Premium
Imaging in three‐dimensional conformal radiation therapy
Author(s) -
Mohan Radhe,
Rothenberg Lawrence,
Reinstien Lawrence,
Ling C. Clifton
Publication year - 1995
Publication title -
international journal of imaging systems and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.359
H-Index - 47
eISSN - 1098-1098
pISSN - 0899-9457
DOI - 10.1002/ima.1850060105
Subject(s) - radiation treatment planning , computer science , radiation therapy , magnetic resonance imaging , computer vision , artificial intelligence , visualization , merge (version control) , segmentation , conformal map , medical physics , medicine , radiology , mathematics , information retrieval , mathematical analysis
By and large, radiation therapy is a noninvasive method of the treatment of cancer requiring knowledge of the precise location and extent of the disease to be destroyed and the organs to be protected from radiation damage. Images have always played a central role in providing the requisite information for this mode of cancer treatment. Different types of images, such as computed tomography (CT); magnetic resonance imaging (MRI), positron emission tomographic (PET), simulator, etc., are used to varying degrees depending upon their relevance to radiation oncology as well as their accessibility. It is often necessary to merge data from various types of images. The availability of three‐dimensional information from tomographic images has allowed the introduction of three‐dimensional conformal radiation therapy (3DCRT) methods. Images are employed for diagnosing and establishing the extent of the disease, planning and delivery treatments, and evaluating the effectiveness of the treatment in controlling the disease and assessing the damage to normal tissues. Each image type has a unique informational content of importance to radiation oncology. To extract the maximum information from images, it is necessary to employ various image processing tools. These tools allow us to perform such functions as (1) image enhancement; (2) image correlation to register information from various images; (3) segmentation of images to extract the surface outlines of the tumor volume and normal anatomic structures; and (4) two‐ and three dimensional data visualization. One important aspect of planning radiation treatments is the computation of dose distribution in the patient for a proposed configuration of radiation beams. This step requires tracing rays in a three‐dimensional CT image data set to compute radiologic path lengths through the patient's body. Although images are employed to a great advantage in radiation oncology, many problems still remain to be solved. Of the various 3DCRT tasks, the outlining of contours of the volume of intended treatment and normal anatomy on images is highly labor‐intensive and fraught with uncertainty. In addition, the integration of data from various imaging modalities is difficult and error prone because of distortions inherent in imaging and also because of the motion, deformation, and displacement of patients and their internal anatomy. Investigations are in progress to find solutions to these problems.