Volume 10 Issue 04 Published, November 9, 2006 ISSN 1535-864X DOI: 10.1535/itj.1004.

06

Intel®
Technology
Journal
Autonomic Computing

A Self-Managing Framework
for Health Monitoring

More information, including current and past issues of Intel Technology Journal, can be found at:
http://developer.intel.com/technology/itj/index.htm
 A Self-Managing Framework for Health Monitoring
Amit Baxi, Corporate Technology Group, Intel Corporation
Nagaraju Kodalapura, Corporate Technology Group, Intel Corporation

Index words: Body Area Network, Zigbee, patient monitoring, autonomic, self-management, power
management, fail-safe mechanisms, biometric authentication

systems would be able to autonomically handle several

ABSTRACT alarm conditions for efficient operation, would require
Existing medical devices and equipment monitor less intervention by doctors and would reduce errors
abnormal physiological conditions and trigger audio- related to patient management.
visual alarms. However, these devices require high levels
In this paper we describe a self-managed generic
of interaction by a doctor since they do not have any self-
framework for digital health monitoring that can be
managing capabilities to reduce user intervention or to
applied in different use cases. The self-managing
autonomically handle alerts.
capabilities of such a framework are useful to efficiently
Recent advances in low-power wireless communication manage several tasks that usually require user
technologies have led to the development of small form- intervention.
factor, wireless, body-wearable biomedical devices for
We believe that such a system design will enable medical
health monitoring [1]. This enables patient mobility and
device manufacturers to design their biomedical front-end
comfort and enables continuity of care from hospital to
devices to be compatible with a generic framework and
home. In this paper we propose a self-managing
this will drive standardization of hardware and software
framework for health monitoring using body-wearable
interfaces for medical devices.
bio-devices [2] that can reduce the doctor’s intervention
for patient management. This will enable doctors to Moreover, this framework utilizes the advances in
effectively manage a greater number of patients and Personal Computer (PC) and server technology to provide
reduce the number of errors inherent in paper-based reliable, generic, cost-effective and efficiently self-
processes. managed patient monitoring solutions.
We illustrate three usage models where this self-managing
framework can be applied: remotely monitoring patients at
home, monitoring fetal well-being in a maternity ward,
and monitoring critical patients in an Intensive Care Unit
(ICU).

INTRODUCTION
There has been a significant development in bio-medical
instrumentation over the last two decades. A number of
high-tech medical diagnostic and therapeutic systems are
used by doctors for diagnostics, for monitoring critical
patients, and for delivering treatment. Such devices detect
abnormal physiological states, and they trigger audio-
visual alarms when the limits are crossed. However, they
require frequent intervention by doctors to manage these
alarms.
If such patient monitoring systems could be equipped with
some intelligence and self-managing capabilities, the

