Functional block diagram and other fun things!

Winslow made this lovely block diagram for our device!

In one of my favorite colors, sea-foam green! Muahaha.

The gait sensor remains on standby until the user indicates that they are ready to use the device through pressing a pushbutton. Once the pushbutton is depressed, the microcontroller turns on the red LED to indicate that it is calibration to calculate the zero values for the accelerometer readings in the X and Y directions for a straight foot orientation. Once the calibration is complete, the LED turns off and the SD card reader will respond to the Arduino indicating that it is ready to start logging information. Afterwards, the microcontroller will start logging the sensor outputs.  

Firstly, a total of 4 analog signals from the X and Y acceleration readings (2 accelerometers) are captured by the integrated ADC converter found on the Arduino board used here. Secondly, each pressure sensor output is separately converted into a pseudo-analog signal through a RC circuit created with a 0.1 uF capacitor and variable resistance from the bend sensors. Each of the six pressure sensors signals are appended to the captured analog signals from the accelerometers. Then this combined data string is then written onto the SDHC 8 Gb card through the SD card shield. The SD card then replies that the recording was successful, and this process repeats every 300 milliseconds. 

After the trial duration has ended, the user inputs another button press and the microcontroller stops logging data. This SD card can then be removed from the device and placed into a card reader for a computer to read. A Matlab GUI, which is our software component, then reads the text file generated on the SD card and shows the changing orientations of the feet based on the accelerometer readings and the changing pressure distribution on the bottom of the feet through the pressure sensors with respect to time.

We should probably provide a specification list as well... here are some tables for your perusal:

Specification	Value, Value range, or quality relating to specification.	Comments about or rationale for specification
Battery Life	24-48 hours	Patients will generally need to record a maximum of one full day of activity. Physician replaces battery; 48 hours preferred in case of longer desired recording times.
Foot sole pressure distribution mapping	v Variable-resistance bend sensor specifications (by Flexpoint Sensor Systems, Inc.) v 6 used (3 for left foot, 3 for right foot) v resolution minimum 1° of bend, 0.1° ideal v resistance range 5kΩ-14kΩ*	*Sensor areas can be custom-fit to insert as needed. Resistance range adjustable by manufacturer; may need different ranges depending user weight. The provided values are just the values of the standard sensors that they sent us in the sample lot. Sensors arranged length-wise across insert surface so that they are bent during compression of the insert base by the curvature of the foot.
Foot orientation tracking	v Triple-axis accelerometer breakout board specifications (by Analog Devices) v 1 accelerometer used per foot v resolution of 30°	Unfortunately, appears to be a previous iteration of current board being offered (ADXL335), but specifications should be identical. New accelerometer will be used in revised designs.
Long-term data logging on removable media	SD card, 2 GB or SDHC card, 8 GB	2 GB minimum, 8 GB ideal. We found that 2 GB would be sufficient to record roughly 24 hours worth of data, but considering the relatively low additional cost of upgrading and the possibility of multiple recording sessions, we find that 8 GB would be a good option.

Feature-benefit table:

We think these features should be sufficient to address the major needs of physicians and podiatrists who are interested in monitoring gait patterns: in general, they want to know the heel- and toe-off patterns over time (the pressure distribution in these regions as the patient takes steps) and the use of "high" vs. "low" gear propulsion (which metatarsal heads--1st and 2nd, or 2nd through 5th--are being used to generate most of the propulsion during gait). These simple measurements would easily be captured by our device, and focusing on addressing the most-needed features would reduce the final costs (materials, software development, testing, etc.).