While last week was spent laying a lot of general groundwork for the internship, this week has focused in on beginning our first design project: a phototherapy dosing meter.

• Background. The particular phototherapy devices we are targeting treat neonatal jaundice, a condition that effects up to 60% of premature babies. Neonatal jaundice is caused by a buildup of bilirubin (a protein) in the bloodstream, and is often harmless. In more extreme cases, however, the bilirubin levels will become dangerous and can lead to long-term neurological damage. Queen’s (the hospital down the street from the Poly) uses blue phototherapy lights, designed and built largely from the efforts of both Rice and the Poly, to treat these more dangerous cases of hyperbilirubinaemia. The device is a sort of plate of blue LED lights, all of which output the same wavelength of blue light. The LEDs are placed a short distance above the baby’s body so the light passes through the skin, breaks down bilirubin, and allows the baby to process the protein and prevent buildup.

• The problem. It is very difficult for a doctor or a nurse to discern how much light the baby is receiving; the intensity of light received by the baby changes significantly with distance from baby to light, and different amounts of light are needed to treat varying intensities of the disorder. The current phototherapy lights have a knob to adjust the light irradiance, however this knob ranges only from the designated “min” to “max,” not numerical measures of irradiance. We’re setting out to design a device that can measure the irradiance (which is essentially a measure of how much light the baby is receiving) of the bililights quickly, accurately, and cheaply.

• Past solutions. This problem isn’t a new one; the particular bililights found at Queens have been around for a few years, and a method to measure the exact dosage of light has been needed for longer. However, the commercial dosing meters available in the US cost thousands of US dollars, making them too expensive to be practical at Queen’s. The current dosing meter used in Queen’s only costs around $40, but is an analog meter. This means the meter suffers from calibration drift: when the mechanical components that control the meter start to wear, the device begins to become inaccurate. Additionally, the analog meter can become inaccurate if it is dropped, bumped, overused, or slightly tilted such that the meter pin falls one way or another due to gravity, disturbing the irradiance measurement. Teams at both the Poly and Rice have developed alternative solutions, however most still use analog meters—which suffer from drift, bulkiness, and inaccuracy—while their digital counterparts haven’t been properly calibrated.
• Our solution. We wanted to use ideas from previous teams who have tackled this challenge, but combine various aspects of these past designs to optimize a device that can hopefully finally meet all of the design criteria needed for a dosing meter. We decided to make a digital device, which will be more accurate in both short and long term as compared to the analog device. The circuit has been mostly redesigned in order to fit the components that we have and to optimize sensitivity. Additionally, the housing has been modified to make the device portable and easy to use, while maintaining accuracy by keeping the meter perpendicular to the light source.

• Progress. We first spent time sifting through the past designs to understand the design decisions they made. This helped us understand what components we wanted to keep vs. what we wanted to change. After deciding on the digital design, Mr. Vweza pointed out the need for proper calibration (turning the current created by a photodiode into a measure of irradiance) as well as a few other design modifications, so we researched solutions to improve our device from previous models. We also had a large brainstorming session on various housing options, and have finally chosen a model that suits the device’s constraints. Finally, we’ve laid out the circuit on a breadboard and begun troubleshooting this new circuit as well as adjusting the Arduino code to obtain the best results.
• The next steps. We will continue to optimize the circuit tomorrow, and hopefully begin building the housing tomorrow as well. With 7 people working all day on this project, things move fast, and we hope to have an initial prototype sometime next week. Then, we must calibrate the device and test with the gold standard of care, and finally make any adjustments needed. Hopefully we will have the device up and running soon!

In other news! We’ve begun searching for our own design project, meeting with many people around the hospital to identify needs. We have a large list going now, but hopefully by the end of next week we’ll have narrowed it down and decided upon a novel design project! We also had a meeting with Mr. Mafuta about the orientation week, but that will be a work in progress throughout this summer. Finally, the five Rice interns in Blantyre hiked part of Mount Mulanje last Sunday, which was incredibly difficult and something my legs were not ready for, but also was beautiful and well worth the effort.