A Self-Managing Framework for Health Monitoring 313

Intel Technology Journal, Volume 10, Issue 4, 2006

Figure 1: A conventional multi-parameter patient monitor

Patient monitors have biomedical signal processing
OVERVIEW OF EXISTING PATIENT algorithms for detecting abnormal physiological
MONITORING SOLUTIONS conditions such as cardiac arrhythmias, apnea, low blood
Patients in the Intensive Care Unit (ICU) are usually wired pressure, etc., and the monitor usually gives an audio-
to a bedside patient monitor known as a Multi-Parameter visual alert when such a condition is detected. The doctor
Patient Monitor (MPM). The MPM monitors one or more usually intervenes when an alert is generated and either
of the patient’s vital physiological signals such as silences the system or takes corrective action to treat the
Electrocardiogram (ECG), Blood Oxygen Saturation patient. Apart from alerts generated by physiological
(SpO2), Non-Invasive Blood Pressure (NiBP), Invasive conditions, other alerts are also generated to notify system
Blood Pressures (IBP), Respiration, Temperature, Airway status or malfunctions such as a low battery condition,
Gases, etc. Patient Monitoring Systems (PMS) are usually sensor coming off the patient, etc. Since a doctor
dedicated embedded systems, each with its own sensors, managing an ICU has to respond to alerts from several
biomedical front-end hardware, LCD display, keypad, patients, it becomes an uphill task for the doctor to
integrated thermal printer, communication interfaces such manage alerts, if the patient monitoring system does not
as RS232 or USB, and battery backup. These systems are have the capability to self-manage such alerts.
quite expensive and thus contribute to the overall high Our proposed framework addresses these issues by
costs of healthcare. These dedicated embedded systems embedding intelligence at various levels in the system, in
have proprietary hardware and software. The biomedical order that the system can manage itself and only require
Original Equipment Manufacturers (OEMs) avoid the use intervention by a doctor for critical physiological alerts.
of general-purpose PC platforms for critical applications
since the PC’s hardware, software, and operating system As seen from Figure 1, the Computing and Results
(OS) are not optimized and reliable enough for medical Display Block (CRDB) in a typical MPM is quite similar
applications. A typical MPM is shown in Figure 1. to that of a general-purpose PC. Hence, it makes sense to
functionally separate the CRDB from the integrated PMS

A Self-Managing Framework for Health Monitoring 314

Intel Technology Journal, Volume 10, Issue 4, 2006

and use a single PC to cater to the computing and results multiple patients at a central location. Each body-wearable
display requirements of multiple patients. sensor, known as a Bio-Front End device (BFE), is
composed of a sensor for bio-signal sensing, the related
A SELF-MANAGING FRAMEWORK FOR front-end hardware, a low-power microcontroller for data
HEALTH MONITORING acquisition and a wireless transceiver for data transfer to
the receiver. There can be multiple BFE devices
Architecture connected to a patient for monitoring multiple parameters.
The evolution of low-power wireless communication An aggregator (AGG) device worn by the patient performs
technologies like Zigbee (IEEE 802.15.4), low-power the function of receiving the wirelessly transmitted data
Bluetooth*, etc. has enabled the development of small, from multiple BFE devices connected to a patient and
body-wearable, wireless sensors for patient monitoring. transmits the aggregated data to a backend PC or server.
The AGG device can be a device like a Portable Digital
As shown in Figure 2, these sensors can be configured to Assistant (PDA) or a scaled down version of that without
form a Wireless Body Area Network (WBAN) [3] and a Liquid Crystal Display (LCD). The AGG and the BFE
enable monitoring of multiple bio-parameters (such as devices form a localized WBAN, with each BFE device
ECG, Pulse Oxygen saturation, Blood Pressure, etc.) of having a unique device ID.

Figure 2: Architecture of a body-wearable, wireless health monitoring platform

Following are the advantages of this architectural • There is better resource utilization since a single
framework: resource (like a network printer) can be shared across
multiple patients.
• The functional partitioning of the PMS allows a
single PC to cater to the computing and results • It enables patient mobility and comfort.
display requirements of multiple patients, thereby
reducing the cost of the overall system. • It allows central monitoring of multiple patients from
a single location.

A Self-Managing Framework for Health Monitoring 315

Intel Technology Journal, Volume 10, Issue 4, 2006

• It allows the entire patient-monitoring framework System Components in the Health

(consisting of WBANs of multiple patients) to be Monitoring Framework
managed from a central location.
There are three major components in the health
• Such a framework does not require the doctor to be in monitoring framework as shown in Figure 3:
close proximity to the patient or to the monitoring
equipment, since the patient’s physiological data can • The BFE device.
be made available to the doctor anytime and • The AGG.
anywhere on his hand-held communication device.
• The backend computational platform such as a PC or
server.

