1 of 38

Traumatic Brain Injury Data Assessment Tool

Austin Dunn // Michaela Gobron // Lydia Lui // Jin Matsumoto // Mandeep Sandhu

2 of 38

Image from : http://klandrylaw.com/wp-content/uploads/2012/11/Bicycle-Accident.jpg

Image from : http://nyulocal.com/wp-content/uploads/2014/01/football-head-shot-collision.jpg

Image from : http://4.bp.blogspot.com/-9P4pyUQXk-s/UQJTTtTJNtI/AAAAAAAAQMo/BiZtl62vqGI/s1600/Elderly_Man_Falling_Down.jpg

3 of 38

Background

4 of 38

Traumatic Brain Injury (TBI) - What is it?

Disruption in normal brain functions

Mental Status
Consciousness

Adverse Effects on

Memory
Emotion
Physical Ability
Sleep

Image from: http://epilepsyu.com/wp-content/uploads/2014/07/TBI.jpg

5 of 38

Motivation

Contributes to 30.5% of injury related deaths in U.S.
2.5 million people (children and adults) suffer from TBI a year in the U.S.

Image from : http://pediatric-house-calls.djmed.net/wp-content/uploads/2013/07/concussion-head-wrap.jpg

50,000 result in death - 137 deaths a day
80,000 irreversible injuries

Every 13 seconds someone sustains TBI

5.3 million Americans live with a TBI related disability (1 out of 60 Americans)

6 of 38

Current Practice

7 of 38

Current Practice

Continuous monitoring to ensure parameters remain within desired ranges

Core Temperature

Fever = longer ICU stay, dangerous in combination with other factors

Mean Arterial Pressure (MAP)

EACH - 10 mmHg = + 20% mortality rate

Intracranial Pressure (ICP)

> 20 mmHg = significantly worse outcome

Partial Pressure of O₂ in Brain Tissue (P_btO₂)

< 8-10 mmHg = high risk of ischemia and increased mortality
< 7 mmHg associated with risk of death, regardless of duration

Image from : https://i.ytimg.com/vi/IvWOwzk0H7g/hqdefault.jpg

Michaela

These four parameters have been shown to be important to a TBI patient’s help. It is of most importance that these variables stay within a predetermined range, for an increased time outside these ranges increases the likelihood that the patient will suffer from secondary injuries or an increased stay at the ICU. The four parameters of interest are Core Temp, Mean Arterial Pressure, Intracranial Pressure, and Partial Pressure of Oxygen in Brain Tissue. Examples of the different parameter ranges can be found on this slide.

Picture = adolescent patient with sensor attached to his skull for ICP monitoring

reference :

ICP decreases sharply between 36° to 35°C, but did not significantly change between 35° and 33°C
ICP 20 to 40 mmHg increased mortality by 3.5-fold and ICP above 40 mmHg increased mortality by 6.9-fold; raised but reducible ICP increases mortality by 3- to 4-fold
P_btO₂strongly correlated with good outcome (P_btO₂ > 35 mmHg), moderate or severe disability (P_btO₂ 26 to 35 mmHg), and a vegetative state or death (P_btO₂ < 26 mmHg).33 P_btO₂ is a very strong predictor of outcome...a worse outcome if P_btO₂ was below 10 mmHg for over 30 minutes; a 50% risk of death was found in patients with P_btO₂ below 15 mmHg for 4 hours.

Increased times outside these ranges increases the likelihood that the patient will suffer from secondary injuries (what are these secondary injuries?)

8 of 38

There is no current treatment for TBI patients.

The best option is MONITORING these critical variables to improve the quality of care.

9 of 38

Problem Statement

10 of 38

Standard bedside monitors do NOT have a system that displays data in a format that allows healthcare professionals to quickly analyze fluctuations of critical variables between ideal, marginal, and critical ranges.

.

11 of 38

Current Standards

Lydia

As seen here, the current method of data analysis is quite inefficient and difficult to discern. Nurses have to manually enter the data in intervals of about two to three hours. If they’re lucky, this may result in about three data points every two hours.Not only is this incredibly arbitrary, it is also extremely inconsistent. If you’re a healthcare provider walking down the hallway and checking in on patients, there is absolutely no way you can determine what’s going on with the patient- much less make informed decisions later on using the information seen here.

Jeff notes: Move this up! Provide context from

Show beginning and end to bring it around / connect everything

Full circle - also don’t put it with real time graph? (M)

Where is this from? Elaborate more what it is. Hospital program?

