Researcher Q&A with Janne Nappi, PhD

Published on November 22, 2011

Updated on February 13, 2018

Janne Nappi, PhD

The Foundation is proud to highlight the colon cancer research of Janne Nappi, PhD, an instructor of radiology at Massachusetts General Hospital. Dr. Nappi received a two-year grant from the Foundation in Fall 2008 to improve the accuracy of computed tomographic colonography (CTC) by developing sophisticated computer algorithms to analyze the surface of the colon and indicate the locations of hard-to-detect flat growths that might develop into cancer.

1.  What led you to the field of colon cancer research?

Although colon cancer is one of the leading causes of cancer deaths, it would be largely preventable if its benign precursor lesions were detected and removed early enough. In 2000, computed tomographic colonography (CTC) was emerging as a better way for colon screening. The idea was to detect precancerous polyps by reviewing virtual three-dimensional images of the colon. However, there were concerns if it would be practical for radiologists to interpret large amounts of CT images. At that time, I had already been working several years on computer-assisted diagnosis for mammography, and I was looking for a new project. I found CTC and its three-dimensionality intriguing. If a computer could be used to detect polyps automatically from CTC images and point them out to radiologists, this could increase the detection sensitivity and consistency of CTC interpretations.

2.  Tell us about your research to improve the accuracy of computed tomographic colonography.

Today, trained radiologists and computers can detect the significant polyps at high sensitivity from CTC data. However, recent studies have revealed that colon cancer can develop also from so-called flat lesions. Flat lesions are less common than polyps, but they have higher potential for cancer. They are also more difficult to detect, because they tend to imitate normal colon. Current computer-aided detection systems have been designed to detect obvious round polyps protruding from the colon’s surface, and therefore they are not so helpful in detecting flat lesions. However, the detection of flat lesions is precisely the kind of thing where radiologists would need assistance.

Prevent Cancer Foundation provided us with the pilot funding to develop new image analysis and pattern recognition algorithms for advancing computer-aided detection to detect not only polyps but also flat lesions in CTC. We are also implementing an observer study to evaluate the improvement in radiologists’ detection performance for flat lesions when aided by such a next-generation computer-aided detection system.

3. What impact could your findings have on preventing other cancers beyond colon cancer?

Our computational methods and ideas are not limited to the colon, but they could also be used to improve the accuracy of image-based diagnosis of other cancers, such as lung cancer. Also, the observer studies provide valuable insight into how human experts are using or not using the computer output. This will help in maximizing the benefits of computer-aided detection in medical imaging in general.

4. Why is it important to fund research in the field of cancer prevention and early detection?

Cancer is still a leading cause of deaths worldwide. However, after several years of research on various aspects of cancer, cancer statistics are showing that we can change this. In particular, we know that prevention and early detection reduce cancer mortality most effectively. At its later stages and particularly after metastasis, cancer becomes increasingly difficult and costly to treat, and the survival rates decrease dramatically. Research in cancer prevention and early detection increases our knowledge and options regarding cancer, thereby ultimately providing us with longer, more productive, and happier lives.

No Comments

Leave a Comment

Your email address will not be published. Required fields are marked *

I accept the Privacy Policy

Sign up to get the latest about cancer prevention and early detection directly in your inbox.