Self-Adjusting GUI Display,

data storage, BFE Device
Control, Alert Management,
Remote Alert generation
Remotely
networked devices
like PCs, PDAs,
Local Display Mobile phones etc.
(optional) A
Application Layer
P
M
L Networking layer Shared resources
S Application Layer A on Network like
E Y printers, scanners
Data formatting & etc.
L E
Networking layer Encryption
F R
M
A Patient Devices
Data formatting &
N Management
S Encryption
A
E G
Application Layer
L E Phisical Layer
BFE Device
F M discovery,
M Networking layer E Identification & WBAN
A N management
N T PC / Server
A Data formatting & Level 3
G Encryption Phisical Layer
E
M
E DAQ and Control DAQ – Data Acquisition
Aggregator
N Level 2 APM – Autonomic Platform Management Layer
T Phisical Layer BFE – Bio Front-end
WBAN – Wireless Body Area Network
GUI – Graphical User Interface

BFE
Device
Level 1

Figure 3: System components of a self-managing framework for health monitoring

The Bio-Front End probe. The microcontroller controls the overall

The BFE device, which is worn by the patient, consists of functionality of the device and performs the functions
the sensor, the front-end hardware and firmware for such as data acquisition, device control, data formatting,
processing the sensor signals, a low-power packet forming, wireless transmission and execution of the
microcontroller and a low-power wireless transceiver commands sent by the AGG device.
device for communication. The communication
The Aggregator
technology between the BFE and the AGG device can be
The AGG device is also worn by the patient and it serves
based on a low-power wireless standard like the 802.15.4
to aggregate the data from multiple BFE devices
(Zigbee) or low-power Bluetooth in order to maximize the
connected to a patient and sends these data wirelessly to a
battery life. The type of sensor and the front-end hardware
PC for further processing. The AGG uses low-power
is specific to the physiological signal being measured. For
wireless technology like Zigbee to communicate with the
example, if the BFE device is designed to monitor ECG,
BFE devices and Wi-Fi or other appropriate
the sensors would be pre-gelled electrodes, whereas for
communication technology to communicate with the PC,
pulse oximetry, the sensor would be an optical finger

A Self-Managing Framework for Health Monitoring 316

Intel Technology Journal, Volume 10, Issue 4, 2006

since the required data rate and range is larger than that SELF-MANAGING THE HEALTH
for a body-area network. The AGG device also serves as MONITORING FRAMEWORK
the intermediate messaging link between the BFE device
and the PC. The flow of events in self-managing a device in the
framework is shown in Figure 4. Self-management
The Personal Computer or Server involves sensing or monitoring parameters from the
The PC is used to collect data from multiple AGG devices device or the system to be managed and then analyzing the
connected to multiple patients and process these data. The parameters to find out whether a corrective action is
PC also serves to display the physiological waveforms and required. Intelligent algorithms and decision trees are used
parameters of several patients on the screen, does further to decide if a management action is required and then to
signal processing and computation, performs feature arrive at the most effective decision to act on the inputs.
extraction, stores data, and communicates with other The decision is then executed by the execution engine in a
terminals over the Internet protocol. The PC also keeps feedback loop to effectively manage the device or the
the doctor updated on the patient status by sending the system without the need for user intervention.
patient information on his mobile phone or PDA. It also
manages resources such as printers, scanners, and other
devices, which are shared across multiple patients.

INTELLIGENCE - PLAN /
DEVICE OR SENSE / CALCULATE /
DECIDE / CHOOSE
SYSTEM MONITOR THE ANALYZE
APPROPRIATE ALTERNATIVES
SYSTEM

ACTION /
EXECUTE

Figure 4: Flow of events in self managing a system

The proposed health-monitoring framework can be network. Hence, the PC is better equipped to make
efficiently self-managed by building limited self-managing optimum decisions by considering the status of the
capabilities in each building block and shifting major entire platform as a whole, rather than the status of a
intelligence and decision-making capabilities to the PC. single device.
The PC makes the control decisions based on status inputs The BFE and the AGG devices should have limited self-
from the connected BFE and AGG devices and then managing capability in order to manage themselves in
instructs the devices to perform certain self-management case the communication link to the PC fails. Such devices
functions. can have fail-safe mechanisms and recovery algorithms to
put the device in a safe state in the absence of PC
Following are the advantages of this self-management
connectivity.
approach:
• BFE devices may not have high computational
capability. Hence, they can send their status
information to the PC and use the processing power
of the PC to make optimal self-management
decisions.
• The PC is the central hub of information gathering
and it is aware of the status of all devices in the

