Google Partnering with Doctors in New Medical Project
June 16, 2018
Google is known for their innovation in technology and internet practices, but the tech giant has decided to branch out even further, partnering with doctors in a new project. This latest endeavor is called Medical Digital Assist, and it will integrate AI with doctor’s visits to help improve wait times and registration processes for all involved.
One of the major ways Medial Digital Assist will help is in taking notes for doctors. The program will use voice recognition software, along with touch and sound tech to provide accurate transcriptions for later research. This is clearly the beginning of something big as Google has already posted four new job listings hoping to amp up their efforts to streamline and motivate this process.
Medical Care and Google
In North America, as well as countries around the world, medical care personnel are often tasked with an overwhelming amount of work and information daily. While doctors, nurses, and emergency services personnel save lives, they often do so without the support required. Google has made it clear that they wish to help these efforts to improve hospitals, clinics, and doctor patient relations.
The project is part of the Google Brain department of the search engine’s network, falling under the artificial intelligence program. While the project is new and in the early stages, Google is reportedly hoping to take it public with trials within the year. It could integrate alternative Google technology already available, such as Home, Translate, and the Google Assistant application.
Ultimately it could lead to something much larger, maybe even being used in alternative forms of medicine, such as dentistry or optometry if things go well. Katherine Chou, one of the co-founders of the Google Brain team, dedicated to their healthcare related projects recently launched a study through Stanford University, which proceeded to test a speech to text tool for doctors. This helped reduce the amount of time typing and writing for physicians, which could be spent listening to patients.
The automated voice to text feature is meant to increase patient doctor communication and decrease time spent filling out paperwork and charts. Of course, this comes with its own set of potential problems, including accuracy, privacy and security of patient information. The tests run through the Stanford Medical project show that the new AI will save physicians approximately two hours of paperwork for every one hour spent with a patient. This is immense, considering the wait times in most medical clinics across North America. In emergency service centers, it will also increase the speed at which doctors move though critical patients.
Privacy and Accuracy
Like all electronic application, speech to text has its faults – one being accuracy. The public can attest to discrepancies in texts sent using voice, much the same as a standard text message can fall prey to autocorrect and spell check. Sometimes information is misheard or misspoken, and the concern is, of course, that artificial intelligence may not catch it the way that a human eye might.
One example where a simple error of this nature could create big problems for patients is in the difference between hypo and hyper. Hyperthyroidism, for example, is the description of an overactive thyroid which is producing too much hormone and causes a heightened metabolism, anxiety, and other issues. Hypothyroidism, though similar in sound, is the exact opposite, causing the metabolism to slow, and causing mood swings, fatigue and miscarriage in pregnant women. Prescribing the wrong medication based on a miscommunication could mean big problems for patients if the mistake isn’t caught.
Although this project is new and hasn’t had much time to accumulate opinion or perspective, feedback has been mainly positive, with many boasting hopes that the program eliminates wait times and other delays related to medical clinics. Comments across the internet indicate hopes that the process will speed things along as promised and not slow them further. With the machine-error ratio to consider, there’s a possibility that doctors will spend more time proof reading for errors than they do recording patient notes to begin with.
The only other negative feedback which seems to be circulating regarding Google’s latest news is regarding the security issues which could result. Privacy is of paramount importance when it comes to doctor patient confidentiality, and while there is always a risk of data leaks when information is stored by computers, the chances of leaks are intensified when alternative forms of data entry and technology are used to process and store the information. These will be among the bugs to be worked out before Google can guarantee the safety of their new product.
Whether it’s smooth sailing from beginning to end, there are bugs to work out, and the most important factor to consider when medicine is involved is accuracy. So long as Google works to fully test and remove bugs from the program before it is released to the public, it sounds like this new venture will prove extremely useful, and could be utilized in several other departments as well.
By the end of the summer, Google hopes to have the first round of studies complete, with phase two beginning in the new year. The recent job postings suggest that Google is seeking a Medical assist Product Manager, and other resources to support this latest project. Interested in jumping on board and giving Google a hand with its latest medical initiative? Check out their job postings for yourself.