Bioengineering Blog: Identifying Gastric Disease from the Inside

Gastrointestinal illnesses can be difficult to diagnose and treat. Some 62 million people in the U.S. are diagnosed every year with some type of GI illness, including Crohn’s disease, irritable bowel syndrome or cancers of various types. The earlier a physician can diagnose some of these abnormalities, the greater the chances are of a positive outcome.

But the GI system can be difficult to navigate without surgery. Standard medical procedures such as colonoscopies and endoscopies are limited in their reach because the GI tract is over 30 feet long. They also are administered infrequently, every five to 10 years, and their invasiveness makes many people avoid the procedures altogether.

Now, researchers at Stanford University’s BAAM lab within the School of Medicine are developing a robotic pill that can move within the GI tract to collect biomarkers over long sampling periods. The untethered pill is fitted with a magnetic core for locomotion, a delayed gate mechanism to control a sampling location based on changes to its environment, and an enrichment module that traps biomarkers.

More for You: FDA Looks for Input on Medical 3D Printing

“Once swallowed, the pill collects information by trapping molecules or pathogens of interest in a sodium polyacrylate gel — an absorbent material,” wrote Fernando Soto, in Advanced Science News. “In an in vitro proof-of-concept experiment, the robotic pill was capable of collecting diverse biomarkers, including proteins and bacteria. After capture, the biomarkers are released from the gel and analyzed in a lab using standard analytical methods, including immunochemical and PCR tests.”

Soto is lead author of a paper describing the work recently published in Advanced Intelligent Systems. He noted that smart pills are already being used to investigate the inner workings of the body. PillCam, for instance, is a capsule fitted with a camera that is used to search for polyps in the GI tract. Others have been used to capture and sense biomarkers in different applications.

”The ability to isolate biomarkers within localized environments could help discover novel early disease signals and monitor health status by generating comprehensive data sets encompassing the shift from healthy to disease states and vice versa,” the authors wrote.

Soto noted that stool or blood samples help in identifying diseases, but they give limited information on the source of the problem. “Thus, there is a great need for creating simple approaches that can take samples at multiple locations in the gut to narrow down and differentiate between different GI cancers, for example, such as pancreatic, intestinal, or colorectal cancer.”

Editor's Pick: Sharper AI, Virtual Testing Coming Soon

The proof-of-concept test involved a device with three compartments with ones for different features, including sampling, delayed opening of the sampling gate, locomotion and keeping the robotic pill in predetermined positions. In testing, a magnet was placed below a table on which the robot was placed, and rotated, causing the robotic pill to rotate. The pill as then “forced” to move across small stones and a piece of pork belly bought at a local supermarket. It was also tested in fluid, using either a plastic tube or a piece of pig’s intestine connected to a fountain pump to generate a closed-loop system.

The next step is to work with large animal models, to evaluate the pill’s ability to capture and diagnose disease. It also will be integrated with automated guidance and imaging for greater control and data collection. According to their research, Soto and his coauthors believe a group of robotic pills be more precisely actuated using a Helmholtz magnetic coil to produce a tunable magnetic field.

Further animal studies would also test the pill under biologic conditions. The pill would be driven using the synchronized contraction and relaxation of GI muscles pushing content forward. Proof-of-concept testing used laminar flow modeling. Soto believes the principles outlined in the work will lay the foundation for a tool that can be used in routine checkups, collecting information that can be used predict disease earlier than now possible.

John Kosowatz is senior editor.