What they do on their own computers, not really a standard

What do you mean by “from nurses”?
Emphasize that there is no uniform spacing, vague time frame for collecting/entering data, - emphasize why this is bad
Not comparing it mp70 (real time data monitoring) anymore but “current standards” (J)

Only “one step up from pen and paper”

12 of 38

Thus...

In order to more effectively monitor tbi patients, our client has requested a new user interface that better displays the data from his traumatic brain injury patients.

13 of 38

Objective

The proposed device must graph critical Traumatic Brain Injury (TBI) variables utilizing a graphical user interface (GUI) with a green-yellow-red color scheme to display every change-of-state from the ideal, marginal, and severe regions so healthcare providers can easily access if and for how long the variables deviate from their prescribed ranges.

14 of 38

Device Goals

Features to assist in data interpretation

Clear interpretation of variable status
Set default ranges for crucial TBI variables
Ability to customize ranges for various patients
Simple and intuitive navigation

15 of 38

Design Concept

Data Input: USB Drive

Hardware: Raspberry Pi 2

Software: Processing Program

Display:

9 inch Touchscreen

16 of 38

Touch Screen Module

USB Drive: Data In

9” Touch

Screen Display

Power

HDMI Cable

17 of 38

Data Acquisition

Data Input: USB Drive

Hospital Monitor: IntelliVue MP70

Third Party Application: MediCollector

18 of 38

Design Solution

Austin

Once the program is executed, the user is prompted to select a file to anaylze. The user may select either a text or csv file.

Depending how many data sets the file contains, the user selects the proper number of parameters.

Depending on the type of data, the user selects the proper parameter to display. The selected parameter contains default critical ranges, depicted by the red, yellow, green shadings on the right image.8

Additionally, the user has the option to manually change the critical ranges because every patient is different. Every patient will have unique ranges depending on characteristics such as obesity, diabetes, etc.

We incorporated several human factors engineering concepts by grouping similar functional buttons together. For example, in the parameters section, all buttons are grouped in the middle. Additionally, we have have a left to right number system to show how the user should progress. Finally we colored the different buttons corresponding to their function.

Once the green “Analyze” button is pressed, the device will generate the plot.

As you can see, the plot clearly depicts the ideal marginal and critical ranges with the green, yellow, red color scheme and point indicators so that healthcare pAZrofessionals can quickly and accurately make diagnoses.

19 of 38

20 of 38

Design for Manufacturing

21 of 38

Lydia

The entire device is then physically housed in a delrin shell with a stainless steel handle.

Following the principles of green engineering and design for manufacturing, the device is designed to use a minimal amount of materials, while maintaining durability and reusability, meaning that the device will have a long lifetime of use. Simplicity was a central theme throughout. The number of different materials were drastically reduced to cut waste and increase cost-efficiency. Delrin was selected for being lightweight, durable and long-lasting, while kerfed stainless steel was used for the handle because of its strength and hardness.

Show box itself during presentation?
Simplicity, from one sheet - green engineering
Cheap, minimal materials
Design for manufacturing
(design!) smooth, seamless fit
White piece - for clarity - client to tell difference between black and white as well as help double wall insertion to slots
Internal walls to keep electronics organized

22 of 38

As you can see, the simple geometric features of each part ensure that the interference fits are reliable and have no need for other fixtures. The design completely opts out of the use of small parts such as screws or nails. It’s not unlike a jigsaw puzzle that has been reduced to just enough material to house the tight placement of the screen and electronics. Both major components were designed to be mass produced and laser cut in one step from a single sheet of material, saving both time and resources that would have otherwise been spent on multiple manufacturing steps.

Reference:

Lastly, you may notice that only a single one of the parts are manufactured out of white delrin. While we don’t intend for the client to ever need to disassemble, we still prefer to prepare for such possibilities. The contrasting color is to ensure that the customer is able to reinsert the more fragile screen in right direction,

23 of 38

Verification

24 of 38

Design Goal Assessment

Metric	Marginal Goal	Ideal Goal	Verification Method	Output
Weight	<15	< 12 Pounds	Measurement verified on scale	6.8 Pounds
Volume	< 2304 in³ (16” x 16” x 9”)	< 1536 in³ (16” x 16” x 6”)	Measurement calculated through dimension of case	358 in³
Memory Size	4 GB	4 GB	Observational	8 GB
Start up/ Calibration time	< 180 Seconds	< 60 Seconds	Timing the process from turning on device to start of program	39 Seconds
Cost	~2000 US Dollars	< 2000 US Dollars	Observational	684.36 US Dollars
Font Size	25 Points	72 Points	Observational	30 Points
External Power Consumption	300 Watts	< 300 Watts	Calculated by using the known voltage and current input (P=VI)	29.5 Watts

25 of 38

GUI Features

Clear Interpretation of Parameter Status

Color coded data points

Quick Trend/Pattern Recognition
High Resolution
Compatible with a .txt or .csv file

26 of 38

Validation

27 of 38

Validation: Survey Results

Survey dispersed to numerous healthcare professionals, including our client. From the survey responses, we changed the design output.

