Tech Research Wins Prestigious Competition
A Georgia Tech project has won an international competition that singles out the best operations research project by an organization.
Every year, the Franz Edelman competition recognizes outstanding examples of operations research (O.R.) projects that have transformed companies, entire industries and people's lives. O.R. uses advanced analytical methods to help make better decisions and is a disciplined way by which management can improve organizational performance in a wide variety of situations, in nearly any type of organization in the public or private sector.
This year's Franz Edelman finalists included Coca-Cola, The U.S. Coast Guard, Hewlett-Packard and Daimler-Chrysler. Past winners have included Motorola, Merrill Lynch, Canadian Pacific Railway and IBM.
Eva K. Lee, an associate professor at Georgia Tech's H. Milton Stewart School of Industrial and Systems Engineering, worked with Dr. Marco Zaider, head of Brachytherapy Physics at Memorial Sloan-Kettering Cancer Center (MSKCC), to devise sophisticated optimization modeling and computational techniques to implement an intra-operative 3D treatment planning system for brachytherapy (the placement of radioactive 'seeds' inside a tumor) that offers a safer and more reliable treatment.
Lee's optimization models and algorithms guide doctors toward the most effective dose provided by each radioactive seed, the shape of the organ being treated, the locations of tumor cells within the organ and critical structures for which radiation dose should be limited, the sensitivity of tissues to radiation, and the expected shrinkage of the organ after treatment. The system's goal is to provide consistent tumor-killing radiation doses to the tumor cells while limiting potentially damaging doses to nearby critical structures.
"The system can be used in real time," said Lee. "The patient can come in, the imaging is done and we can then do the planning and implantation right away. There is no delay between the imaging, planning and implantation of the seeds."
The real-time intra-operative planning system eliminates pre-operation simulation and post-implant imaging analysis. Based on the range of costs of these procedures, Lee estimates conservatively that their elimination nationwide could save on the order of $450 million a year for prostate cancer care alone.
By exposing healthy tissue to lower doses of radiation, the system reduces treatment complications by 40 percent to 65 percent and has a profound impact on the cost for interventions to manage side-effects. The procedure also uses significantly fewer seeds and needles compared to current best-practice procedures, according to Lee and Zaider, reducing procedure time, invasiveness and blood loss. As a result, patients experience less pain and have faster recoveries.
National distribution of this system will allow achievement of consistent treatment planning across different clinics, thus reducing the variability in the quality of treatment plans. The resulting plans limit urethral dose, decrease operator dependency and reduce the influence of the learning curve associated with prostate brachytherapy. These all have important consequences for the outcome of treated patients.
The system allows for dynamic dose correction, thus helping the training of clinicians and residents to develop effective and safe treatment plans.
The patented system is based on optimization techniques known as mixed integer programming. It was licensed to Prowess in 2004 and converted to a commercial product. Prowess added the new algorithms to treatment planning systems it already has in operation at more than 700 clinics in the United States.
Beyond prostate cancer therapy, the mixed-integer algorithms can also be used to optimize radioactive seed and external beam radiation treatment for a broad range of other cancers.
With support from the National Science Foundation, National Institutes of Health and Whitaker Foundation, Lee has also been working with medical specialists on improving treatments for breast, lung, cervical, brain and liver cancers.
"The cancer instances are really hard to solve, and our team has worked very hard in advancing the algorithmic frontier. Now we can use this in many different applications and it works very well for improving local tumor control," Lee said. "I feel really good about seeing this applied in the clinic to improve treatment to patients."