Pearson's R
Heart Rate: 0.192798089
Temperature: 0.690806489
Design Flaws and Recommendations
Design Flaws and Recommendations
A flaw in the design of the Spree band was that it was measuring heart rate on the head. The most accurate way to measure heart rate is in the chest or on main arteries where the pulse is the strongest. This design flaw also affected the measuring of the body temperature. Body temperature is usually measured from the mouth or the underarm. The head is not typically used for body temperature measurement. The data supports this because there was a statistical significance to the golden standard. The Pearson’s r value for the heart rate is 0.691 and the Pearson’s r value for temperature is 0.193. Both of these values are very high, indicating a low correlation between the values.
Experimental Design of Own Device
Our Experiment
Our experiment will include different parts that will all test the technological components of our device. Each individual experiment will be run at least three times, with multiple subjects in each trial. In these experiments, our technologies will be tested against an accurate current technology or device.
Experiment 1- Heart Rate Monitor:
To determine whether the heart rate monitor in our watch is accurate, an experiment similar to the one we analyzed in class will be performed. The heart rate data from our watch will be compared to the the heart rate data from a reliable heart rate measuring device, the pulse oximeter. Inferential statistics will need to be performed in order to compare the data from our heart rate monitor and the pulse oximeter. The mean, standard deviation, and Pearson’s r value will need to be calculated in order to compare the data. Optimal results would consist of values with a low variance and small Pearson’s r value (p<0.05).
This experiment will need to be run with three separate groups, each group participating in two trials. As testing heart rate is a noninvasive procedure and does not take much time, the cost of trial will be a lot cheaper than most medical trials. For these reasons, a sample size of 30 people per testing group seems feasible.
The experiment will require a total of 60 minutes per patient, divided into three subunits. At the beginning of the first 20 minutes, the resting heart rate of the patient will be tested using the pulse oximeter and our heart rate measuring device. The heart rate will be measured every 2 minutes throughout the 20 minute period using both the pulse oximeter and our heart rate monitoring device. In the next 20 minutes, the patient will rest in order to restore resting heart rate. In the last 20 minutes, the experiment performed in the first 20 minutes will be repeated. Subjects of the experiment can be tested in constant rotation, maximizing our use of time and reducing the time spent on the trial.
Experiment 2- GPS Accuracy:'
In order to test the watches GPS accuracy several tests would need to be performed in a variety of situations. We would run a random test between two people. One person wearing the watch and another tracking the the watch. The watch wearer would drive, walk, or run to a random location not known to either of the participants and take their exact longitude and latitude with one of the most accurate positioners on the market. Then the other participant would go to the location the watch GPS shows and take their exact longitude and latitude with an identical device. Doing this with 50 people and putting them into 25 random pairs, one person would be the runner with a given random location and the other a tracker. The average distance between the actual longitude and latitude vs the GPS longitude and latitude would give the accuracy of the watches GPS.
Experiment 3- Watch to App Connectivity (response times):
The ability for the caretaker to set reminders for specific times and receive a timely response from the patient via the watch is an essential feature of our design. It is important the the caregiver is able to receive real-time data from the patient, as the purpose of the watch is to reduce the amount of time and stress the caregiver expends and increase the independence of the patient. If the watch to app connectivity fails, the entire function of the watch fails.
The watch and the app will both connect to a cellular network in order to interact with each other. This connectivity and its accuracy are being tested in this experiment. Similar to our second experiment, 50 people will be put into 25 pairs of a caretaker and a patient. These pairs will be in different locations. The patient will test the watch functions by pushing the panic button, setting alarms, and receiving voice memos. The time that it takes for the caretaker to receive these notifications are essential, and will be recorded multiple times. Ideally, the data will be received immediately by both parties.
BME 100 Spring 2017
