Device Prototype

To ensure that our design is produced with appropriate considerations, we aimed to produce a working physical prototype to illustrate our vision for multiplexed diagnostics. In this process, it was imperative that the end-user was a priority and that the device was composed of economically feasible materials. Through collaboration with the transformers, detectors and our human practices sub-teams we were able to manufacture and develop a working prototype that is compatible with our biological system. We designed and manufactured an initial prototype which was then analyzed and altered via rendered drawings and fluid flow analysis to arrive at our final design. Our physical system is based on many aspects of informed design based on research of where this device would be implemented and what requirements it must meet as dictated by international policy.

Process

Our device is portable, user friendly, and manufactured specifically with informed design considerations for the developing world. A drop of blood from the patient undergoes sample preparation to extract target DNA out of the sample. A sample with extracted DNA then flows into the Polymerase Chain Reaction (PCR) chamber where target DNA is amplified via isothermal PCR. Upon completion of the PCR reaction, a plunger is manually pushed down forcing the sample through the tubing into the chambers containing B. subtilis engineered to identify the target disease. Target DNA is up taken by B. subtilis via transformation process. In the digitalized version of the device, a colour sensor reads the colour by shining blue, green, and red lights and measuring the reflection from the sample. The results are then sent to the Arduino Uno microcontroller which outputs the test result on the LCD display after analyzing the input data from the colour sensor. B. subtilis. In each of the chambers, B. subtilis is engineered to detect a specific disease allowing for multidiagnosis of diseases with similar symptoms. Additional considerations regarding the ideal temperature of our device have been explored using a circuit with our Arduino microcontroller, a temperature sensor, and various proposed methods ranging from highly resistive wire to a small and inexpensive heating pad. It is important to consider all environmental factors when creating this device.

Economic feasibility - Two-Tiered Model

Economic feasibility is of the utmost importance when dealing with diagnostics in the developing world and therefore our team chose to tackle this issue using a two-tiered development model. An inexpensive model would utilize the basics of our colourimetric system in which a simple colour change would indicate presence of the specified disease. This signal is easily visible and readable and requires very little technology other than a chamber for Polymerase Chain Reaction (PCR) to occur and consequential separation of the sample into detection wells representing different diseases. Additionally in order to ensure that there is development potential should this modular system expand to other industries a completely digitized device was also explored (Diagnostic Platform). The colourimetric signal of the reporter is picked up by a colour sensor that detects the reflected light of this signal, converts it into a voltage and is output to an LCD screen. This exercises the potential of being able to send results of a test directly to a patient's file with bluetooth application.

Cost Analysis

Our device only uses a very small amount of reagents and the resulting price of these materials is negligible in comparison to the materials that are used to create our device. Materials used in manufacturing the physical device contribute the most to the price of the device. The table below illustrates price approximations for the materials used in our device. The tubing was purchased by our team at a price of $20 per 100ft of tubing. The price for the detection chambers was estimated based on the bulk price for Teflon sheets provided by the machine shop. Bulk prices are comparable to what we would be paying for these materials if were to mass produce our device.

Acrylic covers were used to cover the chambers to use our device with the colour sensor and are not included in the cost analysis above. The cost of this component is very minimal, and it's purpose in the device is simply to provide assurance that the system is not contaminated and that the colour sensor can rest on the chambers. Therefore, we concluded that the cost would total to an absolute maximum $5 which is comparable to other diagnostics for this purpose.

Manufacturing

The goal of producing a physical prototype brought different perspectives of the best way to produce such a device. Through expert consultation it was determined that microfluidics would increase the price greatly and would not be ideal for our device. Using polytetrafluoroethylene tubing would alleviate this issue and prevent cells from adhering to the inner walls of the device. Both options of 3D printing and machine manufacturing were explored, however it was determined that manufacturing the device would allow us to have direct control of the use of polytetrafluoroethylene for both the chambers and the transport tubing. In collaboration with a mechanical engineering student William Gill from the Schulich School of Engineering Machine Shop - a design was created, analyzed, and produced from the desired materials.

Figure 3: 3D model of the initial design

Figure 4: 3D model of the first manufactured device

Colour Sensor

For the digitalized version of our device, we are using a ColorPal Parallax colour and light sensor to detect colour change in the bacteria culture in each of the three chambers. This is to ensure that there is minimal error associated with determining whether there is a positive or negative result; the colour sensor will allow us to quantify the colour and decrease the chance of random errors and inconsistencies. The Parallax sensor uses its’ LED lights to shine red, green, and blue at the sample. Then the sensor measures the reflected light and outputs the voltage proportional to all the light that it sees. The Arduino code then converts the sensor values to red, green, and blue components and outputs it to the screen (or LCD display). Since our prototype has three different chambers with bacteria, the initial plan with the ColorPal Parallax colour and light sensor was to measure the colour in all three chambers at once. In order to achieve that, we would need to have three different colours for the chambers. In theory, we expected to see noticeable change in red, green, and blue components when just one of the colours was present versus two different colours present at a time (co-infection). Figure 1 shows the set up that we used to test the colour sensor with three chambers. However, the results were occasionally inconsistent as the sensor had difficultly differentiating between multiple colours closely aligned together. This system does not optimize the accuracy of the sensor and makes contamination a larger issue. For example, if one of the stripes got contaminated with RFP, we would not be able to tell using the colour sensor.

Figure 5: Colour sensor testing set up: three stripes within 1 cm range imitating the three chambers. Four test squares clockwise: 3 blank; 1 red and 2 blank; 1 red, 1 blue, and 1 blank; 1 blue and 2 blank.

After initial testing, we decided to measure each chamber individually. The sensor was much more effective when detecting one colour at a time, even if the sample was rather faint and less distinguishable. Initially, the code for the colour sensor was referenced the from Parallax website, but was thoroughly modified to suit our specific purposes. The sensor needs to output five consistent measurements before the colour is outputted. Once the colour is changed, the sensor detects new measurements and outputs the new colour after 5 consistent measurements.

Design Optimization and New Design

Our first design was analyzed and optimized to make the prototype more cost effective and compact, as well as allow for the addition of extra chambers. The main change to the design is the location of B. subtilis chambers. The three chambers are now aligned radially and symmetrically (120° angle between the chambers) around the PCR chamber. First of all, this allows us to use shorter pieces of tubing for transport of the sample reducing the cost of the device and making it more compact. Secondly, this design allows us to easily add more chambers and adjust this number if needed. Analysis proved that as long as the chambers are symmetrical and radial around the PCR chamber, the fluid would flow equally into each chamber. This means when the design is altered for a specific purpose or set of disesases, fluid flow analysis would not be necessary as the flow would be guaranteed to be equal. The volume of the B. subtilis chambers was reduced so the size of the chambers was also reduced making the device even more compact. The device can now fit on a circular aluminum pad with a diameter of 4 inches. Alternative materials were explored as opposed to the steel plate from the first design. This was for the purpose of showcasing our device over the course of several months, however cheaper alternatives were explored for implementation of our final device. Through these design iterations we were able to render an image of our new design.

We chose not to manufacture our new device, as our original prototype demonstrates our proof of concept and the fact that sample can flow through our physical system. The two major changes in our device are a more cost-efficient and light-weight base and radial alignment of the chambers resulting in use of less materials. Both of these changes will result in a more economically feasible device that is better suited for the developing world.

Figure 6: 3D model of the new design with three chambers located radially around the PCR chamber to ensure equal fluid flow to each chamber (120° between the chambers)