A device capable of identifying a wide variety of prescription medications and providing timely information about the drug

According to the US Centers for Disease Control there were 123.8 million emergency room (ER) visits in the US in 2010, many involving potentially life-threatening injuries or conditions where swift triage and treatment is often required. Medical professionals must quickly ascertain any medications a patient may be taking. Patients may be unresponsive, unaware of the names of their medications, or fail to carry their medications in the original prescription containers, making identification of the drugs difficult and potentially time consuming. Further, with the large number of drugs on the market, medical personnel may not be immediately familiar with a drug’s side effects and contraindications.

We propose a device for prescription drug identification (PDI) capable of identifying a wide variety of prescription medications and providing timely information about the drug from on-line or built-in databases. Drugs can be identified via a camera or by scanning a bar code from a prescription. Upon successful identification of the drug, pertinent information about the drug is displayed.

The first working prototype of PDI was developed in 6 months. The system implements 3 drug databases with the maximum one has 18,000 types of drugs approved by FDA. The automation process for input from camera takes less than 15 seconds with the accuracy of 86.2%. For the input via bar-code module, the system output results with accuracy of 100% in less than 2 seconds. The result above is from the test performed on over 39 types of drugs that we collect from student health center in our college.

The device took top honors at the Cornell Cup USA 2012 competition.