When it comes to combating cancer, almost all technologies can be explored as potential solutions. Thanks to a team of engineers, physicians could be provided with a new Artificial Intelligence (AI) program to help them and their patients fight the deadly disease.
Researchers from the Center for Research in Computer Vision at the University of Central Florida have developed an artificial intelligence impeccable "vision" for the detection of small tumors in CT scans. Most human radiologists have a success rate of 65 percent in identifying smaller tumors on a scan, the team noted. This new AI system thrusts that percentage down to  95 percent accuracy and gives radiologists a second set of keen eyes to use for them.
AI Training to Detect Tumors
As with many AI systems, one had to learn what exactly a tumor looked like and how these tumors change in different parts of the body.
"We used the brain as a model to build our system," said Rodney LaLonde, a graduate student. (LaLonde is not only a computer engineer, but also captain of the UCF ice hockey team .) LaLonde was assisted by the technical assistant Professor Ulas Bagci.
"You know how to strengthen and learn the connections between neurons in the brain during development? We used this design to help our system understand how to look for patterns in CT scans and themselves teaches you how to find those tiny tumors, "LaLonde continued.
The engineers resorted to common face recognition algorithms like the ones in the iPhone X. Most of these software constantly scan faces to find a particular pattern, and eventually find a match in those patterns.
In this case, the team worked with the National Institute of Health and the Mayo Clinic to scan thousands of CT scans into their AI software. Each CT scan taught the AI the tumor size and shape
Improving Opportunities Through Improvement
"I believe this will have a very big impact," Bagci said. "Bagci has worked extensively in biomedical imaging and machine learning, hoping to improve the state of clinical imaging, and before joining UCF, was head of the laboratory at NIH's Center for Infectious Diseases Imaging Lab in Radiology and Imaging Sciences
Number one cancer killer in the United States and when detected at late stages, the survival rate is only 1
Lung cancer is in fact the largest cancer killer in the United States, which claims nearly 800,000 lives each year.The AI could one day be used for detecting tumors in other parts of the body: colon and rectal cancer the second highest killer and has one of the highest rates in stage 4.
The pressure to improve the AI is something that motivates that Research Team Continues: Lung cancer, prostate, colon and breast cancers account for 45 percent of all cancer deaths in the US . Engineers want to reach AI in a hospital and quickly  "I think we're all came here because we wanted to use our passion for engineering to save lives and have a big impact, "said LaLonde.
Via: University of Central Florida