Home

Noninvasive Intracranial Pressure Monitor

WINTER 2015 MAE 156B SPONSORED PROJECT

UNIVERSITY OF CALIFORNIA, SAN DIEGO

SPONSORED BY UC SAN DIEGO, SCHOOL OF MEDICINE

Team 3: Brad Adams,  Joeran Melby,  Minjae Lee,  Peggy Ip

Figure 1. Isometric View of Final Design

Figure 2. Assembly of Final Design

Background:

Intracranial pressure (ICP) is of great importance when dealing with head trauma and neurological disorders. Normal intracranial pressure,  determined by the volume of brain, blood, and cerebrospinal fluid (CSF) within the rigid skull, ranges from 7–15 mmHg. As intracranial volume changes, pressure changes occur. For example, a patient who suffers a traumatic brain injury may have hemorrhage and swelling of cerebral tissue, causing increased volume, which raises intracranial pressure against the fixed skull. Alternatively, a patient with obstruction of CSF will have an elevated ICP, referred to as hydrocephalus, as fluid builds up in and around the brain. This pathology is often present in infants. An elevated ICP over 20 mmHg is associated with significant morbidity and mortality if not treated.

Objective:

The project objective was to design, build, and test a non-invasive pressure monitoring device for use on patients with craniectomy or unfused fontanelles. Inspired by an ocular tonometer (measures intraocular pressure), the device measures intracranial pressure by applying a force sensor to the scalp (without underlying bone). Initially the device was tested in the laboratory on ICP models such as inflated balloons with known pressures to assess its accuracy. Ultimately, applying for approval to test it in a clinical setting would allow comparison between the device and invasive ICP monitors already in place.

Overview of Final Design

Figure 3. Exploded CAD of Final Design

The device consists of an outer shell, a load cell, force sensitive resistors, LEDs, an instrumentation amplifier, a microcontroller, a shift register, and a LCD. Applanation principle was applied in the design of the device, enabling the force against the load cell to be translated into pressure. The microcontroller (Arduino Uno) controls the measurement process and reads inputs from the load cell and the force sensitive resistors. The load cell, one of the most important components, works like a cantilever beam. Shear strain gauges outputs a voltage proportional to the load. The probe transfers the force required to flatten a known area of the skin to the load cell.

The outer shell is 3D printed and ergonomically shaped. It provides a comfortable grip, and the device is small and light enough to be portable. The measured pressure is displayed on a big backlit LCD, and LEDs guide the operator in placing the device correctly on the patient. Force sensitive resistors on the bottom of the device are responsible for verifying that the contacting area is large enough for an accurate measurement. Each LED represents one of the force sensitive resistors at the bottom of the device. If a force sensitive resistor is not contacting the patient’s skin, the associated LED will remain off. If it is contacting the skin surface with proper force, the LED will light green. If too much force is applied, the light will be red. This color coding allows the operator to adjust the positioning of the device in order to ensure a correct measurement value. The measurement is only taken when all four LED’s are green simultaneously.

Animation 1: CAD model design    

Summary of Test Results:

Figure 4. Final Calibration Curve

The device was calibrated with synthetic skin with a thickness of 0.17 mm. The standard error of the regression was determined to be 2.1 mmHg, and the standard error of forecasted value is 2.3 mmHg. The accuracy of the calibration curve was evaluated by comparing the forecasted pressure in mmHg to a known pressure as indicated by a manometer. The result was an average error of 1.195 mmHg, and an average percent error of 7.62%. No significant problems with hysteresis were found.

The final device was able to measure pressure pressures from 0 – 35 mmHg with ± 2mmHg. The device gave less accurate results when adding different skin thicknesses, indicating the further calibration was needed in order to account for skin thickness variations. Increasing the pressure range resulted in lower accuracy. The force sensitive resistors were not the optimal solution for controlling the applied force and the flattening of the measurement area. They were unable to accommodate variation in the measurement situation, such as change in shape, thickness and elasticity of the skin. It was also found that the shape of the bulging skin had more impact on the measurements than first hypothesized, due to changing the points of contact with the force sensitive resistors.

The power consumption of the device was larger than anticipated. One standard 9 V battery was able to endure multiple test cycles, but for continuous use, a larger battery or an AC adapter is recommended.

The device was a successful proof of concept, but requires further improvements before it can be used in a clinical setting. Most important is finding an appropriate alternative to the force sensitive resistors, expanding the measurement range to 50 mmHg and improving the battery life. Additional functionality can be added for ease of use, and a disposable sterile film on the contacting surface allows use of the device on several patients in succession.