MD Anderson Cancer Center UTHealth
Graduate School of Biomedical Sciences

Student feature: Using computer technology to cure cancer patients around the world

March 06, 2019
Kelly Kisling, GSBS student

This article was written by GSBS student Kelly Kisling, the second place winner of the GSBS Student Science Writing Contest. Kisling is affiliated with the Program in Medical Physics and his advisor is Laurence Court, Ph.D.

Image of map showing cancer rates throughout the world

Hearing a diagnosis of cancer from your doctor is scary and leads to many questions. “What are the best treatment options for me to cure my cancer with minimal side effects? Which hospital and doctor should I see?” Now imagine you live in the developing world. There the question is more basic: “Can I even get treatment?” Too often, the answer is no. 

Many people are surprised to hear about cancer in the developing world. But more people die from cancer in developing countries than from HIV, tuberculosis, and malaria combined. Cancer rates are on the rise globally and they are rising most rapidly in less-developed countries. In these regions, millions of people get cancer every year. 

Despite the prevalence of cancer, developing countries still lack the bare minimum resources needed to treat and cure cancer, especially when it comes to radiation therapy, a cornerstone of cancer treatment. In developing countries, less than 50 percent of cancer patients have access to radiation therapy. The map on the right-hand page highlights this, with dark red indicating the least access. The few countries with adequate radiation therapy are in yellow. This map shows the enormity of the global need. In less-developed countries — where more than 80 percent of the world’s population lives — only 15 percent of the world’s radiation treatment facilities exist. 

The global shortage of radiation therapy is a multifaceted problem. One major issue is a shortage of the specialized staff needed to safely and effectively deliver radiation therapy. A recent study found that developing countries have only 58 percent of the trained radiation therapy staff needed for treating cancer. In total, the study estimates that more than 35,000 additional staffers are required to meet the current need. This deficit will only increase as cancer rates and populations rise. 

This may seem like an insurmountable deficit, but we think computers can help. Through automation, computers can reduce the amount of work needed for each patient, thus relieving the staff shortage. Our research group aims to create computer programs that automate some of the most time-consuming, complex work done by staff, allowing them to focus on other vital tasks and treat more patients. 

Image of CT scan of a chest tumor

One of the most time-consuming tasks in preparing a patient for radiation therapy is known as treatment planning. During treatment planning, highly trained staff design precise radiation treatments tailored to each patient for the best chance of curing his or her specific cancer. Staff use a CT scan to “see” inside a patient and view that patient’s specific anatomy and tumor. (See image above as an example.) To design the radiation treatment, they interact with specialized computers. The process is lengthy, involving hours of manual work. Computers are already helping, but why not let the computers do all the work? 

This is exactly what I am pursuing in my Ph.D. research: turning computers into skilled treatment planners. To do this, I am developing and testing software to automate every task that is usually done manually by a human. Where staff use a patient’s CT scan to see the tumor and surrounding organs, now a computer can, too. In the case of a breast cancer patient, for example, a computer can determine where the tumor is and precisely aim the radiation there. It can also recognize where the healthy organs are, such as the heart and lungs, and limit the radiation exposure in those areas. Computers are skilled at this because, like humans, they can see what the scan looks like. But where a person sees the scan as an image consisting of pixels with varying shades of gray, computers see the scan as pixels with different number values. Black is zero, white is 4095, and the shades of gray are all the numbers in between. 

In traditional treatment planning, staff manually identify the anatomy on a CT scan and then design a treatment to target the tumor and avoid the healthy organs. To do so, they aim the radiation beams and try to determine the optimal combination of treatment parameters. Computers excel at this—they can quickly search through thousands of combinations of parameters and find the one best solution creating an optimal, individualized treatment for each patient. 

It requires time and skill to tailor each treatment plan for each patient’s cancer. When there is a shortage of staff, it can mean fewer patients receive the life-saving radiation treatments that they need. Even when patients are able to receive treatments, not all clinics have the time to tailor the radiation to each patient’s individual cancer and anatomy. The difference in these treatments can mean an increase in the severity of side effects a patient experiences from radiation damage to healthy tissues. Or even worse, it means the difference in a successful cure if the tumor is not treated adequately. 

Our research group has harnessed the computer’s abilities to automate the entire treatment planning process. In my research, I have focused specifically on treatments of breast and cervical cancer because these are some of the most common cancers in developing countries. We have been working hand-in-hand with physicians in South Africa to hone our computational techniques and provide optimal treatments for their patients. In the next year, we will roll out our automated computer technology in hospitals in South Africa. There, we will assess this technology’s ability to reduce the treatment planning workload on staff and evaluate the impact on patients receiving optimized radiation treatment plans. 

Our long-term goal is to apply the techniques developed in our research to all types of cancer that are treated with radiation therapy globally. If successful, more people would have access to radiation therapy in places where they are struggling to provide it. As a result, we could ease human suffering from cancer around the world.

site var = gsbs