Merging computer science and robotic technology to modernize processing of radioisotopes

In a new effort to modernize and improve medical isotope production, scientists at the U.S. Department of Energy's (DOE) Argonne National Laboratory have embarked on a project that harnesses the power of computer-driven robotic technology. The project aims to increase safety and reproducibility, while driving down cost.

Thanks to new funding the technologies used to process diagnostic and therapeutic radioisotopes used in cancer treatment will be modernized. They will be replacing systems based on technologies used since the 1940s.

This project, led by Argonne, is multidisciplinary. It includes scientists and engineers from across the Lab as well as faculty and students from four colleges and universities.

The DOE's Isotope Program helps to fund and direct the research required to bring important medical isotopes from concept to reliable delivery to end users for pre-clinical and clinical trials.

For oncologists on the front lines of the battle against cancer, these medical radioisotopes represent an indispensable weapon. Their advantage lies in the ability to be selectively targeted against malignant tumors with reduced collateral damage to surrounding healthy tissue.

Preliminary research with some radioisotopes has indicated promising advances in cancer treatment. As a result, in many cases the demand for these emerging isotopes is far beyond the available supply. This limits the rate of progress in developing these advanced cancer treatments.

Research and production of medical isotopes are currently conducted at national laboratories and university-based accelerators. A handful of private companies are also participating.

Scientists painstakingly create these isotopes by irradiating targets composed of enriched stable isotopes. The desired radioisotopes are created by a nuclear transmutation. However, a significant challenge arises once these radioisotopes are produced in the targets. The very small amounts of the useful radioisotopes produced need to be chemically separated from the bulk mass of the target material and impurities.

The radiochemical separation can take one of two routes. One is bench-top, manual processing in a "glove box," but this is fraught with radiation exposure to the researchers involved, limiting the throughput of the production batches. The other route involves processing in heavily shielded, dedicated "hot cells." These cells still use 1940s-era mechanical manipulators. They are quite maintenance-intensive and expensive and have limited mechanical capability.

The tele-operated robotic system that Argonne plans to develop will introduce a new kind of processing station for radioisotopes. It will feature a "hot box" that optimally combines elements of a hot cell and glove box, said Argonne nuclear physicist Jerry Nolen.

An immediate goal is to carry out the basic research required to capitalize on recent developments in remote or tele-operation. These include new robotics technologies, 3D vision technologies, 3D dynamic modeling with high-speed computing, and cost-effective, capable manipulators that enable complex remote processing.

"Very flexible manipulator capabilities are required to carry out the intricate operations of the radio-chemical separations," said Millicent Firestone, Argonne deputy associate laboratory director and senior science advisor for Physical Sciences and Engineering.

"Just by gaining the ability to do the manipulation of the sample from across the room means that we can safely handle samples up to 10 times as radioactive without requiring the use of hot cells. This dramatically increases our ability to produce these valuable and necessary isotopes," Nolen said.

The new robotic hot box will operate remotely through augmented reality. The user is seated away from the radioactive sample and uses 3D computer vision and immersive display technologies to visualize the hot box. Also employed will be advanced software to remotely control specially designed robotic components located in the hot box.

Argonne computer scientist Nicola Ferrier said that the technology for the augmented reality is likely to be based on the Omniverse technology developed by the U.S. company Nvidia, paired with optimization of models for real-time interaction. An off-the-shelf headset will give the remote operator a 3D virtual reality view of the interior of the hot box. Omniverse enables the development of 3D workflows using Universal Scene Description (an open standard).

"Precise, real-time visualization of the remote radiochemical processing in the hot box requires a complex 3D workflow that is integrated with the robot system," she said.

Argonne's Young Soo Park has experience building robotic manipulators. He said the introduction of a radioactive environment adds another degree of challenge. "Successfully building and operating this hot box requires a multi-step process to make the augmented reality experience intuitive for the user," he said. Park leads Argonne's Robotics and Remote Systems program in the Laboratory's Applied Materials division.

"This new funding will allow Argonne to modernize its medical isotope infrastructure, bringing cutting-edge solutions to nuclear medicine," said Kawtar Hafidi, Argonne's associate laboratory director for Physical Sciences and Engineering.

According to Nolen, a goal of the project is for Argonne and other DOE facilities to produce steady quantities of a variety of isotopes. From there, DOE could sell them to hospitals through its Isotope Program.

This project builds on existing robotics and computing capabilities at Argonne, plus collaborations with four colleges and universities: Northwestern University, the University of Illinois at Chicago, Morehouse College and Florida A&M University. This multi-institutional, multidisciplinary team also will involve students and early-career scientists and engineers. Emphasis will be placed on increasing the diversity of the technical workforce of the future.

An advisory team with members from four private companies has been formed to help ensure rapid transition of the technology being developed from the basic research stage to commercial deployment.

Provided by Argonne National Laboratory