Multisource Data Collection for Sleep Disorder Studies (Widener University)

Daniel Santosusso, Elvin Izaguirre, and Michael Malkoun

ABSTRACT:

The goal of this project is to develop a portable and affordable device to collect multisource physiological and environmental data for sleep disorder study. Millions of Americans suffer from various sleep disorders with very few of them being able to afford or even have a sleep laboratory nearby to seek diagnosis and treatment. These labs are very expensive and health insurance companies usually do not cover the cost. In addition, a sleep disorder diagnosis requires the patient to stay from one to multiple nights at the laboratory, which cannot sufficiently emulate the patient’s home environment, comprising the accuracy of diagnosis. Theproposed device uses open-source hardware, mainly the Arduino, which is based on an ATMega microcontroller, to monitor and record data in real-time. A graphical user interface (GUI) was designed to help medical professionals visualize and interpret the diagnosis results. A prototype device with supporting software has been successfully implemented and tested. It can collect and transmit room temperature, ambient light in the room, and on-body temperature. In the future development of this device, more sensor modalities will be added to the device, and the power efficiency will be optimized.

BACKGROUND:

Developing a sleep disorder is a growing problem in the United States. Diagnosing these disorders is currently a very expensive and aging process. A typical sleep disorder treatment requires that patients travel to an unfamiliar sleep center, spend one night or longer, and be evaluated by a medical staff. This presents multiple challenges to collect accurate data. First, the patient is in an uncomfortable situation and asked to act normally. Second, the patient has many wire leads connected to them while they sleep. This is not a natural feeling and may cause inaccuracies in the results. Third, the sheer cost of using a sleep center is an issue. The average cost to a patient, per night, in a sleep center is $3000.00. This is a steep cost that some medical insurers only cover partially or do not cover at all. This leaves a lot of the cost burden on the patient to bear.

PROBLEM STATEMENT:

The goal of this project is to design a device that will replicate the functions of a sleep center.This device would have to be compact and light enough to be carried easily as well as user friendly and not too complicated. The device will be useful if it can collect physiological data wirelessly and collect the data for future analysis. The device should also monitor environmental conditions of the patient’s surroundings. The wireless transmission will free sleepers who may want to change positions during the night. The device can acquire multisource physiological and environmental data based upon which healthcare professionals may provide treatment solutions. Ambient temperature, ambient light and body temperature will be the only factors considered for the scope of this project. Sensors to measure these factors will be the only ones used for this device. The perfection of these sensors will lead to the expansion of the devices capabilities in the future. It was determined that no one has begun prototyping a device quite like ours: A device to collect environmental temperatures as well as physiological data.This product could one day make peoples’ lives easier by helping them get the care they need.

METHODS/APPROACH:

After speakingwith medical professionals at sleep centers, the design was proposed as follows:
The device will utilize Arduino Hardware for functions. The reasons Arduino hardware was selected are that it is readily available, cost effective, user friendly, and open source. This means replacing parts will never be an issue and usability will not be a problem either. The equipment used is open source with worldwide support. The Arduino microcontrollers use the ATMega328 processors to process our data and send to the computer or storage device for later evaluation. A computer will be connected to the Arduino and a program was developed that will collect the sensory data in real time and save it to a file with a timestamp attached with each sample.
In order to alleviate the issues with wires hindering a person’s normal sleep habits, wireless communication is necessary. The XBee Series 1 RF modules was utilized to implement the wireless communication between the processing computer and the on-body sensing modules. These wireless communication chips are designed for low power consumption as well as a short range of secure data transmission.

Hardware Implementation

The electronics research concluded with the selection and procurement of the following components: Arduino Uno, Arduino LillyPad XBee, Arduino XBee shield, LillyPad, LillyPad power supply, Series 1 (S1) XBee RF modulces, MCP9700 thermistor, NTCLE100E3 thermistor, GL5528 Photocell.

