Microsoft Shows Searches Can Boost Early Detection of Lung Cancer
Microsoft Corp. researchers want to give patients and doctors a new tool in the quest to find cancers earlier: web searches.
Lung cancer can be detected a year prior to current methods of diagnosis in more than one-third of cases by analyzing a patient's internet searches for symptoms and demographic data that put them at higher risk, according to research from Microsoft published Thursday in the journal JAMA Oncology. The study shows it's possible to use search data to give patients or doctors enough reason to seek cancer screenings earlier, improving the prospects for treatment for lung cancer, which is the leading cause of cancer deaths worldwide.
To train their algorithms, researchers Ryen White and Eric Horvitz scanned anonymous queries in Bing, the company's search engine. They took searchers who had asked Bing something that indicated a recent lung cancer diagnosis, such as questions about specific treatments or the phrase "I was just diagnosed with lung cancer."
Then they went back over the user's previous searches to see if there were other queries that might have indicated the possibility of cancer prior to diagnosis. They looked for searches such as those related to symptoms, including bronchitis, chest pain and blood in sputum. The researchers reviewed other risk factors such as gender, age, race and whether searchers lived in areas with high levels of asbestos and radon, both of which increase the risk of lung cancer. And they looked for indications the user was a smoker, such as people searching for smoking cessation products like Nicorette gum.
How effective this method can be depends on how many false positives -- people who don't end up having cancer but are told they may -- you are willing to tolerate, the researchers said. More false positives also mean catching more cases early. With one false positive in 1,000, 39 percent of cases can be caught a year earlier, according to the study. Dropping to one false positive per 100,000 still could allow researchers to catch 3 percent of cases a year earlier, Horvitz said. The company published similar research on pancreatic cancer in June.
Microsoft's research arm is working on a variety of different technological approaches to improving cancer care and diagnosis. In September, the company published research on using machine learning algorithms to better tailor customized combinations of cancer drugs to a particular tumor and an effort to one day allow scientists to program cells to fight disease.
"These are diseases that present with very general symptoms and so they often take some time for a physician to route a patient with those symptoms to a lung cancer diagnosis or a pancreatic cancer diagnosis," said Trever Bivona, an associate professor of hematology and oncology at the University of California, San Francisco. "While this would need to be rigorously tested I could envision how this could be a useful adjunct to the current paradigm of diagnosis."
In many cases, these cancers are found late, when treatment is unlikely to be successful, or even an option. That's why the team is searching for new ways to increase early detection, said Horvitz, who has a medical degree and is a long-time artificial intelligence researcher.
"I have very stark memories of my first lung cancer patient as a medical student. I remember saying is this operable? He had just come in, but the lung cancer was already in his brain," Horvitz said. "The prospect we could save patients' lives' with one or more methods of pre-screening is intriguing."
"This is a very powerful tool," said Joe Gray, a professor of biomedical engineering at Oregon Health & Science University. "We are interested in identifying cohorts of patients that might be at risk of cancer. One of the difficulties we have in the general community is if we really want to focus on developing better screening, how do you find them?"
The work may also be able to identify new, previously unknown risk factors for lung cancer and correlations with environmental issues, Horvitz said. The search data showed a possible association between lung cancer and older homes, which may not have been constructed to avoid the build-up of radon gas, and with frequent air travel, he said. It would be worth further study on whether these factors do increase the risk of developing lung cancer, Horvitz added.
Now that the study has determined the web search approach can work, researchers and clinicians need to figure out ways to use it, with patient consent. One option, said Horvitz, is to offer users apps that provide patient monitoring and could be used to alert them to red flags.
Horvitz and Gray are also trying to design a study that will get cancer patients to volunteer to have their medical records and search logs examined to find even stronger causality between cancer symptoms, risks and diagnosis.