A Self-Managing Framework for Health Monitoring 317

Intel Technology Journal, Volume 10, Issue 4, 2006

INTERNET

PAN of PATIENT- 1 PC/Server Remote

BFE-1
PC / Server

AGG

BFE-k

Multiple patients
PDA /
BFE - Bio Front End Mobile
in a Ward AGG - Aggregator
PAN - Personal Area Network Phone
PAN of PATIENT- n
Device Self-Management
BFE-1
Physiological and Status Data
AGG Management commands from Aggregator
Management commands from PC / Server
BFE-k
Commands from remote Hand Held
Device

Figure 5: Self managing framework for health monitoring

Figure 5 shows a self-managing framework for health
Self-Managing the BFE Device
monitoring. Each BFE device connected to the patient
self-manages itself to some extent and can also be The self-managing capabilities of the BFE device may
managed by its WBAN AGG device and by the back-end vary according to the type of the device. An ECG BFE
PC/server. The management scope of a BFE device is device may self-manage itself differently than a SpO2
limited to itself. The management scope of an AGG BFE device. Typically, a BFE device can exhibit the
device is limited to itself and also to manage all the BFE following self-management capabilities:
devices in the WBAN of that particular patient. The PC is • Device to patient binding. This involves binding all
the central hub of information gathering and does the the BFE devices and the AGG device connected to a
system-level management, making decisions based on patient to a unique patient-ID and a unique Personal
inputs from all the devices in the system. The management Area Network (PAN) identification (ID) number. The
scope of PC extends to itself, all AGG devices in its PAN-ID of each WBAN is closely coupled to the
network and all the associated BFE devices. The PC also Patient-ID (i.e., the patient registration number
interacts with other servers that are remotely connected assigned to the patient by the hospital). The unique
via the Internet link. The remotely connected hand-held Patient-ID and the patient history may be
devices and servers can be used to manually manage the programmed in a small Radio Frequency ID (RFID)
BFE and AGG devices connected to the network after wrist strap attached to the patient, when the patient is
secure authentication. The PC manages and communicates admitted to the hospital. The BFE devices may have
with the BFE devices in the network via the intermediate RFID reading capability to read the Patient-ID and be
AGG devices. associated with the unique Patient-ID, when they are
The self-managing capabilities of different blocks of the brought into close contact with the patient’s RFID
framework are described below. wrist strap. Once each BFE device and AGG device
gets associated with the Patient-ID, all the connected
devices can form a unique PAN-ID for the patient.

