Radiation therapy for cancer is one of the first areas in which computers were widely used in support of clinical practice, dating back at least twenty years. The initial application, still the major one today, was the calculation of radiation dose in the tumor region and in other locations of the patient's body. However, knowledge about clinical implications of radiation treatment has been difficult to systematize in a form suitable for computer processing. The number and range of treatment options is large. Treatment optimization based on dose computations with variation of parameters can only handle a few of the possible options. As an alternative approach to systematizing clinical knowledge of radiation effects, this project will develop an expert system for planning radiation therapy. The problem of planning radiation therapy presents challenges in the design of expert systems. Efficiency and flexibility are important because of the large problem space. In addition, some of the knowledge of radiation treatment planning is most naturally expressed in the form of constraints, rather than production rules. Most important, decision-making normally relies on a treatment simulation system to provide dose and geometric information about proposed treatment strategies. To meet these requirements this project will develop and investigate the properties of an expert system that integrates with a treatmet simulation system. The two systems will be coupled using a message passing technique previously developed for the simulation system internal operation. The expert system will use both frames and production rules as appropriate tothe problem domain. Constraint mechanisms will be introduced which allow constraints to be strengthened or weakened as the system preceeds toward a solution. The simulation system will be coupled to the expert system as a means of checking constraint satisfaction and graphically displaying the solution(s). The integration of a simulation system with appropriate knowledge representation techniques should provide new insight into methods of constraint satisfaction, and also extend understanding of the design and behavior of hybrid systems.
| Ketting, C H; Austin-Seymour, M; Kalet, I et al. (1997) Consistency of three-dimensional planning target volumes across physicians and institutions. Int J Radiat Oncol Biol Phys 37:445-53 |
| Ketting, C H; Austin-Seymour, M; Kalet, I et al. (1997) Automated planning target volume generation: an evaluation pitting a computer-based tool against human experts. Int J Radiat Oncol Biol Phys 37:697-704 |
| Kalet, I J; Jacky, J P; Risler, R et al. (1997) Integration of radiotherapy planning systems and radiotherapy treatment equipment: 11 years experience. Int J Radiat Oncol Biol Phys 38:213-21 |
| Kalet, I J; Austin-Seymour, M M (1997) The use of medical images in planning and delivery of radiation therapy. J Am Med Inform Assoc 4:327-39 |
| Kalet, I J; Jacky, J P; Austin-Seymour, M M et al. (1996) Prism: a new approach to radiotherapy planning software. Int J Radiat Oncol Biol Phys 36:451-61 |
| Ketting, C H; Austin-Seymour, M M; Kalet, I J et al. (1995) Evaluation of an expert system producing geometric solids as output. Proc Annu Symp Comput Appl Med Care :683-7 |
| Austin-Seymour, M; Kalet, I; McDonald, J et al. (1995) Three dimensional planning target volumes: a model and a software tool. Int J Radiat Oncol Biol Phys 33:1073-80 |
| Kalet, I J; Paluszynski, W (1990) Knowledge-based computer systems for radiotherapy planning. Am J Clin Oncol 13:344-51 |
| Kalet, I J; Jacky, J P (1988) Radiation therapy treatment planning using concurrent programming. Comput Methods Programs Biomed 26:115-22 |
| Kalet, I J (1987) Multiview three-dimensional treatment planning. Front Radiat Ther Oncol 21:33-43 |
Showing the most recent 10 out of 11 publications
