PatientLink (North Carolina State University)

Daniel Bieber, Richard Daniels, Timothy Martin


Lack of access to communication for hospital patients with disabilities causes a reduction in the overall effectiveness of care and places a financial burden on the healthcare industry.  Some solutions exist to address these communication issues for individuals with special needs, but these solutions are unreliable and even entirely ineffective in some cases.  In conjunction with NC State University and WakeMed Hospital in Raleigh, NC, we set out to develop a method of communication that is both reliable and effective for all hospital patients, regardless of disability.  Using electromyography, we developed a novel, wireless communication device that can better serve the needs of both patients and their caregivers.  While more refinement is needed to commercialize the product we have named ‘PatientLink’, evaluation of prototype devices have demonstrated the capability to allow simple communication using the raise of an eyebrow.


In a survey of sentinel events in hospitals, defined as an occurrence causing risk of death or serious injury, the Joint Commission found that communication was a top underlying cause in each of the past eight years [1].  Patient-caregiver communication in particular is critical to timely assessment, accurate diagnosis, and proper treatment [2-5].  When lapses in communication occur, patients may be removed from their right to know their medical status and be involved in the decision making process of their treatment.  Furthermore, patient discomfort may be increased, quality of life decreased, hospital stays may be lengthened, and in extreme cases, death may occur.  Barriers to communication may present in many forms, and as many as 305,000 patients with common communication disabling disorders are treated in US hospitals each year.

Figure 1. List of CCN diseases/conditions and patient populations.

Figure 1. List of CCN diseases/conditions and patient populations.

Inside a hospital or assisted living facility, all patients are required to have access to nurse call, including those with disabilities.  Standard nurse call accessories, such as pillow speakers and call cords require function of a patient’s hands to activate.  Patients who present with common chronic disabling illnesses such as amyotrophic lateral sclerosis (ALS) or acute injury such as spinal cord injuries (SCI) may have quadriplegia and be unable to use these devices.  The needs of these patients, those who have severe communications disabilities, are commonly referred to as “complex communications needs” (CCN).  The importance of providing access to augmentative and alternative communication (AAC) equipment for these patients is well documented [2].


Figure 2. Pneumatic switch (left) and pressure pad (right).

Figure 2. Pneumatic switch (left) and pressure pad (right).

Adaptive technologies for persons with these types of physical disabilities rely on some function of voluntary muscles innervated by cranial nerves.  Examples include inhalation and exhalation controlled sip-and-puff, pressure pad switches which are typically placed on the shoulder and activated by lateral head movements, eye gaze tracking, and speech recognition.  Of these technologies, only sip-and-puff and pressure pad switches are widely available for commercial nurse call applications.  One recent research study successfully demonstrated a “sniff” sensor for call bell activation, though the device is not yet commercially available [11].

Still, common practices and interventions used on hospital patients, particularly those with severe conditions, cause the use of all existing devices and technologies to be impossible or unreliable. Access to the mouth for a sip-and-puff sensor is blocked by ventilation masks, as would access to the nose for a sniff sensor be. In what may be considered the worst-case scenario, patients with “locked-in syndrome” may be completely paralyzed and only have some residual control of facial muscles.  There are no commercially available solutions that can be rapidly deployed and reliably used by any patient who needs to use nurse call.

Our group found out about the problem of inadequate communication available to patients with disabilities in the hospital during a design and development course at North Carolina State University. As graduate students, Richard Daniels and Timothy Martin started with meeting with our clinical partners at WakeMed Hospital in Raleigh, North Carolina. After defining the problem with current devices, we met monthly with the rehabilitation staff at WakeMed (occupational therapists, physical therapists, intensive care nurses, biomedical engineering and clinical management) to discuss the user needs and requirements of a prototypical communication system. As part of our grant through the National Collegiate Inventors and Innovators Alliance, we had the opportunity to meet with healthcare practitioners in facilities in the greater Los Angeles area. We met with clinicians from Kaiser Permanente, Cedars-Sinai, and the Rancho Los Amigos National Rehabilitation Center. At these facilities we received a new perspective on the problem and confirmed that the problems and drawbacks of current assistive technology for calling the nurse were not regional but present nationwide.


Design Goals: To provide patients with disabilities a nurse call system that they are consistently able to use, provides vital feedback to the user, and reduces the number of false calls.


To address the need for reliable, equal access to communication, we propose PatientLink, a novel wireless communication platform for patients with disabilities.  The PatientLink system is composed of a patient input controller, a wall adaptor, wireless radio technology, and software for data processing and patient feedback.   PatientLink utilizes a novel and patent pending method for allowing a patient to simply raise their eyebrows to call the nurse. The facial sensor component accomplishes three important tasks: calibration, biosensing, and data processing. Calibration allows PatientLink to adapt to any patient and can be used on different parts of the body if desired. PatientLink combines a custom designed electromyography sensor with three electrodes to provide the signal. The data processing in PatientLink limits the false calls seen in other ineffective technology by including a two stage nurse call. First, when the patient raises their eyebrows beyond the custom threshold, the device LED blinks to let the user know the process has started. Second, if the eyebrow raise is held through the initial stage, the device completes the call and provides an extended LED blink signifying the call has been made. Wireless technology utilizing the Zigbee protocol links the facial sensor (sensing and processing) and the adaptor (feedback and nurse call). The adaptor allows nurse call to be activated by interfacing with existing nurse call system wall units.

