OUTCOME OBJECTIVES:

Mobile computer assisted, beside

incentive spirometry can be used to

detect and calculate the frequency,

amplitude, and duration of inspiration.

METHODS

A clinical patient-centered model in an

academic research setting was used to

demonstrate that computer assisted

incentive spirometry could analyze the

frequency, amplitude, duration, and

calculated volume of inspiration. Sixty

inspiratory cycles were recorded from a

smartphone (n=30) and laptop computer

(n=30) utilizing an external microphone

and traditional incentive spirometer

(Voldyne 5000, Hudson RCI, Temecula,

CA). Multiple microphone-spirometer

positions were investigated. Recordings

were processed and analyzed utilizing

Matlab (R2012a, The Mathworks, Inc.,

Natick, MA, USA).

RESULTS:

All inspiratory cycles were captured and

recorded successfully. All smartphone

recorded inspiratory cycles did not

exceed 1.5kHz in broadband noise.

Increasing amplitude of inspiratory

sounds did not produce a significantly

increased flow over microphone. No

significant power difference was found

between the smartphone and computer

recordings when investigated 2-cm in

front of lips and immediately behind the

open-ended spirometry mouthpiece.

CONCLUSIONS:

Mobile, computer assisted recording and

analysis of inspiratory cycles is a low-

cost, and reproducible method for

detecting detailed characteristics of

inspiration. Future investigations of

portable and automatically implemented

signal processing may improve patient

care and post-operative recovery.

Utility of Mobile Computer Assisted Incentive SpirometryKaram W Badran, MD1; Chloe C. Krasnoff, BSc2; Zachary Taylor, PhD3; Maie A. St. John, MD PhD1,2,3

1Department of Head and Neck Surgery, UCLA Medical Center, Los Angeles, CA2David Geffen School of Medicine at UCLA, Los Angeles, CA

3Department of Bioengineering, UCLA, Los Angeles, CA

Sixty inspiratory cycles were recorded using the system

illustrated in Figure 1. Through an external microphone

(3.5mm EarPod) thirty samples were recorded using a

smartphone (iPhone 6s, Apple Inc., Cupertino, CA) and

the remainder using a laptop computer (MacBook Pro

2013). For each computing device, ten samples were

recorded at three locations from the user. Figure 1 (A)

illustrates a microphone 2cm away from the user’s lips,

(B) immediately behind the incentive spirometry, and (C)

within the spirometer’s chamber.

The dataset of audio samples were analyzed utilizing

Matlab (R2012a, The Mathworks, Inc., Natick, MA,

USA). The frequency domain representation of the

original signal was associated to a function of time by

use of Fourier transformation. The data graphed

represents noise amplitude, frequency, and duration.

INTRODUCTION

METHODS

RESULTS

ABSTRACT

Karam W Badran, MD

Department of Head & Neck Surgery

UCLA Medical Center

Email: kbadran@mednet.ucla.edu

CONTACT

DISCUSSION

Utilizing a microphone and headphone jack all

inspiratory cycles were successfully recorded and

analyzed (Figures 2-4). The frequency, amplitude, and

duration of inspiration were recorded. The maximum

recorded amplitude of noise identified through Fourier

transformation on iPhone (x=0.18 ± 0.02) was on

average less and differed significantly from those

recorded by computer (Figure 2, x=0.25 ± 0.02,

p=0.027) when recorded in front of the lips and when

placed inside the spirometer chamber (p<0.01). The

range of frequencies captured by computer (range: 0-

2kHz) at all microphone positions was wider than that

recorded by iPhone (Figure 3, range: 0-1.5kHz), however

all frequencies captured were less within the chamber

than when outside of the spirometer or within the

mouthpiece (Figure 4, p<0.01). Representative

measurements of air flow at 500 Hz, 1000 Hz, and 1500

Hz demonstrated variable amplitude at each microphone

position and method of recording.

REFERENCES

1. Overend TJ, Anderson CM, Lucy SD, Bhatia C, Jonsson BI, Timmermans C. The effect of

incentive spirometry on post- operative pulmonary complications: a systematic review.

Chest. 2001;120:971-978.

Smartphone computing allows for the efficient storage of

datasets readily available for computer-assisted analysis.

Our study demonstrates a low-cost and easily

reproducible system for recording inspiratory cycles.

Future work is required prior to clinical application.

Improvements include a frequency threshold and range

restriction in order to exclude extraneous noise (ie.

bedside alarms). Additionally, we expect that power and

flow would correlate with noise amplitude, a made more

significant in a reduced frequency range. With the

appropriate user interface, the smartphone incentive

spirometer has the ability to increase patient compliance,

reduce hospital costs, increase accessibility, and improve

patient care in the post operative setting.

Incentive spirometry is a near universal component of

medical and post-operative care. The goal of this

technique is to encourage patients to take deep and slow

breaths to assist in expansion of the lungs after surgery.

This is crucial as many individuals are at risk of atelectasis

secondary to pain and resultant shallow breathing.1

Conventional incentive spirometry is accomplished by the

use of a large device that provides the patient with a visual

feedback when they inhale for a minimum of 3 to 5

seconds for 5 to 10 consecutive cycles.1 The self-

administered device however, suffers from low patient

compliance. It is bulky, not user-intuitive, may induce

pain, and does not monitor patient use or progression.

Recently, advances in mobile and smartphone computing

have penetrated everyday personal and clinical practice.

With the increasing sophistication of software for these

devices, mobile, point-of-care detection of patient focused

goals are possible. Our study aims to assess the utility of a

smartphone assisted incentive spirometry system.

Figure 1. Workflow demonstrating a mobile and smart method to capture

inspiratory flow. A, B, and C indicate the three areas where flow past a

microphone are recorded

Figure 2. Recording from a MacBook Pro computer. The graphs above illustrate a single inspiration directly in front of the mouth. From left to right, the duration of

inspiration is graphed as a function of amplitude, frequency, and both.

Figure 3. Recording from an iPhone 6s. The orientation from left to right is the same as Figure 2. The middle graph illustrates frequency v. time. The hue increases in

intensity as amplitude increases for each frequency identified.

Figure 4. Recording on MacBook from within the incentive spirometry chamber demonstrates high amplitude though not in consistent frequency bands.