half of all medication-related errors and 41% of deaths related to these errors. These mistakes often come during prescription and dispensing, which can cause fatal overdoses or ineffective treatments from overly low doses.
Children are about three times more likely to experience some types of prescription mistakes than adults are, which makes them more vulnerable to this issue.
Over 50% of dosing errors in children lead to overdose, and small mistakes like confusion between mg and mL or conversion errors between kg and lbs can cause a 10-fold fatal overdose in children.
Our project aims to reduce the frequency of such errors by allowing doctors to double check their prescription before giving it to children or adults. By doing so, small mistakes that lead to big consequences can be prevented. This project is not intended to give an exact dose, but rather act as a tool for doctors and a safety measure to help mitigate mistakes, as well as a tool for parents of kids taking medications or people taking medicine in general to have peace of mind that their prescription is correct.
How AI is used: We prompt-engineered our frontend and model using Claude AI and our backend using ChatGPT. We also used it for debugging and troubleshooting the connection from our backend to our frontend.
Who benefits from your solution: Adults and children who take medication, parents of kids who take medicine, and healthcare providers that make prescriptions. All of these parties benefit as the project provides reassurance that their prescriptions are correct and won’t result in overdoses or ineffective treatments for themselves, their kids, or their patients.
Log in or sign up for Devpost to join the conversation.