BME100 s2014:T Group16 L3

From OpenWetWare
Jump to: navigation, search
Owwnotebook icon.png BME 100 Fall 2013 Home
Lab Write-Up 1 | Lab Write-Up 2 | Lab Write-Up 3
Lab Write-Up 4 | Lab Write-Up 5 | Lab Write-Up 6
Course Logistics For Instructors
Wiki Editing Help
BME494 Asu logo.png


Name: Ryan Hess
Name: Jossel D. Nkunzi
Name: Alireza Momeni
Name: Angie Chan
Name: Jospeh Sidman


Descriptive Statistics


Oral (Gold Standard)

Mean: 97.3 degrees Fahrenheit

Standard Deviation: +/-0.49 degrees Fahrenheit

Standard Error: +/- 0.001484848


Mean: 98.0 degrees Fahrenheit

Standard Deviation: +/-0.69 degrees Fahrenheit

Standard Error: +/- 0.002090909

Blood Pressure:

Blood Pressure Cuff (Gold Standard)

Mean: 120.1 mmHg

Standard Deviation:+/-8.54 mmHg

Standard Error: +/- 0.025878788

Watch Sensor

Mean: 125.1 mmHg

Standard Deviation:+/-12.57 mmHg

Standard Error: +/- 0.038090909


Pulse Ox (Gold Standard)

Mean: 87.1 beats per minute

Standard Deviation: +/-17.8 beats per minute

Standard Error: +/- 0.019606061

Watch Sensor

Mean: 84.9 beats per minute

Standard Deviation: +/-15.7 beats per minute

Standard Error: +/- 0.019818182







We used a T-test to compare the data collected from the bp cuff with the data from the watch sensor. We did this because there were only 2 sets of data that we needed to compare, and used a paired test because the data was from the same subject and therefore related. Our T-test value resulted in 0.2125 and our Pearson's R value was 0.4682. Pearson's R tests for the direction of correlation between two sets of data. Since our R value is between 0 and 1, it has a positive correlation, meaning the BP cuff and the watch sensor directly and positively correlate with one another's values. The t-test tests the significant difference between the two sets of data. Because 0.2125 derived from our data is well above 0.05, then we can conclude that the results from the experiment may not have been reliable. This means that the watch sensor is not reliable in comparison to the gold standard of the bp cuff.

Following the same logic as above, we used the same procedure to find the statistical similarity between heart rate as measured by the pulse oximeter and watch sensor. The T-test value was 0.0170 and the R value was 0.8279. While this data is much more statistically similar than the last test, the R value is closest to 1 (out of the numbers 0 and -1), resulting in a positive correlation between the two products. The t-test is less than 0.05, making the results reliable since the error is very insignificant.

Again, we followed the same procedure for our Temperature data from the oral thermometer and RAINN sensor, we there to be an R value of 95.91 and a T-test value of .00026. This means that the data was statistically similar enough to be significant, meaning the sensor is reliable when compared to the gold standard of the oral thermometer.


Issues and solutions

Problem: The underarm temperature sensor often lost contact with the area being measured and had to be held under the arm with uncontrolled pressure.

Possible solution: Instead of having to tape or hold the device, the RAIIN sensor could have a Velcro belt that would strap around the patients chest to ensure a firm and comfortable fit. This strap can help keep the device in place with constant pressure which could help avoid any error in the report that could have been caused by not exerting the correct pressure or the position of it. The strap can also have a tightening function in which the patient may adjust the pressure to the standard.

Problem: Data may have been skewed because every test subject had been performing many different activities which could include walking, sitting, or running. The class data may not be reliable because the subjects had not been doing the exact same exercises at the exact same time (time of day is important as well since the temperature and weather could affect results). People involved in more strenuous activities possibly had higher body temperatures than that of a person after sitting for several minutes.

Possible solution: The device could provide several settings that could manipulate the data and calculate the exercise along with the data. It could measure and display the regular data but also calculate and show the measurement in terms of a certain or standard exercise if the patient were to be doing that. For example, a patient walking about 2.3 miles per hour may have a body temperature of 97.4, then the sensor would display also the body temperature of the patient if he or she was actually walking 2.5 miles per hour. That way, everyone would have "walked" at the same speed. Another, more practical solution would be making sure the patients all begin and end the activity at the same time and are performing their task or exercise in the same speed and strength.


Target Population and Need

Target Population

Our product is targeted towards people who regularly do not have a sufficient amount of sleep. Our device can diagnose and monitor the amount and quality of sleep using Electroencephalography. The device can also be linked to the user's automated coffee maker to make the perfect amount of coffee to stimulate the user based on how rested they are. Usually, this technique can be costly and inconvenient as it requires the device to be perfectly and intricately assembled onto the patient. With our device, we can design the product with minimal parts in order to make it affordable and convenient for our customers.


Businesspeople, travelers, college students and other schedule-tight people are known to be on-the-go with their daily routines. Because they are constantly busy throughout the day, there are many times where sleep is their least priority. However, having little to no sleep is extremely unhealthy and could cause many problems such as losing alertness, trouble with memory, and fatique. Many people turn to caffeinated drinks to keep them going, but either take too much or too little to perfectly boost their alertness. With our product, we can easily and efficiently measure brainwave voltage during the patient's sleep in order to diagnose the the amount and quality of sleep. This way, the patient will be well aware of how they are damaging their health. Another component to our product is adding caffeine into the equation. The device will signal the coffee or tea machine and call to brew the right amount of caffeine with the drink depending on how well the patient slept over the night. This can help the customer drink just the right and healthy amount of caffeine for the day.

Device Design


Inferential Statistics

Table 1: Data Results of Patient Using Our Sleep Monitor and Electroencephalography Device (Gold Standard)

Time Elapsed (min) Sleep Monitor(Hz) Electroencephalography Device (Gold Standard) (Hz)
0 1 1
20 3 4
40 4 4
60 3 5
80 3 1
100 1 7
120 4 10
140 4 9
160 14 4
180 7 18
200 15 18
220 16 17
240 17 19
260 13 17
280 19 18
300 18 17
320 9 12
340 12 15
360 19 16
380 13 17
400 15 17
420 16 15
440 12 7
460 14 6
480 12 9
500 17 8

Table 2: Statistical Analyzations

' average stdev std error pearsons r t-test
Our device 10.80769231 6.059829904 1.211965981 0.6808649 0.687899446
EEG 11.19230769 6.020094556 1.204018911