Using artificial intelligence to spot breast cancer

Reading Time: 4 minutes

Breast cancer is one of the modern world diseases that greatly influence women’s lives (0,5%-1% of the cases occur to men) with a high death rate – in 2020 out of 2.3 million women with breast cancer, 685 000 of them died. It is the most frequently diagnosed cancer among women worldwide. The key to combatting it in the early stage is by screening however the current methods as self-checking or mammalogy are seen as not accurate enough due to women’s unawareness of proper screening methods or the stigma of being physically tested by a doctor. With the rise of the importance of AI and women’s irritation, new possible solutions started to be introduced around the world. 

In 2016, Geetha Manjunath, a data analyst from India, lost her cousin who died out of breast cancer. Geetha blamed the mammology method as it didn’t discover breast cancer in time. She stated that mammographs have high difficulties in discovering cancer among young women and that there needs to be some improvements. 

“As I was working with multiple image modalities for other projects at Xerox, I chanced upon thermal imaging.” – Geetha Manjunath said.

She came up with the idea of combing thermal screening with AI. The whole process consists of thermal scanning the women and then comparing the results to the already existing pictures of patients with cancer with the help of AI to analyse loads of past cases and asses whether a pattern on a thermal picture could be a sign of early-stage breast cancer. Another advantage of Geetha’s project called Thermalytix is that is affordable, which is a major game-changer in a country like India where a lot of women are not affluent enough to be able to spend their savings on breast cancer screening. Another key factor, especially among young women is privacy. During the visit, nobody touches the patient and the doctor only sees a thermal image of the chest on the screen. It is also radiation-free and non-invasive so it doesn’t hurt and is not harmful to the body. So far, Geetha has screened over 75 000 women across India and plans to expand her innovative idea. Her mission is to make access to screening to every woman on the globe with the current progress of introducing her methods into countries like the Philippines and Kenya. She also started creating camps in poor areas of India to help poor women detect the disease before it’s too late, as they are the most vulnerable. 

Yet, researchers alarm that in the current stage the are some challenges to using AI in screening breast cancer. Firstly, AI databases are currently limited in comparison to the desired accuracy – AI bases its prediction on the pictures that are already in the database, therefore any new cancer pattern or any other abnormalities might be not registered by AI. Secondly, the electronic medical database is still not fully updated and well organised in many countries, even those well-developed as well. It leads to poor sources for AI to create the right prediction. There would also be a need for a global medical easy-access software system for all of the hospitals, which is extremely hard to organise on a worldwide and national scale. Moreover, there would have to be some social trust towards the doctors who would be making their decisions based on AI results which might nowadays rise some doubts and reluctance. Following that, hospitals would have to make special training for the doctors to implement using AI in screening breast cancer which takes time and money.

https://www.kpwashingtonresearch.org/news-and-events/blog/2020/artificial-intelligence-aiding-not-replacing-radiologists

The method is visionary and promising as it creates a comfortable way of detecting breast cancer conducted by a professional. It could minimise the scope of the problem drastically, helping all no matter their financial background. Yet, due to its early stage, there are some limitations as relatively small databases, the low popularity of medical data stored on the internet in many third-world countries and the question of AI deciding our health. Nonetheless, I believe there is a great future with this method and the right people and funds it can revolutionise the world.

Bibliography:

https://www.cureus.com/articles/106594-artificial-intelligence-in-breast-cancer-screening-and-diagnosis

https://www.cureus.com/articles/106594-artificial-intelligence-in-breast-cancer-screening-and-diagnosis

https://www.who.int/news-room/fact-sheets/detail/breast-cancer

https://www.bbc.com/news/av/stories-63755128

Tagged ,

One thought on “Using artificial intelligence to spot breast cancer

  1. 47576 says:

    This illness is one of the worst things a woman can imagine. Not only it is difficult to overcome but also in many cases breast must be removed, which deprives women of their femininity. The fact that this procedure exists and is safe for women makes me extremely pleased. I hope that thousands of people will be saved using this technique in the near future.

Leave a Reply