To help protect the first responders, researchers at the University of Central Florida have developed a synthetic intelligence approach that not only temporarily and remotely stumbles upon the harsh drug fentanyl, but also learns to stumble upon any unknown derivatives of the past manufactured in underground batches.
The method, recently published in the journal Scientific Reports, uses infrared spectroscopy and can be used on a portable tabletop device.
Fentanyl is one of the leading causes of drug overdose death in the United States. This and its derivatives have a low lethal dose and can lead to the user’s death, can pose a danger to lifeguards and even be used as a spray. “
Fentanyl, which is 50 to one hundred times more potent than morphine according to the U.S. Centers for Disease Control and Prevention, can be legally prescribed to treat patients with severe pain, but is also manufactured and used illegally.
Subith Vasu, professor in the Department of Mechanical and Aerospace Engineering at UCF, co-led the study.
He said the immediate identity strategies of known and emerging fentanyl opioids can contribute to the protection of law enforcement and the army workers’ corps who will have to minimize their contact with those substances.
“This set of AI rules will be used in a detection device we are building for the Advanced Defense Research Projects Agency,” Vasu said.
For the study, the researchers used a national database of biological molecules to identify molecules that have at least one of the functional equipment discovered in the original compound fentanyl. Using this data, they built device learning algorithms to identify the molecules discovered in their infrared spectral properties. Then they tested the accuracy of the algorithms. The AI approach had an accuracy rate of 92.5% to identify fentanyl-related molecules.
Xu stated that this is the first time that systematic research has been conducted that fentanyl-like functional equipment is known from infrared spectral knowledge and uses device learning tools and statistical research.
The study’s co-author, Chun-Hung Wang, is a postdoctoral researcher at the UCF Center for Nanoscience Technology and has helped examine the spectral homes of the compounds. He said it is difficult to identify fentanyls because there are many formulations of fentanyl and carfentanyl analogues.
Artem Masunov, co-author and associate professor at the Center for Nanoscience Technology and the Department of Chemistry at UCF, studied functional equipment that is not unusual for the chemical structures of fentanyl and its analogues.
He said that despite the differences in analogues, they have non-unusual functional groups, which are structural similarities that allow compounds to join receptors in the framework and perform a similar function.
Anthony Terracciano, co-author of the studio and study engineer in the Department of Mechanical and Aerospace Engineering at UCF, worked with Wang to read about infrared spectrum homes. He stated that profiling and infrared spectra studies were fast, very accurate and can be done with a table machine.
Current studies have used infrared spectral knowledge of compounds in the form of gas, however, researchers are conducting a similar examination to use the learning of devices to trip over fentanyl and its powdered derivatives. The generation product will be ready for immediate on-site identity until 2021.
University of Central Florida
Xu, M. et al. (2020) Learning identity of high-precision devices of molecular composite classification applicable to fentanyl research of constituent functional groups. Scientific reports. doi.org/10.1038/s41598-020-70471-7.
News-Medical.net – An AZoNetwork site
Ownership and operation through AZoNetwork, © 2000-2020