PatientLink Alpha

Figure 3. PatientLink Alpha: Arduino based system with a third party EMG circuit purchased to capture the signal.

PatientLink Alpha’s design is focused on two main components: the Arduino microcontroller and the Advancer Tech EMG circuit. Tim authored custom software that allowed the Arduino to read the analog signal from the EMG sensor and process the signal. The main components of the signal processing described in the Solution section were present in this software code and were honed using the Arduino. The microcontroller calibrated the initial signal, processed the real time signal and would provide a two stage activation with the initial alert through blinking a LED and then a confirmation of a call with a sustained LED illumination. The EMG sensor was found through an NCSU connection and allowed us to get the signal we needed with battery power. Below is a video showing the complete Alpha prototype being used.

PatientLink Beta

Figure 4. PatientLink Beta: custom EMG circuit with Synapse wireless microcontroller.

PatientLink Beta is the second iteration of the PatientLink design and builds on the first prototype by accomplishing three main things: custom EMG with reduced components, wireless radio capability, and a small, portable design that is battery powered. With the help of Elliott Tech, a local electronic design company, we were able to take our breadboard design for our custom EMG to a PCB form factor that is battery powered. The Synapse Wireless radio and microcontroller contains the wireless communication hardware and the microprocessor for PatientLink Beta. The software for the SNAP module is a modified version of Python and was written by Daniel Bieber and Timothy Martin. There are three electrodes that connect to gel based EMG electrodes placed on the forehead as seen in the Alpha and Beta demonstration videos. The cost of the current prototypes range around $100 for all components involved. Our longterm goal is to produce a disposable sensor that would last for one week on a battery and would be sold for around $50. We have received feedback from clinicians and potential users that this price would be attractive.


PatientLink Alpha was tested with our software to confirm the signal processing that we developed through the code written for the Arduino platform. When tested, the signal coming from the EMG through the Arduino after calibration with no eyebrow raise was around 300-400. When an eyebrow was raised, the signal would go up to 800-1023. This shows that our sensor was able to detect the eyebrow raise and distinguish between the resting signal of the forehead and when the sensor was purposely activated.

Figure 5. Sensor reading from PatientLink Alpha.

PatientLink Beta was tested both in an electronics lab to further refine the upgraded design and to fully understand the performance of our sensor. Using an oscilloscope, we were able to determine the performance of our custom designed EMG. The signal from our EMG had a resting value of between 750-800 mV and would go up to 1000-1100mV when the forehead sensor was activated. This was concurrent with our design and showed that we could differentiate between the two signals. Future development will center around conditioning the signal and perfecting the software related to calibration of the signal and processing of the nurse call.

Figure 6. Oscilloscope uncalibrated data from PatientLink Beta showing resting transition to eyebrow raise.


PatientLink is a device that will provide hope and satisfaction to patients with disabilities using the current assistive technology for calling the nurse. By utilizing a simple and unique approach, the PatientLink team designed a communication platform that enables any patient in the hospital to call the nurse when they want to. Our design and device has come a long way in two years and delivers the core features needed for a universal communication platform for individuals with a wide range of disabilities. Our original goal for our partners at WakeMed was to deliver a solution that helped to improve the experience of patients with the most severe impairments that enter the doors. That goal is what drives us forward to deliver a product that fulfills the patients’ desire to have some control over their experience while they are in a traumatic or frustrating environment.


 1. The Joint Commission (2011). Sentinel Event Data – Root Causes by Event Type (2004 – Third Quarter 2011). Retrieved from


2. Finke, Erinn (2008). “A systematic review of the effectiveness of nurse communication with patients with complex communication needs with a focus on the use of augmentative and alternative communication”. Journal of Clinical Nursing 17 (16).

3. Priest, Susanna Hornig (2010). “Physician-Patient Communication” in Encyclopedia of science and technology communication (1-4129-5920-9, 978-1-4129-5920-9).

 4. Ayers, Susan (2007). “Healthcare professional–patient communication” in Cambridge handbook of psychology, health and medicine (0-521-60510-5, 978-0-521-60510-6).

 5. Doss, Sheila (2011). “Patient-nurse partnerships”. Nephrology nursing journal : journal of the American Nephrology Nurses’ Association, 38 (2), p. 115.

 6. National Spinal Cord Injury Statistical Center (2011). Spinal Cord Injury Facts and Figures at a Glance. Retrieved from:

 7. Dean, G (1994). “How many people in the world have multiple sclerosis?”. Neuroepidemiology, 13(1-2), p. 1.

 8. National Instititute of Neurological Disorders and Stroke (2010). ALS (Amyotrophic Lateral Sclerosis) Fact Sheet. Retrieved from:

 9. National Instititute of Neurological Disorders and Stroke (2004). Parkinson’s Disease

Backgrounder. Retrieved from:

 10. National Institute of Neurological Disorders and Stroke (2002). Traumatic brain

injury: Hope through research. Retrieved from:


 11. Plotkin, Anton (2010). “Sniffing enables communication and environmental control for the severely disabled”. PNAS, 107 (32), p. 14413.


Professor Andrew DiMeo for providing guidance and mentorship during the project.

The staff at WakeMed Hospital in Raleigh, NC for providing us key clinical information.

Elliott Tech in Holly Springs, NC for providing electronic design consulting.

NCIIA E-Team grant and development program.

North Carolina Translational and Clinical Sciences Association Kickstart Commercialization Grant.


Tim Martin


, , , ,

Powered by WordPress. Designed by WooThemes

Skip to toolbar