Unite EMR

Inspiration

Every year countless hours are lost to hospitals who take the time to convert data between other hospitals or even internally. In many cases, tests which don't need to be rerun are rerun, questions that don't need to be asked are reasked, and many other pieces of vital information are thrown away. Additionally, doctors often make these types of retests because they don't have the time to utilize what information is already there. These types of situations can be incredibly stressful to patients both mentally and financially. Imagine being a patient who is struggling with your new diagnosis and is being asked the same questions over and over by different sections of the same hospital never leaving your bed. It would drive anyone insane, and make you doubt the integrity of your care.

Countless doctor hours that could be spent saving lives are spent on menial tasks supporting cross-hospital or internal hospital data when they could be spent saving real lives. Imagine a world where electronic medical records of any type could be converted to seamlessly integrate with different formats. That is the world that Unite EMR is vying to create.

What it does

Our purpose is to enable fast conversion between electronic medical record data types. You can convert using any of the formats we handle into a universal format. HL7 FHIR, chosen for its wide implementation across many commonly used EMR software such as EPIC, Oracle, MEDITECH, and many more. During conversion, you can customize as much as you want depending on the headers/information you would like to extract from your document. Then, you can choose to export it as is or into one of the many file formats we offer. This includes an option to submit an example of the format you would like your data to be exported in.

How we built it

We built Unite EMR using a combination of different methods for different types of conversion pipelines. Currently we support three main different distinct methods for conversions; file, image, and audio.

We also support conversion to text and speech so that doctors who don't have time to review a patient's entire medical history have the chance to listen to or read a summary.

Challenges we ran into

As patient information is extremely protected and susceptible to privacy risks, we have taken that into account with parts of our implementation to show how it could be done on a larger scale.

Accomplishments that we're proud of

Successfully converts file types of audio, HL7 files, CSV's, and images!

What we learned

The true complexity of electronic medical records and the intricacies or preserving data integrity.

What's next for Unite EMR

Security! Security! Security! The next steps in this prototype's development is converting every conversion method into a local system, such as a local model or otherwise, to improve any potential security risks. Currently, we encrypt data for most of our methods before sending them to any type of processing on the cloud, but we did not have the time or capacity to complete this process across all methods.