A hardwired connection was made between the input of the Arduino LillyPad and the MCP9700 thermistor. One of the XBee RF modules was then connected to the designated location on the LilyPad, Pin D0. The LilyPad was powered by lithium ion power supplies. Another XBee was also connected to the Arduino Uno board using the Arduino XBee Shield. This shield allows for easy connection of the XBee Series 1 module to the inputs and outputs of the Arduino Uno while still allowing access to all other input/output pins of the Arduino Uno. This is important as the environmental sensors utilize these ports during operation.
The GL5528 photocell and the NTCLE100E3 thermistor used for ambient condition monitoring were both placed on a breadboard with their respective wire leads connected to the Arduino Uno analog input ports A0 and A1 as stated previously. A diagram of the pin connections for the Arduino UNO can be seen above as Figure 1.

Software Implementation

In order to ensure the correct configuration of the Arduino, the Arduino IDE was used to program each microcontroller in the network. A computer program was also written based on the Arduino software to perform data acquisition for the on-body and environmental sensors. Analog pin 0 on the Arduino Uno board recorded a voltage that was converted into temperature using a formula found on the data sheet for the thermistor. A suitable program was also written for the photocell in order to monitor the changes in ambient light intensity. The wireless transmission between the XBee S1 RF modules and the microcontrollers on the network was successfully implemented as a result of the execution of the code that was created in the Arduino software.
The serial port of the Arduino Uno was used to collect raw data from the on-body and environmental sensors. A USB cable connected to the Arduino Uno board and a computer allows for the hard wire transfer of the collected data. This data is then compiled on this computer preloaded with the Arduino programming software. After a successful compilation, the data collection is then exported to a file written in Java. Java takes the data and saves it to a comma separated values (.csv ) file. This file is comma delimited so MATLAB can interpret the data. The MATLAB software then generates a Graphical User Interface (GUI) from the Java file, which can be used to easily interpret and analyze the results.

RESULTS:

The device that was set out to be made has been successfully completed. This device monitors both environmental conditions as well as physiological indicators as well. The device is affordable and portable, just as desired. The device also records physiological data wirelessly. Overall, the device was a success and the data being collected is accurate. The following details the actual results the device will obtain whenever operational.
The data collected were on-body temperature, room temperature, and ambient light in the room. The results were collected from the subject who would run a program that was written in Java that collects the sensor values and timestamps them, the run dialog is seen in Figure 2. Note also within Figure 2, the outputted .csv file is shown. The program then saves the data to a .csv file. This .csv file is then turned into a graph using MATLAB, Figure 3. The program has a graphical user interface (GUI) which makes it very easy for the medical professional to use the program. This program compiles data and runs as intended. A screenshot of this program is shown in Figure 3.
The data is then streamed into a GUI from the .csv file once the run button in Figure 2 has been pressed. This program imports the .csv file that was created and a medical professional can analyze all the data that was collected. This program was written in MATLAB. This program also provides a GUI for easy usage. This program can plot up to three different sensor data streams. A screenshot of the program can be seen as Figure 3.

DISCUSSION/OUTCOMES/FUTURE WORK:

While the device was successfully implemented to meet the original goals of our design there are a few areas in which improvement to our design should be made. Additional sensors will be added to the current framework. For instance, the addition of a Rapid Eye Movement (REM) sensor will allow for medical professionals to tell for certain whether or not a patient is getting the recommended amount of sleep. Also, sensors that could measure the blood pressure and oxygen levels in the blood would be highly recommended for functionality to be the most akin to a sleep center. The battery life is still being optimized in order to hit the 10 day to 2 week time frame initially recommended.The implementation of a wireless shield to make the device self-sustainable is also recommended. This would allow the data collected to be stored on line or on a “cloud” network drive for future analysis. Also, the testing and evaluation sections of our project have lead us to believe that the next version of the device should look at the Series 2 XBee wireless transmitters.

COST/IMPLICATIONS:

For under $200, a device was constructed with an Arduino Uno board, an Arduino Lilypad chip, Arduino accessories, and a few sensors. This device can be used to collect data overnight to help study sleep disorders and this is a very minimal cost compared to spending one night in a sleep center where it may cost upwards of $3,000. Of course, in order to get the data to the doctors and for their professional opinion, that may cost some extra money but much less than $3,000. This device would help a lot with regards to studying and diagnosing sleep disorders for those who need it by making the process much more affordable.

, , ,

Powered by WordPress. Designed by WooThemes

Skip to toolbar