How A.I. Is Learning To Fight Cancer
In 2013, Hisashi Kambe, a computer systems engineer, developed a unique system that used artificial intelligence (AI) to identify baked goods at a Japanese bakery. The BakeryScan system, for instance, could distinguish a cake from a croissant and a Danish from a turnover.
BakeryScan was featured on television and a doctor at the Louis Pasteur Center for Medical Research in Kyoto, Japan, watched the segment. While the program focused on baked goods, the doctor thought that cancer cells, under a microscope, sort of looked like bread.
The doctor contacted BRAIN, the company Kambe had formed and asked if he could develop a type of BakeryScan for pathologists. Kambe made changes to the technology to create, Cyto-AISCAN, an AI system that can detect cancer instead of donuts.
For years, researchers have been working on a cure for cancer, the second leading cause of death, after heart disease, in the United States, according to the Centers for Disease Control and Prevention (CDC).
Since the cancer survival rate is higher when the disease is diagnosed early, scientists and medical professionals are hoping to save more lives by using AI and machine learning (ML).
Artificial intelligence involves a computer performing tasks commonly associated with human intelligence. ML is a branch of AI in which computer programs can access massive amounts of data and learn how to use it for themselves instead of being programmed by a human to use the data.
According to the World Economic Forum, there are three ways in which AI and ML can help in the fight against cancer:
- AI analysis of images can accurately detect cancer.
- AI and ML can use a patient’s electronic medical history and come up with different treatment plans.
- AI can be used in developing new drugs to treat cancer.
AI and ML are welcomed by medical professionals in overburdened health care systems preparing for more cancer cases over the next two decades. According to estimates from the World Health Organization’s (WHO) agency for cancer research, the global cancer burden is expected to climb 47 percent, from 19.3 million new cancer cases in 2020 to 28.4 million new cases in 2040.
Opening A New Window Into Cancer Research
The U.S.-based company that made “PowerPoint” and “Windows” international brand names is now working on projects in the healthcare industry. Microsoft launched Healthcare NExT, an initiative that combines AI, health research, technology, and the expertise of medical professionals to find a cure for cancer or effective treatment for cancer.
“It’s a big challenge,” Peter Lee, Corporate Vice President at Microsoft Research NExT said in a blog post about the company’s initiative. “But we believe technology—specifically the cloud, AI and collaboration, and business optimization tools—will be central to health care transformation.”
Microsoft researchers, located in the company’s biological computation labs in Cambridge, United Kingdom (U.K.), are developing technology that could reprogram cells to fight cancer and other diseases.
“We are trying to change the way research is done on a daily basis in biology,” Jasmin Fisher said in a press release about the project. Fisher is a biologist who works in the programming principles and tools group in Microsoft’s U.K. lab.
The project calls for developing a computer-type system made from DNA that would live inside human cells and identify potential cancer cells. Once the system detects the cancerous invaders, the computer would restart and clean out the dangerous cells.
Another project involves combining ML with computer vision to give radiologists a more detailed understanding of how their patient’s tumors are progressing.
Joining Together to Fight Cancer
Time is of the essence when it comes to diagnosing and treating cancer. So, AI and ML are bringing computer scientists, engineers, and medical professionals together to diagnose cancer and find the best treatment as soon as possible.
As an example, researchers at the Institute of Cancer Research London and the University of Edinburgh are testing a new treatment developed by ML that can predict cancer growth. With this prediction, doctors can design the most effective treatment for their patients.
Similarly, scientists in Finland developed an ML model that accurately predicts how combining different cancer drugs can kill different types of cancer cells, according to the study published in the journal, Nature Communications. The ML model is designed to help doctors find the best drug combinations that can selectively kill cancer cells with specific genetic or functional makeup.
Researchers at Aalto University, the University of Helsinki, and the University of Turku trained the AI model with a large set of data from previous studies that explored the association between drugs and cancer cells. Interestingly, the model found associations between drugs and cancer cells that had not been noticed before.
In the United States, Intel, the world’s largest manufacturer of PC microprocessors, partnered with The Knight Cancer Institute at Oregon Health & Science University to create a “Collaborative Cancer Cloud.”
The collaborative project of the California-based Intel, and the university in Portland, Oregon, uses ML and a set of molecular and imaging data to:
- Diagnose cancer.
- Determine how genes interact to drive disease in individual patients.
- Predict the best possible treatment for patients.
The end goal of the project is to empower researchers and doctors to give patients a diagnosis within 24 hours and create a targeted treatment plan, said Eric Dishman, vice president, and general manager of Intel’s Health and Life Sciences Group.
Although critical patient data will be shared among medical and tech experts, patients’ privacy will be maintained, said Dr. Brain Druker, director of the OSHU Knight Cancer Institute.
“By securely sharing clinical and research data among institutions, while maintaining patient privacy, our goal is to turn a process that’s agonizing and uncertain for countless patients into a highly tailored, one-day diagnosis and treatment recommendation,” Druker said.
Meanwhile, Kambe said the cancer detector, Cyto-AISCAN, is now being tested in two hospitals in Kobe and Kyoto, Japan. So far, the detector has a 99 percent accuracy rate, Kambe said.
Links:
https://www.newyorker.com/tech/annals-of-technology/the-pastry-ai-that-learned-to-fight-cancer
https://www.cdc.gov/chronicdisease/resources/publications/factsheets/cancer.htm
https://www.weforum.org/agenda/2021/02/cancer-treatment-ai-machine-learning/
https://acsjournals.onlinelibrary.wiley.com/doi/full/10.3322/caac.21660
https://futurism.com/neoscope/microsoft-ai-machine-learning-discover-cure-cancer
https://www.forbes.com/sites/forbestechcouncil/2019/04/22/ai-in-the-fight-against-cancer-and-other-diseases/?sh=3d453cab4872
https://www.cancer.gov/research/areas/diagnosis/artificial-intelligence
https://www.thedailybeast.com/in-2020-could-artificial-intelligence-help-cure-cancer