“Looks Awesome!”

Kia Shahlaie, M.D., Ph.D., Neurosurgeon, Assistant Professor, Client

Simple

Intuitive

Manny

We started off with the graphical representation of the data because this was the core of our client needs, and we went through several iterations of graphs. Starting with the graph on the right, the client liked the highlighting but he felt that at certain areas were misleadign the representation of the data. For exmaple if you look at the red its ambiguous if the range is actually in red or not. So we sent him another graph with a more simple display and he liked it, but then we took it a step further and incorporated color coded dots so there is no confusion on status of a data point and he really liked. He said and i quote “Looks Awesome!” We were also able to confirm that this representation was both simple and intuitive.

Explain how well we executed the value added

Add surveys for testing

Incorporation of suggestions - beauty of software project!

Chart - mimic verification table for questions:

I.e. Is this device intuitive / usable in a hospital environment? *checkmark*
Sell this

28 of 38

Validation Continued

“Extremely helpful feature! Normalizing the variables to each patient is a feature the current monitors lack. Very impressive.”

William Hammontree, R.N., B.S.N.

29 of 38

Value Added - Impact on Patient Treatment

Clear interpretation of the data impact decisions on treatment plans

Track Progression
Locate Trends

Customizable ranges for all parameter

Personalized Care
Accurate Representation

Variable Modularity

Multiple Intensive Care Units

30 of 38

Future Improvements

Expand the preprogrammed ranges to more than the TBI-related ones
Save custom ranges created by healthcare providers
Access patient data directly from hospital database
Total time elapsed in each time zone displayed

Integration to show duration

Display More Graphs

4+ Variables on multiple screens
Zoom and Scroll Bar Features

Exportable Graph
Battery Powered
Real Time data graphs from the MP70

Michaela

Our device, being very software heavy, has the potential to be upgraded to include a multitude of upgraded features.

In future iterations, we hope that this data analysis tool will be modular, allowing for other hospitals and intensive care units to monitor variables pertaining to the injuries or diseases they monitor, whether that be heart rate or respiratory rate. Also, we would like for the software to be able to save and store custom ranges inputted by the healthcare provider. The ranges preprogrammed might need to be altered due to a patient’s physical characteristics like height or weight. We would like for our device to be able to bypass having to input a flashdrive to gather saved data, but grab it on the fly from the hospital databases. Time outside the ideal ranges are critical, so we hope to be able to display an ellapsed time timer for each variable being monitored. We would even like the user to be able to visualize however many graphs they would like, and to show them on mulitple screens so that the resolution can remain pristine. We would also like these graphs to be exported so the data is readily available elsewhere. Perhaps even a more standalone device without any cables, where it runs on battery power. And last, but certainly not least, a future improvement is that our device can graph real time data directly from the MP70 current bedside monitors

31 of 38

Future Works: Real Time Data

32 of 38

Future Improvements Continued

Continuous Monitoring beside the MP70 for real time treatment methods

33 of 38

There is no current treatment for TBI patients

Therefore, the best option is MONITORING them to improve their quality of care

34 of 38

Questions?

35 of 38

Appendix

36 of 38

Motivation/Background

Problem

Objective/Design Solution

Features

How Design Works

Assessment of Solutions

Surveys

Assembly of Device

Value Added

Big Reveal

REAL TIME DATA - TEMP SENSOR

Future Works

37 of 38

Temperature Sensor

38 of 38

Material Properties

Material	Tensile Strength, psi	Rockwell Hardness	Impact Strength, ft.-lbs./in.	Coefficient of Friction	Dielectric Strength, volts/0.001	Water Absorption, %	Density lbs./in.3	Thermal Expansion, in./in./° F
Delrin® Acetal Resin	9,000- 11,000	M89-M94	1-2.4	0.2	435-500	0.2-0.4	0.051	4.7 to 12.2^10–5