Researchers at google have come up with an Artificial intelligence(AI) algorithm which detects cancer cells in real-time.
It is applied to an Augmented reality(AR) enabled microscope.
According to a survey conducted in February 2018, cancer is the second leading cause of death worldwide accounting for around 8.8 million deaths worldwide. The numbers as well as the malignancy are both a matter of great concern. Thus, detection of cancer in its early stages has now gained far more attention than ever before.
There are big time players like IBM Watson among many others and now Google has also joined the party. Researchers at Google have come up with an Artificial intelligence (AI) and Augmented reality(AR) enabled model inspired by deep learning to detect cancer cells even at a very primitive stage.
This sounds interesting, but the big question is how does this work?
How it Works :
In its prior attempts, Google has spoken about how convolutional neural networks (CNNs) have helped to detect breast cancer metastases in lymph nodes with accuracy close to that of trained pathologists.
The system was first made to study a bunch of human tissues and the slides were then exposed to the lens of a light microscope fitted with an AR lens. The image seen through the microscope is then sent to a computer and this is where the deep learning model helps the computer to detect the cancer cells.
This might sound time consuming but all this happens in real-time and quite quickly as the results are stored in the computer. The hardware system consists of a modified light microscope equipped with an AR lens which enables the real-time image analysis.
(image source-google AI blog)
But the best part is that this system can be easily retrofitted in the existing models being used by doctors at a very minimal cost and using readily available components. The researchers tested the model for two different types of cancer- breast cancer and prostate cancer with convincing results. The AUC (area under the curve) factor for breast cancer was found out to be 0.98 and 0.96 for the latter.
The details of the system can be read on Google AI blog- https://ai.googleblog.com/2018/04/an-augmented-reality-microscope.html
So Let’s Summarize :
Google has had a hands-on experience with a similar system to detect breast cancer which was trained by convolutional neural networks. With this real time model, Google aims at widening its horizon and the results are quite promising. Sophisticated systems like these promise better diagnosis in a lesser amount of time.These AR microscopes offer a gamut of applications to be executed under a single cap.
“We believe that the ARM has potential for a large impact on global health, particularly for the diagnosis of infectious diseases, including tuberculosis and malaria, in the developing countries,” Google noted.