Interview: Teresa Sanders Discusses TI Tech That Helps Parkinson’s Patients
This file type includes high resolution graphics and schematics.
According to the Parkinson’s Disease Foundation, Parkinson’s patients spend an average of $2500 a year on medication. As symptoms such as tremors and instability change, though, so does the need for that medication. Effective treatment depends on precise evaluation of those symptoms, but doctors can’t continuously observe patients and make incremental adjustments.
Related Articles
- Build A Wrist Heart-Rate Monitor Using An Ultra-Low-Power MCU
- Choose The Right Resistors For Medical Applications
- Biomedical Devices Improve And Extend Patient Lives
Teresa Sanders, a Texas Instruments Fellow and PhD student at Georgia Tech, was working with Emory University Parkinson’s disease specialists when she received an eZ430-Chronos watch from TI. She then realized that a watch with an accelerometer could measure limb tremors and be paired with other tools to provide information for accurately medicating Parkinson’s patients, even at home (see the figure).
The watch measures the patient’s tremors and sends the data to a smart phone, which the patient also wears to track speed and stability. The phone processes the information and remotely transmits it to medical personnel for evaluation. We spoke with Sanders about the hardware and software behind the system.
ED: What makes the eZ430-Chronos watch a good fit for the system?
TS: Our primary application for the watch was using the accelerometer to monitor limb tremor. The eZ430 CC430 system, together with the USB-based CC1111 wireless interface (915 MHz), allowed our software to access and control the CC430 on-board three-axis accelerometer from either the PC (Python) or the Samsung Galaxy S3 smart phone (Java).
ED: How does the system use accelerometers measure movement?
TS: The system uses the CMA3000-D01 low-power three-axis accelerometer for limb tremor measurement, while the Samsung S3 smart-phone accelerometer and gyroscope are used to measure the other signs of Parkinson’s disease.
ED: How does the system distinguish between tremors caused by Parkinson’s disease and natural arm movements?
TS: Parkinson’s tremors have a distinctive frequency profile. After removing signal artifacts, we detect this profile from the watch accelerometer data using band-pass frequency filters. Since we simultaneously track body trunk movements using the smart-phone sensors, we can then isolate true Parkinsonian tremors that occur at rest from signals related to overall body movements.
ED: Did you need to develop new software to enable the watch and phone to monitor tremors?
TS: Yes, we had to create a smart-phone Java app to enable wirelessly controlling and reading from the watch (and saving the accelerometer data to the smart phone). A second app simultaneously collects data from the smart-phone on-board sensors to allow analysis of additional Parkinson’s disease (PD) signs (other than tremor). We also created offline analysis algorithms and software (Matlab) for detecting the tremors and other PD signs from the saved data.
ED: What about smart-phone compatibility? Will it work with Apple, Android, and Windows devices?
TS: Our Java app should work in Android smart phones with USB host mode capability. However, so far, it has only been tested in the Galaxy S3. We have not yet created an app for Apple or Windows mobile devices.
ED: How did you develop the user interface so patients can easily use it?
TS: The data collection apps use standard Android user interfaces. However, the system is currently designed for medical professionals to set up. The system should then run continuously without patient intervention (although the patient can, of course, remove the smart phone and watch as needed). We have plans for an eventual doctor/patient data interface tool.
ED: Could the system be adapted for use with patients who have other neurological disorders that affect motor functions?
TS: Yes, the system could be adapted to detect motor signs related to other disorders such as multiple sclerosis (e.g., loss of balance, postural and/or intention tremors) or ALS (e.g., twitching, reduced movement/weakness on one side of the body).
ED: Do you have a timetable for when you expect to begin clinical trials?
TS: Our on-site IRB (institutional review board) approval allowed us to collect preliminary data in our lab. We have been talking with a potential physician partner for pre-clinical trials and hope to begin these in August.
After obtaining a master’s degree in electrical engineering from UCLA, Teresa Sanders spent several years in industry designing, simulating, prototyping, and field-testing infrared and radar missile systems. In 2008, she returned to graduate school for a PhD in bioengineering at the Georgia Institute of Technology. While at Georgia Tech, she gained advanced knowledge in biochemistry, systems physiology, and neuroscience and applied her signal processing and embedded systems knowledge to design novel algorithms and devices for subcellular particle tracking, cancer staging, and neural decoding. She has authored numerous papers, holds one signal processing patent with another patent pending, and is the recipient of both Texas Instruments Leadership University and NSF Teaching Fellowships.
This file type includes high resolution graphics and schematics.