Inspiration
When we all attended the postoperative track pitch, we were told the reality of what many women face during and after childbirth in low income countries. Due to poor resources in underdeveloped countries, dealing with postpartum hemorrhaging (PPH) is difficult, as people are unable to obtain proper medical equipment, lack medical expertise, and are understaffed for the constant monitoring PPH requires. These deaths are all preventable, as this condition is not nearly as lethal in developed countries. We were all disturbed by these odds, and felt compelled to come up with a solution.
What it does
Our device measures light resistance in blood by shining a light through the blood, or finger if noninvasive, and picks up the light that filters through with the light sensor. Resistance levels are collected by the device. This data is then put into the Hemolite website along with a username, preoperative hematocrit levels, and anemia status. The preoperative hematocrit levels will be transformed into hemoglobin levels, which will be used as the standard to compare future data points to. The scaling factor achieved when transforming from hematocrit to hemoglobin will be used when the difference between standardized and new data is calculated. If the difference passes a certain threshold or the hemoglobin level is below 7g/dL, then there will be an alert on the web page stating that this patient is at high risk of hemorrhaging.
How we built it
We connected an Arduino Uno board and a light sensor, and then programmed the Arduino using its language. We would get the resistance reading from the light sensor. For example, when there is low light, the resistance is higher. Higher resistance also correlates to higher hemoglobin levels as light intensity is decreasing. This relation will be pertinent to determining hemoglobin levels and risk of hemorrhaging.
We utilized Java in Eclipse in order to create a program that will take in the new patient input data. This data will be put into an arraylist which will contain initial data and subsequent data (hemoglobin levels) collected. The data will be analyzed and compared to the initial standardized values to determine risk of hemorrhaging (low or high) which will then be sent out to the website. Each patient will get its own new arraylist. There is also an arraylist that compiles all the data for all patients, creating a local database that can then be called up at any time to see past lab readings and results. The backend was then built using Spring, a Java framework.
We also used React, a Javascript framework, to create three user friendly pages, including the login page, data entry page, and sign up page. Bootstrap and Material-UI were used to make an aesthetic and welcoming website. Google cloud platform and domain was used to get the domain page.
Challenges we ran into
Initially, we attempted to use light absorption and hematocrit levels. However, as we do not have the data we need to make this project work, we had to switch to light resistance and hemoglobin levels in order to demonstrate how the project would work through the coding. In addition, we considered using a color sensor to register wavelength, but supplies ran out. Another complication that arose near the end was getting the code to interact to create the webpage.
Accomplishments that we're proud of
We were able to integrate many concepts that we weren’t fully familiar with into our final project. We worked together well, listening to each other’s suggestions and taking them into consideration.
What we learned
We learned how there are many low-income areas that face various barriers when dealing with effective patient care, whether those barriers be limited staff, insufficient technology, or limited finances. Once we decided that postpartum hemorrhage was the issue we were going to tackle, we were faced with the challenge of identifying both external and internal bleeding. Through our research and discussions with various health care professionals, we learned about hematocrit and hemoglobin, and how they can be used as identifiers in bleeding. We had a hunch that we could use light to measure them, which was confirmed through more research. We also learned how to make a website and work through any issues that came with the coding process.
What's next for Hemolite
In our future endeavors, we aim to collect data on the light resistance of blood samples depending on their level of hemoglobin. During data sampling, scattering properties of blood must be taken into consideration as they can cause resistance readings to be lower than actual readings. We can accommodate for this by taking several data points that can be averaged to one hemoglobin level. Another solution would be creating a mathematical function which takes these factors into account. Though it is unable to be implemented at the time being, this mathematical method has been done before in determining the absorption coefficient of blood with hematocrit levels of 45%, making it a feasible option (see “A literature review and novel theoretical approach on the optical properties of whole blood”). Once we obtain the desired data, we would be able to use various light measurements, such as resistance, intensity, etc. We can collect either hematocrit or hemoglobin levels, as both have the same importance when testing for blood loss. There also could be the possibility of implementing blood pressure, heart rate, and breathing rate as factors in our program, as they indicate hypovolemic shock and could potentially alert medical staff of hemorrhage sooner than hemoglobin levels alone. However, this would not happen in the near future, as identifying potential relationships and algorithms between all these variables would require large amounts of patient data, and possibly AI and machine learning.
Log in or sign up for Devpost to join the conversation.