A Self-Managing Framework for Health Monitoring 318

Intel Technology Journal, Volume 10, Issue 4, 2006

Alternatively, each BFE device may have the • Power management. The BFE device can
capability to autonomically identify a patient based dynamically turn OFF the idle peripherals when not in
on some biometric technique. Data packets sent by use. For example, the BFE device microcontroller can
the BFE devices to the AGG device and to the PC are turn OFF its communication ports, data acquisition,
tagged with the unique Patient-ID in order to bind and control peripherals when not in use in a particular
them to the patient’s individual database. mode of operation. Battery capacity monitoring can
be done at periodic intervals by the BFE device and
• Network association. All the BFE devices and the
the battery capacity information can be sent to the PC
AGG device connected to a patient are programmed
via the AGG device.
to have a unique PAN-ID for their network. As
mentioned above, the PAN-ID can be programmed The BFE device can reduce the clock frequency of its
into the BFE devices and AGG device before microcontroller and related peripherals if too much
connecting them to a patient. When the BFE device is computing is not required by the BFE device sensor.
turned ON, it looks for beacons sent by the AGG Very low data rate BFE devices such as body
device to check the AGG device’s PAN-ID and gets temperature sensors can enter sleep state between
itself associated with the network if it matches with its periodic measurements, whereas, the medium data
own PAN-ID. The AGG in turn identifies the newly rate devices such as SpO2 may turn OFF the wireless
joined BFE device, determines the type (i.e., ECG, transmitter’s RF circuitry between data transmissions
SpO2 etc.) of the BFE device and intimates the same in order to conserve battery power.
to the PC. The PC dynamically adjusts the Graphical
The BFE device may either execute certain power
User Interface (GUI) on its screen to effectively
management functions autonomically or execute them
optimize the screen space to display the signals from
when instructed by the PC. For example, when the
the newly networked BFE device.
battery voltage falls below a certain threshold value,
• Self-configuration. When a BFE device is turned ON the PC may instruct an ECG BFE device to reduce
it automatically configures itself to a default state. It the ECG sampling rate, and the PC may interpolate
also checks the attached sensors and re-configures the sub-sampled ECG data for proper visual
itself for its mode of operation. For example, for an representation. Alternatively, depending on the status
ECG BFE device, there are two modes of operation: of other sensors attached to the patient, the PC may
diagnostic and monitoring. In diagnostic mode a 10 instruct the ECG BFE device to reduce the number of
lead cable is used to sense the ECG signals and the ECG signals to be sampled and reduce the amount of
ECG signals are digitized at 500 samples per second wireless data transmission to conserve battery power.
per lead. However, in the monitoring mode, a 3 or 5
Near the end of battery capacity, the PC and the BFE
lead ECG cable is used to sense the ECG signals and
device may issue audio and visual alerts to inform the
the ECG signals are digitized at a sampling rate of
hospital staff to change the battery. The BFE device
200 samples per second per lead. At start-up, the BFE
also intimates to the PC its decision to turn OFF to a
device can identify the type of ECG cable connected
safe state when the battery capacity reaches 0%.
at its input and automatically adjust its mode of
operation, the number of channels to be sampled, the • Fail-safe mechanisms. It is very important for the
sampling rate per channel, and it can reserve the BFE device to have intelligent algorithms to handle
bandwidth required for the wireless data transfer. alerts in case of malfunctions and put the BFE device
in a safe state to avoid patient injury.
• Self-calibration and self-check. At power ON or
when requested by the PC, the BFE device can also Some important malfunctions could be
execute self-calibration and self-check routines to microcontroller/hardware malfunction, sensor
dynamically adjust its settings for maximum malfunction and communication failure.
accuracy. For example, an ECG device may input a
• Hardware malfunction. These types of malfunctions
fixed reference voltage to its Analog to Digital
can be induced by failure of electronic components or
Converter (ADC), convert it to a digital
can be due to effects of electromagnetic interference
representation, and then adjust its gain and offset
(EMI). EMI can cause the BFE device
errors to compensate for the errors in the ADC.
microcontroller to enter a runaway state that can be
Similarly, a SpO2 device may calibrate its optical
very dangerous. This can be mitigated by ensuring
sensor characteristics, whereas, a NiBP BFE device
that, in the case of malfunction, a watchdog timer
may calibrate its air pressure sensor and self-check its
running inside the microcontroller device either resets
air release solenoid valves and the air pump for
proper operation.

