Inflammation Detection
Design matrix comparing all methods for detecting inflammation. Digital Image Analysis was chosen because it provided a way to robustly evaluate inflammation non-invasively; so that outside factors such as skin conditions and tone would minimally affect risk assessment, and did so at a feasible price for a project prototype.
| Invasivity | Robustness | Cost | Ease of Use | |
| Digital Image Analysis | Takes pictures of the skin. | Can detect patterns of inflammation beyond just color changes, is not affected by skin color, requires uniform lighting | 130.00 | Large amount of documentation and code examples available for camera modules and arduino boards, group member previous experience with arduino. |
| Multi-Spectral Imaging | Records the multiple wavelengths of light emitted by skin | Not affected by skin color, can not discern different patterns of inflammation present on skin, requires precise lighting | 435.00 | Large amount of documentation and code examples available for arduino boards, limited documentation and code examples for multi-spectral modules, group member no previous experience with multi-spectral imaging. |
| RGB Sensor | Records the red, green, and blue light coming off the skin | Affected by skin color, can not discern different patterns of inflammation present on skin, doesn’t require specific lighting | 20.00 | Large amount of documentation and code examples available for rgb modules and arduino boards, group member previous experience with arduino |
Obtaining Systemic Oxygen Levels
Comparison between a number of bluetooth modules. All bluetooth modules that had been considered have similar specifications. The only significant difference is the board that each was compatible with. Based on this companitbality factor MK71251-02 Spresense camera module was chosen by.
| Bluetooth Module | Cost | Input/Output Capabilities | Power Consumption | Compatibility |
| Adafruit 2479 Bluetooth module | 17.5 | Yes | 76mW | Arduino Uno |
| MK71251-02 Spresense/ Name: ROHM SPRESENSE-BLE-EVK-701 | 41.75 | Yes | 24.12 mW | Sony Spresense |
| Adafruit Feather nRF52840 Express | 24.95 | Yes | 44.4 mW | Adafruit Feather |
| Oximeter | Cost | Accuracy | Compatibility |
| OxySmart™ Fingertip Pulse Oximeter | $49.00 | +/- 2% (advertised) | Connects to app through bluetooth |
| Finger Pulse Oximeter with Bluetooth LE – BM1000 | 49.95 | +/- 2% (advertised) | Adafruit industries. |
Size & Adjustability
For adjustability the anklet will be put on the subject with a stretchable perforated fabric and adjustable clip in that its length can be changed to accommodate the circumference of the subject’s ankle.
Due to the patient having to wear the device for long periods of time, comfortability is paramount in deciding the proper material for the strap. A buckle was initially thought of as an option but due to the possible discomfort a more appropriate option was determined to be a vinyl polyurethane leather strap. This mechanism will be durable enough to hold the electrical components as well as due to the perforation, will be breathable and more comfortable to wear for the patient over long periods of time.
Design matrix comparing possible strap material considerations for the adjustability mechanism.
| Strap/Buckle | Cost | Size | Biocompatibility | Comfortability | Durability |
| Vinyl Fabric Perforated | $10.99 | 54” x 36” | Yes | High | High |
| Roll Set w/Clips | $9.99 | 4” x 5’ | Yes | High | Moderate |
| HOSSPORA Cotton Elastic Bandage with separate strap | $10.19 + $5.99 | 2” x 13-15’ | Yes | Moderate | Moderate |
Battery Life and Power Consumption
Since using a battery pack is much safer than a wire to operate the device, the team calculated the device’s theoretical energy consumption required to power the device for at least 12 hours, which comes out to be approximately 15 W*hr (Hand Calculations 2). A design matrix was then used to determine the ideal battery that will be used to power the prototype.
Battery design Matrix shows why battery pack (REV 16000) will be used for testing as it can be obtained for free, then (INIU) Battery pack will be used to power the final prototype for it has the largest battery density and is the cheapest.
| Battery Pack | Cost | Power Supply(＞15Wh) | Dimensions(This will be reviewed in the final design) | Output Power( >1.25 W) |
| REV 16000 power bank | 60(can also be free for testing) | 59.2 Wh | 5.9*0.59*2.76 inches = 9.61 in^3 | 5V*2.4A=12W or5V*1A=5W |
| Ceeya Power Bank | 34.99 | 134Wh | 5.40 x 2.70 x 1.20 inches = 17.49 in3 | 5V*2.1A=10.5W |
| INIU Portable Charger | 19.98 | 100Wh | 5.3 x 2.7 x 1 inches = 14.31 in^3 | 5V*3A=15W |
Threshold and Alarm Systems
The alarm component of this device was considered a non-critical component for functionality. The Mini Metal Speaker w/ Wires – 8 ohm 0.5W was selected due to its wide range of compatibility with Arduino boards and low price. The final design of this device did not incorporate a on-device speaker system, instead alarms are triggered from the connected computer system to allow multiple device alarms to be monitored from one location.
Arduino Board and Camera
In order to control the lighting and imaging systems on the device an Arduino microcontroller was selected. The Sony Spresense was selected specifically selected for its compatibility with Sony Spresense Camera Board that had the imaging properties necessary for digital image analysis required for our inflammation detection.
| Arduino Board | Cost | Size | Ease of Use | Max Power Output |
| Arduino Uno | $18.40 | 68.6 mm x 53.4 mm | 2 | 250mW |
| Sony SpresenseMain board | $65.00($130 w/ Ext Board and Camera) | 50.0 mm x 20.6 mm | 3 | 220mW |
| Arduino Feather Huzzah | $16.95 | 50.0 mm x 20.6 mm | 3 | 1650mW |
| Camera Sensor | Cost | CCD/Global Shutter | Y/Cb/Cr Output | Resolution (＞752 x 582) | Ease of Use | Sampling Rate (＞1Hz) | Max Power Output |
| Sony SpresenseCamera Module | $35.00 | No | Yes | Yes | 3 | Yes | 555mW |
| Hi-Res 6.0mm Lens, PAL, Color DSP Board Level Camera | $137.50 | Yes | No | Yes | 1 | Yes | 1560mW |
| Blackfly S USB3 | $365.00 | Yes | No | Yes | 2 | Yes | 3000 mW |
| OpenMV Cam H7 – MicroPython Embedded Vision + Machine Learning | $65.00 | No | No | No | 3 | Yes | 495 mW |
| Foxeer Arrow Micro Pro 600TVL CCD FPV Camera w/ OSD | $19.90 | Yes | No | No | 2 | Yes | 750 m W |
| OpenMV Global Shutter Camera Module | $50.00 | Yes | No | No | 3 | Yes | 160 mW |
LED Lighting System
An LED lighting system was required when illuminate patient skin when collecting skin images for analysis. The White LED Backlight Module was selected due its uniform lighting area in order allow the skin area of interest to be visible without producing lighting artifacts. This module was also selected for its relatively low power consumption when compared to led lights with similar lighting properties.
| LED Light | Cost | Uniformity | Max Power Output |
| White LED Backlight Module – Large 45mm x 86mm | 3.19 | 4.5×8.6 cm | 3V*20mA=60 mW |
| SparkFun Electronics COM-10800 | 15.00 | 10×10 cm | 680mW |