A new research partnership between the University of Rochester Medical Center (URMC) and a Rochester-based startup company led by former Eastman Kodak Co. executives will apply new image processing technology to scans from patients with the goal of significantly advancing medical tomography. URMC is partnering with TIES LLC, which was founded in 2003 by Vaseem Chengazi, M. Akram Sandhu, K. Bradley Paxton and URMC internist Bilal Ahmed. Sandhu and Paxton are both former Kodak executives. TIES (which stands for Tomographic Image Enhancement Systems) has patented a new image processing technology called "image surgery" that allows scientists and radiologists to selectively focus on a specific organ or region of the body to create clearer and more precise side by side images. The company will work with researchers in the URMC Department of Imaging Sciences to apply the technology to images from actual patients. TIES executives said the company's imaging technology potentially overcomes what have been significant limitations in medical tomography or 3-D imaging. Today's advanced imaging technologies such as gamma cameras, CT, MRI and PET scanners reconstruct images by converting a sequence of 2-D images which are captured by a receptor as it rotates around the patient into a 3-D image. While these technologies have provided doctors an invaluable view into the human body, the images often contain flaws. Radiologist Chengazi, an associate professor at the University of Rochester and TIES chief technology officer, said that scanners work very well to create images from stationary subjects. "However, the problem in the human world is that the body is dynamic and not stationary," he said. "The body moves and breathes, the heart beats, the bladder accumulates urine, and so on. Consequently, images of these areas of the body are often marked by artifacts or distortions." The clarity of a specific image often also depends upon the composition of objects that are nearby. Natural objects, such as organs and bones, and man-made objects, such as artificial hips or surgical clips, can interfere with the images of adjacent organs or tissue because they are in the way or, in the case of artificial objects such as prosthetics, are far more dense than the surrounding tissue and can throw off a scanner's sensitive instrumentation. These distortions can ultimately make it more difficult to spot smaller objects, such as tumors. "Radiologists have attempted to compensate for these problems by doing faster scans, and then once they have an image they filter it to decrease the artifact," said Chengazi. "But these artifacts are already 'baked' into the image by the process of reconstruction." The TIES technology, which Chengazi first began to develop as a PhD student at the University of London, overcomes these problems by segmenting the raw data before it is converted into an image. It lets radiologists exclude objects that are not of interest and heighten the resolution of the remaining target image. The technology could represent a fundamentally new direction in medical tomography, and has a potentially vast application in clinical care and biomedical research in fields such as cancer, musculoskeletal conditions and cardiovascular disease, Chengazi said. Under the research agreement, the raw data from the university's nuclear medicine gamma cameras will be run through TIES software so that scientists can compare it with the image generated using standard technologies. The technology is applicable to other types of scanners as well, he said. TIES plans to have its first image enhancement product for SPET (single-photon emission tomography) applications commercially available by next year. For more information, visit: www.urmc.rochester.edu