About DBIS   | Story archive   | Contact DBIS  | DBIS home

Saving Lives: Detecting Lung Cancer Faster

Radiologists Use New Technology to Spot Lung Cancer More Easily, Earlier

January 1, 2011

Radiologists are using a new technology to find lung cancer earlier. Computer software can pinpoint areas of concern on a traditional X-ray. These areas can then be analyzed more closely by a doctor. Traditionally, lung cancer can evade detection because the ribs and blood vessels can obstruct the view in a chest X-ray. Studies have shown that the new technology can detect up to 50 percent of lung cancers that were missed in initial imaging, and early detection means a higher survival rate.

read the full story...

Science Insider

ABOUT LUNG CANCER: The lungs are sponge-like organs found in the chest cavity. When we breathe, air goes into the lung through the windpipe (trachea), which is divided into smaller branches leading to tiny air sacs called alveoli. These contain tiny blood vessels that absorb oxygen from the inhaled air into the bloodstream and release carbon dioxide. Most lung cancers start in the lining of the smaller branches, developing over several years. The cancer cells produce chemicals that cause new blood vessels to form, which feed the cells until they form a tumor. Cells can break away from the original tumor and spread to other parts of the body, a process called metastasis. About 90 percent of lung cancers develop in smokers or former smokers. Symptoms of lung cancer include coughing up blood, shortness of breath or wheezing; chest pain; repeated chest infections, asthma, or fluid around the lung; and swelling in the neck or face.

MACHINE LEARNING: A computer that learns to recognize lung cancer or a car that interprets the behavior of its driver and can offer alerts about when that driver seems to be impaired is an example of machine learning. Typically, this concept refers to a computer that can gather and interpret data via an algorithm and then improve the performance over time. Many consider machine learning an attempt to automate the scientific method. If a well-programmed machine encounters enough examples it will begin to recognize certain patterns and be able to identify or diagnose problems. Machine learning algorithms are used in everything from computer vision to search engines, and credit card fraud detection to handwriting recognition.

Video help

Latest stories

  • A Satellite Named Violet and a Student Named Amanda
  • Behind the Scenes with the K-Team
  • Deep Space Discoveries
  • Dogs Fighting Cancer
  • Earthquake! What's Your Risk

More information on this story

To Go Inside This Science:
Sharon Boston
Media Relations Manager
University of Maryland Medical Center
and University of Maryland School of Medicine
Baltimore, MD
sboston@umm.edu


© 2011 American Institute of Physics