How Can AI Help Save Lives?

Artificial intelligence is not new, but recent years have been particularly interesting, bringing many exciting opportunities, developments, and experiences. Especially in the healthcare industry. 

Today, cognitive computing systems can diagnose heart diseases better than cardiologists do, smartphone apps can detect skin cancer with expert accuracy, and algorithms can diagnose eye diseases just like specialized doctors.

Major technology companies — Google, Microsoft, IBM — are investing in the development of AI for healthcare and research. The number of AI startup companies, oriented towards medical AI, is also increasing. The numbers are impressive!!

The total public and private sector investment in medical AI is expected to reach $6,6 billion by 2021. What is more, medical AI is forecasted to become a $10 billion market in the US by 2025.Accenture, a leading global professional services company, predicts that the top AI healthcare applications may result in annual savings of $150 billion by 2026.It is expected that by 2025, AI systems will be implemented in 90 percent of the US and 60 percent of the global hospitals and insurance companies, replacing as much as 80 percent of what doctors currently do. 

Excellent patient outcomes, reduced treatment costs, and elimination of unnecessary hospital procedures with easier hospital workflows and patient-centric treatment plans are the prime reasons for the wide adoption and successive growth of the AI market in the healthcare industry.

What is more, AI has the potential of solving some of the biggest problems of our time: including cancer treatment.Every year, more and more people are getting diagnosed with cancer.According to UK Cancer Research Magazine, over 17 million cancer cases were diagnosed across the globe throughout 2018. The same research suggests that there will be 27,5 million new cancer cases diagnosed each year by 2040.

When it comes to treating cancer, time is one of the most essential things. Between diagnosis and the first day of treatment, days and even weeks may pass by until doctors decide on the treatment plan, perform tests and gather all the necessary information.

“If a patient is diagnosed early, the chance of survival increases exponentially. It is known that more than 80 percent of breast, ovarian, prostate, lung cancer deaths are entirely preventable if detected early. Early detection and diagnosis are key to higher cancer survival rate,” said Brandon Suh, CEO of Lunit, whose chest imaging solution Lunit INSIGHT CXR is currently being used in hospitals in Mexico, Dubai, and South Korea.

This is where AI steps in: it can help to better predict patient’s diagnosis and determine the most effective treatment plan for their type of cancer, based on other patients with similar medical histories. 

With this vision in mind, the Knight Cancer Institute at Oregon Health & Science University (OHSU), in partnership with Intel, created a “collaborative cancer cloud“: once doctors have met a patient, within hours, they can diagnose him and begin treatment. To be more specific, the diagnosis-to-treatment time period is expected to be shortened to 24 hours in 2020. 

What is more, researchers from the Houston Methodist Research Institute in Texas have developed AI software that can accurately predict breast cancer risk. To compare, a manual review of 50 charts took 2 clinicians 50 to 70 hours, whereas the AI software reviewed 500 charts in just a few hours, saving the human doctors as much as 500 hours of their time. 

Stephen Wong, chair of the Department of Systems Medicine and Bioengineering at the Institute, noted that “This software intelligently reviews millions of records in a short amount of time, enabling us to determine breast cancer risk more efficiently using a patient’s mammogram. This has the potential to decrease unnecessary biopsies. An accurate review of this many charts would be practically impossible without AI.“

Although medical AI is still in its early stages, it has already achieved high levels of accuracy. 

For example, AI technology developed by the RIKEN Center for Advanced Intelligence Project (AIP) in Japan was able to identify features relevant to cancer prognosis that were not previously noted by pathologists, leading to higher accuracy of prostate cancer recurrence compared to pathologist-based diagnosis.

According to Resistance to Medical Artificial Intelligence study, published in the Journal 

of Consumer Research, in some cases, AI truly outperforms human healthcare providers. 

We have a few examples for you right here.

When the performance of IBM’s Watson was compared to human experts for 1000 cancer diagnoses, Watson found treatment options that doctors missed in 30 percent of the cases. 

When UK researchers compared the accuracy of 3 diagnoses made by doctors to those made by AI, doctors were found to be correct 77,5 percent of the time, whereas AI reached an accuracy rate of 90,2 percent.

Doctor Jack Kreindler, a physician, psychologist ad serial technology entrepreneur, founder, and director of The Centre for Health and Human Performance says that “I would sooner today trust computer scientists and data scientists to tell me how to treat a really complex system like cancer than my fellow oncologists. I would not have said that two to three years ago.“

However, Maciej Mazurowski, associate professor of radiology and electrical computer engineering at Duke University noted that “The final question, even if we can show that artificial intelligence work as well as humans, will be whether and to what extent it will be adopted into the healthcare system. It’s not just whether it works.”

Today, medical AI is being used for a range of healthcare and research purposes: detection of diseases, management of chronic conditions, delivery of health services and even drug discovery. 

However, although machine learning is a powerful tool, it requires a huge amount of data. 

Up until now, one of the main things, causing slower progress, was that researchers struggled with finding enough data, all in the same place and all in the same format, which would be accessible to everyone.

You see, as data sharing has always happened between colleagues and experts who know of each other’s work, it has gained traction on an international level just in recent years. 

Dr. Robert Miller, medical director of CancerLinq, which gathers anonymized patient data from electronic health records nationwide, noted that “There are more than 100 types of cancers with different genetic variations. Data sharing and AI have the potential to help us further personalize care for each individual patient. For example, if a physician is treating a patient with rare cancer, he or she can examine the outcomes of patients across the country with same cancer and similar characteristics to help choose the right therapy for the right patient at the right time.“

As medical AI is gaining more and more momentum all over the world, one question stands out: what about the future?

Experts in the healthcare industry seem to have similar opinions.

“AI is never going to fully replace physicians, but researchers have made impressive strides in developing AI that specializes in particular tasks, like reading images from CT scans or pathology slides,“ says Ryan Schoenfeld, vice president of scientific research at the Mark Foundation for Cancer Research, a New York City-based philanthropy. “I expect we will continue to 

see further advances in this area.“

“All of this AI stuff is really new. It’s been out for about five years and hasn’t taken hold in many places. It has a lot of promise: I think the way we treat cancer in 10 years will be a lot different than we do today,“ added Dr. John Vu, a board-certified medical oncologist and director of clinical informatics at Baptist MD Anderson Cancer Center in Jacksonville, Florida.

Although AI is not the cure for cancer that the world has been waiting for, it sure is a phenomenal way to improve treatment, until the cure is discovered. And who knows, maybe one day AI will also help scientists to discover the cure itself.