Source: Deep Learning on Medium
Google AI for breast cancer detection beats Doctors.
The AI developed by Google detects breast cancer with higher accuracy
The goal of an AI is to create algorithms, robots, and technology capable of functioning in an intelligent (human-like) manner. One of Google’s AI tools has shown skills of detection of breast cancer which are similar if not better than those of a trained doctor.
In a study published in Nature (see here), An Artificial Intelligence (AI) developed by Google has improved the early detection process of breast cancer, reducing false negatives and false positives.
Breast cancer is the second leading cause of death from cancer in women. One of the key aspects of Breast cancer is early detection, which can considerably improve the outcomes of breast cancer. Women starting between the ages of 40 and 50 are often advised to go for a mammography screening. And although these measures contribute to early detection, there can still be cases of false negatives due to the difficulty of interpreting the screening images correctly.
A false negative is when the doctors erroneously diagnostic a negative when in reality there is breast cancer in the patient. This only makes breast cancer worst to treat as it develops and is found once the visibility of the tumor increases.
The AI tool for breast cancer detection explained:
The AI developed by Google analyses X-ray images know as mammograms and reduces the number of false negatives by 9.4 percent and false positives by 5.7 percent for women in the US. While for women in the UK it cuts 2.7 percent in false negatives and 1.2 percent in false positives.
Although this system outperformed doctors in most cases, there were other cases where doctors flagged breast cancer that was missed by the AI model.
This new AI Tool developed by Google is just one of many new technological developments within the field of Computer-vision. This field has seen major improvements in the latest years. Within the last 10 years, algorithms are much more capable of detecting objects and analyzing large visual datasets. The field of Deep learning (a branch of machine learning) brought what is known as Neural Networks as a way to analyze large and complex datasets. Furthermore, Convolutional Neural Networks (CNNs) is what has made a huge revolution in the field of Computer Vision.
In addition to Google AI being able to detect breast cancer more accurately than doctors, CNNs allow many other applications such as Face recognition, Surveillance, Biometrics, and Autonomous Driving.
Computer-Vision problems are mostly tasks such as localization, image classification, and object detection.
According to the research paper, this is the structure of the AI:
“The AI system consisted of an ensemble of three deep learning models, each operating on a different level of analysis (individual lesions, individual breasts, and the full case). Each model produces a cancer risk score between 0 and 1 for the entire mammography case. The final prediction of the system was the mean of the predictions from the three independent models.”
The problem of Breast cancer detection is a problem of image classification. Where breast cancer images must be classified as positive or negative. According to the research paper’s supplementary information which addresses the algorithms developed for this AI, CNNs were implemented.
What does this mean for the future (Conclusions)
It can result easy to think the age of the machine is coming, but in reality, these tools are merely tools. Ultimately, they are meant to be used by doctors (radiologists) to improve diagnostics. This contribution to science will definitely be saving lives once it is fully adopted by doctors.
These are without a doubt promising times for the field of computer- vision, which is a very important subset of AI these days.