Top A.I. Algorithms In Medicine – The Medical Futurist


It’s inevitable. A.I is going to change healthcare as we know
it. But in this video, I want to show you a few
medical specialties where it’s already making a difference. Radiology In 2018, a deep-learning-based algorithm was
developed using more than 50 thousand normal chest images and almost 7 thousand scans with
active TB. The algorithm became so good that in performance
tests it easily beat radiologists. Of course, it has its shortcomings, but this
test shows that even the A.I of today can be a helpful second reader for physicians,
while the A.I. of tomorrow can bring screenings and precise diagnostics to even less developed
and rural areas where medical professionals are not available. Dermatology A.I is advancing elsewhere too. Researchers in Germany, the US and France
trained a deep learning neural network to identify skin cancer by feeding it with more
than 100,000 images of malignant melanomas and benign moles. After its training, they compared its performance
with 58 international dermatologists and the results were remarkable. While the dermatologists accurately detected
more than 86% of the melanomas, the neural network detected 95% of them. Oncology One of the biggest promises of A.I. is that
one day it could crack the code of individualized cancer diagnosis and treatment. And Watson, IBM’s very own A.I, is a powerful
tool that is mainly being used and tested in the field of Oncology. So far, dozens of hospitals have adopted this
technology and it’s been used in conjunction with medical judgment. While its promise is strong, it has not yet
been able to live up to the expectations in the fight against cancer. Cardiology Cardiovascular diseases are the number one
cause of death globally. For those affected, early detection is critical
for both management and treatment. And in the future, A.I. based predictions
could be a life-saver. Since studies have shown that markers of cardiovascular
disease can often manifest in the eye, scientists are using deep-learning methods to identify
risk factors such as age, gender, smoking status, and blood pressure only by looking
at the eye. These new studies still need to be validated
and repeated on more people before they could gain broader acceptance, but since retinal
images can be obtained quickly, cheaply and non-invasively, this will probably open new
horizons in healthcare. A.I. also has several limitations. Most of these studies have not been tested
under clinical circumstances and algorithms are only precise in a specific task while
clinical life is much more diverse. Nevertheless, what matters here is that A.I.
has amazing promise.

6 thoughts on “Top A.I. Algorithms In Medicine – The Medical Futurist

  1. Early detection does nothing. count mortality not OS. it proves that people live same amount with cancer treated and untreated. but untreated has better lives. It's a scam. Of course There are few examples like blood cancers where treatment sort of works but in breast cancer for example it fails miserably. Thats why we do not advice to treat older women

Leave a Reply

Your email address will not be published. Required fields are marked *