The discovery, announced in late January by a team of researchers at Harvard’s Massachusetts General Hospital and the Massachusetts Institute of Technology, is part of a growing medical trend of using algorithms to predict everything from breast and prostate cancer to the likelihood of tumor recurrence. Although research is growing, scientists say more trials are needed before these products can be fully rolled out in clinical settings.
The instrument is called a sibyl, named after a prophetess in ancient Greek literature. It’s a deep-learning model, meaning computers parse through huge data sets to identify and classify patterns. Sybil was trained for six years to scan the lungs of patients in the United States and Taiwan, the researchers said.
The results of the study showed that Sybil scored scientifically as “good” and “strong” in predicting lung cancer over six years. This was in line with its one-year prediction rates, the study scientists noted.
Lung cancer is “the biggest cancer killer because it’s relatively common and relatively difficult to treat,” said Florian Fintelmann, an interventional radiologist at Massachusetts General Cancer Center and study co-author. “If you detect lung cancer early, the long-term outcome is significantly better.”
Cancer is the second leading cause of death worldwide, and as advances in artificial intelligence software and computing power increase, it has become a ripe field for researchers to harness the technology in hopes of helping doctors diagnose it.
Researchers are using artificial intelligence to track the progression of prostate cancer, breast cancer or tumor regrowth after treatment.
Many techniques involve analytics Medical scans, data sets or large troves of images, then feeding them into complex artificial intelligence software. From there, computers are trained to detect images of tumors or other abnormalities that researchers claim can be more accurate and faster than the human eye.
In recent years, there has been an increase in new treatments to fight lung cancer, study researchers said, but many patients still die from the disease due to complications.
The elderly and poor cannot afford screenings because of limited federal funding. Many patients diagnosed with lung cancer have either never smoked or are ex-smokers who quit more than 15 years ago, MIT researchers said, making them ineligible for screening in the United States.
For those who can be screened, the most common method is to use a low-dose computed tomography scan, called LDCT. The researchers created Sybil to turbocharge the screening process, allowing the software to analyze LDCT images without the assistance of radiologists to predict cancer risk up to six years in advance.
But building Sybil was challenging, the study authors said. Peter Michael, a researcher and associate at MIT’s Jameel Clinic and its Computer Science and Artificial Intelligence Laboratory, described it as “trying to find a needle in a haystack.”
mostly There were no obvious signs of cancer in the imaging data to train Sybil, as early-stage lung cancer occurs in small areas of the lung and can be difficult to see with the naked eye. to spot. To make sure the software could assess cancer risk, the study team “labeled hundreds of CT scans with visible cancerous tumors” and fed them into Sybil before releasing the software with limited signs of cancer on CT scans, the researchers said.
The team used data sets from the National Lung Screening Trial, Massachusetts General Hospital and Chang Gung Memorial Hospital in Taiwan. According to the study, some statistics from white people are highly skewed.
Medical experts warn that more study is needed before cancer software can be put into clinical use, according to government scientists and research studies.
Researchers from Harvard and the Netherlands have said that skills for translating information generated by AI algorithms remain at an “early stage”. Moreover, the benefits that AI can provide to medicine are currently quite narrow. Even with these detection tools, doctors still need to diagnose, design treatment plans and manage overall care.
Other medical experts pointed out the need for more tests to see how well the software works on different patients using different scanners and devices. More work needs to be done to show that the software actually benefits people, either by helping them live longer, prevent cancer or save time and money. How the algorithm works should be transparent, not a “black box,” they said.
The MIT researchers said they will continue their work.
“An exciting next step in research will be testing CIBIL in people at risk of lung cancer who have never smoked or quit smoking decades ago,” said Lesia Sequist, director of Massachusetts General Hospital’s Center for Early Cancer Detection.