Articles from "Microelectronics News"

Fraunhofer Group for Microelectronics

< Previous article

Detecting bowel cancer early

Bowel cancer is the second-most common type of fatal cancer in Germany and the entire Western world. The chances of treating it successfully, however, are between 90 and 100 %, as long as the malignant tissue changes are detected early. Researchers at Fraunhofer IIS have now devised a method that could improve the diagnosis of bowel cancer considerably.


Regular preventive screening can help you to gain foresight into your health. Photo: / Rainer Sturm
Regular preventive screening can help you to gain foresight into your health. Photo: / Rainer Sturm

Most bowel carcinomas develop from polyps, which are macroscopically visible protuberances in the mucous membrane. Health insurance companies recommend regular preventive check-ups in the form of colonoscopies once patients reach a certain age. All the same, 15–25 % of malignant changes are still missed or are discovered too late – often these are small, sunken, or flat tissue changes known as lesions.


Detecting polyps automatically with digital image analysis


In the future, standardized, computeraided detection of polyps based on endoscopic image sequences could help gastroenterologists looking for and interpreting polyps and tumors in the bowel, improving the sensitivity of preventive colonoscopies. In a pilot study, researchers at the Fraunhofer Institute for Integrated Circuits IIS and at Klinikum Bayreuth hospital examined how image processing methods could be used for early detection of tissue changes. As part of the study, the scientists drew up high-definition bowel images from 115 patients. In each case, image sequences with clearly visible polyps were taken, saved and validated histologically. It was seen that marked polyps can be well distinguished from surrounding tissue using texture analysis as part of imaging processes.


For improved certainty and quality


Due to reduced staff, financial, and time resources, gastroenterologists are under a lot of time pressure. A computer system for detecting neoplasms or neoplasm-like areas, connected to an endoscopy system, could help them to improve diagnostic certainty and quality given the number of preventive check-ups required. This would contribute to better check-up quality in both clinics and doctors’ practices. To encourage further research, the Fraunhofer IIS team received the Boston- Scientific Innovation Award 2011 on March 19 of this year for its contribution “Image-based differentiation between colorectal polyps and background tissue using color textural analysis.” The award was presented at the annual general meeting of the Deutsche Gesellschaft für Endoskopie und Bildgebende Verfahren (DGE-BV) in Munich.




Dr. Christian Münzenmayer

Phone +49 9131 776-7310

Dr. Thomas Wittenberg

Phone +49 9131 776-7330

Fraunhofer Institute for Integrated Circuits IIS

Am Wolfsmantel 33

91058 Erlangen