Researchers at MIT have developed a system that relies on wireless radio signals and artificial intelligence (AI) to detect errors when patients self-administer medicine. The new development could have a big impact given the alarming number of patients that fail to adhere to doctors’ orders, which leads to thousands of deaths and billions of dollars in medical costs each year.
The system uses wireless sensing and AI together to determine when a patient is using an insulin pen or inhaler. Potential errors are detected by it when a patient self-administers medicine.
Dina Katabi is the Andrew and Erna Viteri Professor at MIT. Katabi’s research group was responsible for developing the new system.
“Some past work reports that up to 70% of patients do not take their insulin as prescribed, and many patients do not use inhalers properly,” Katabi says.
According to the researchers, the new system can be installed at home and alert patients and caregivers of medication error, which helps reduce unnecessary hospital visits.
The research was published last month in the journal Nature Medicine. The lead authors of the study include Mingmin Zhao, PhD student in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), and Kreshnik Hoti, former visiting scientist at MIT and current faculty member at the University of Prishtina in Kosovo. Co-authors of the research include Hao Wang, former CSAIL postdoc and current faculty member at Rutgers University, and Aniruddh Raghu, CSAIL PhD student.
Drug Delivery Mechanisms
Many drugs require complex delivery mechanisms.
“For example, insulin pens require priming to make sure there are no air bubbles inside. And after injection, you have to hold for 10 seconds,” Zhao says. “All those little steps are necessary to properly deliver the drug to its active site.”
With each additional step comes more chances for errors, which is increased even more if there is no pharmacist present. Since patients often make mistakes without realizing it, the team aimed to create an automated system.
The new system has three broad steps, starting with a sensor that tracks a patient’s movements within a 10-meter radius. This step is done through radio waves that reflect off their body. Then, AI looks at the reflected signals to determine if a patient is self-administering an inhaler or insulin pen. The last step is for the system to alert the patient or health care provider when an error is detected in the self-administration of the medicine.
“One thing nice about this system is that it doesn't require the patient to wear any sensors,” Zhao says. “It can even work through occlusions, similar to how you can access your Wi-FI when you’re in a different room from your router.”
Sensor and Neural Network
The sensor sits in the background of a house while using AI to interpret the modulated radio waves. A neural network was developed to detect patterns in the use of the medicine, and it was trained to perform example movements. Through reinforcement learning, the network successfully detected 96 percent of insulin pen administrations and 99 percent of inhaler uses.
After identifying any errors, the network can also correct them. Proper medicine administration follows similar sequences, which means the system can identify any anomalies in the specific steps. That information can then be sent to the patient or their doctor, which helps correct the technique.
“By breaking it down into these steps, we can not only see how frequently the patient is using their device, but also assess their administration technique to see how well they’re doing,” says Zhao.
“An alternative way to solve this problem is by installing cameras,” Zhao continues. “But using a wireless signal is much less intrusive. It doesn't show peoples’ appearance.”
According to the team, this new system could eventually be adapted for other medications by retraining the neural network.
- Innovative Acoustic Swarm Technology Shapes the Future of In-Room Audio
- 10 Best ChatGPT Prompts for Developing Content Strategy
- Vitalii Romanchenko, CEO & Co-founder of Elai – Interview Series
- How AI Helps With Education Automation And Management
- Murf AI Review: The Ultimate AI Voice Generator in 2023?