Model Predicts the Progress of Disease Detection
By Calla Cofield
Many groups at the APS March Meeting presented their work on new devices for disease detection. Three groups presented work on more robust TB detectors. Another company revealed proof of concept for a handheld device that could do everything from monitoring white blood cells to identifying bacteria in drinking water.
John X. J. Zhang and researchers from the University of Texas at Austin want to improve on current technologies to identify tumor cells that appear in the bloodstream long before tumors become visible through imaging techniques. They are engineering sensors that go out and look for the cancer cells in the sample, in contrast to current technology in which the sensors remain stationary and wait for the cells to pass by.
In this way Zhang and his group have changed the relationship between the number of disease indicators in a volume and the time it takes for them to reach the sensor. According to Ashram Alam, a professor of electrical and computer engineering at Purdue University, this is the underlying challenge for all biosensor technologies.
“On the surface, these technologies, protocols and approaches look very different,” said Alam. “But they all use different manifestations of the same principles. They all rely on how the molecules diffuse to the sensors. [Our work] along with others from Naval Research Lab (led by P.E. Sheehan) and MIT (led by T.M. Squires) speaks mostly to the physics of sensitivity and how far you can go based on the physical principles of how molecules are detected.”
Alam spoke at a press conference at the meeting in Portland and related the task of disease detection to a game of cops and robbers. New technologies hope to increase their ability to find the elusive robbers (disease indicators in large sample volumes) and create better cops (more effective means of capturing and identifying those indicators).
Alam and his group, including postdoctoral researcher Pradeep Nair, studied the zoo of biosensors and disease detection techniques and put the parameters that govern the “cops and the robbers” into a basic model. Those parameters include how well a team’s engineered nanoparticles bind to the disease indicators, the device’s signal to noise ratio, or a detector’s ability to handle a cell gently enough that it doesn’t destroy it in the capture process.
The Purdue researchers are calling their model a “periodic table” because it provides a kind of map of all the current biosensors based on their unique values for each of the individual parameters. The map then reveals gaps in the current technology, and provides developers with the opportunity to fill in those gaps with new devices. The team first developed their model based on technology available in the 1970’s, and found that they could successfully predict technologies available today. Alam says they have already begun to see some of their current model predictions fulfilled.
“All these parameters can now be understood in a global context, so you don’t just start changing one without understanding the others,” said Alam.
The model also predicts that current biosensor technology will reach a wall of sensitivity, limited by the density of disease indicators in a volume, and the amount of time the researchers are willing to wait for the results. The wall can be overcome by new techniques such as amplifying the signal from the disease indicator.
“There really are a huge number of things depending on this kind of understanding,” he said. “And in the end we hope something significant and wonderful will come out of it.”