A Self-Managing Framework for Health Monitoring 319

Intel Technology Journal, Volume 10, Issue 4, 2006

the microcontroller or puts the microcontroller and minutes using a circular memory buffer. In case of
associated hardware in a known safe state. communication failure, the BFE device keeps on
looking for the periodic beacons sent by the AGG
In the case of BFE devices such as a NiBP device, the
device in order to check the PAN-ID and re-establish
blood pressure is measured by occluding the brachial
the network connection.
artery by inflating a cuff tied to the patient’s arm. If
the air inside the cuff is not released due to hardware Communication errors may be introduced in the
malfunction, the blood flow to the patient’s hand communication link between the BFE device, the
stops and may cause serious tissue damage. Hence, it AGG device and the PC due to the presence of other
is very important to have either a redundant networks operating on the same wireless frequency. A
microcontroller or some standby mechanical devices typical example can be the interference of Wi-Fi
(such as normally open solenoid air release valves) to (802.11b) in a Zigbee (802.15.4) network since they
ensure air release even in case of power or hardware both operate in the same frequency band of 2.4 GHz.
failure. In another scenario, two patients may be connected
with BFE devices that use the same frequency
• Sensor malfunction. The BFE device also needs to
channel for communication to their respective AGG
handle situations when the sensors connected to the
device. When such patients come in close proximity,
BFE device malfunction or when they get
their networks conflict with each other causing errors
disconnected from a patient. For example, for an
in communication. The BFE-AGG body area network
ECG BFE device, the number of electrodes connected
needs to have a robust interference detection and
to a patient may vary from three to ten. The BFE
mitigation algorithm by which a conflicting BAN
device needs to continuously monitor the impedance
would dynamically sense and change its operating
of each electrode and generate an alert in case the
frequency channel in case of interference from a
electrode gets disconnected from the patient. The
similar network.
BFE device sends the sensor status to a PC
periodically, and the PC can generate an alert in case In case the BAN is not able to mitigate the
of malfunction. Apart from generating an alert it may interference, robust error detection techniques should
also be necessary to stop displaying the waveform of prevent the BFE device from executing wrong
the disconnected ECG lead to avoid the doctor commands received from either the AGG device or
misinterpreting the noise picked up by the the PC.
disconnected electrode as a valid ECG signal. The
ECG BFE device may, however, continue to monitor Self-Managing the Aggregator Device
the ECG signals from other electrodes. The AGG device serves as a communication link between
In another case, an air leak may be present in the cuff the WBAN formed by the BFE devices and the backend
tied to the patient’s arm for blood pressure PC/server. The AGG device has a low-power wireless
measurement. The NiBP BFE device needs to interface like 802.15.4 or a low-power Bluetooth to
monitor the rate of increase in cuff pressure once the network the BFE devices and also a long range, high data
air pump starts pumping air into the cuff. If the cuff rate interface like Wi-Fi to interface with the back-end
pressure does not increase at the required rate an air- PC. The AGG device may or may not have a local LCD
leak alert is transmitted to the PC. display and control keypad. A typical example for an
AGG device could be a PDA. The self-managing
• Communication failure and communication errors. capabilities of the AGG device may be similar to that of
The BFE device also needs to handle situations when the BFE devices.
the sensors and its hardware components are
functioning normally but when the communication The AGG device manages the BAN and allows BFE
link to the AGG device or the PC fails. In such cases devices to join or leave the BAN seamlessly. Automatic
the BFE device is neither able to send the BFE device discovery, device type identification, and
physiological data and its status to the PC, nor is it BAN networking are the responsibility of the AGG
able to receive autonomic control messages from the device. The AGG should have sufficient capabilities to
PC. In such a scenario, the BFE device may save the manage the BAN efficiently even in the absence of the
patient’s data in its local memory for transmission to communication link to the PC and have sufficient memory
the PC when the communication link is re- for storage of patients’ physiological and status data over
established. If the BFE device does not have extended periods of operation.
sufficient memory to store data for long periods of Another important function of the AGG device is data
disconnection, it may save the status of the last few encryption/decryption for network security.

A Self-Managing Framework for Health Monitoring 320

Intel Technology Journal, Volume 10, Issue 4, 2006

Autonomic Platform Management (APM) by The PC is better equipped to handle self-management

the PC functionality of the health-monitoring platform as a
whole. The PC may run complex bio-signal
The PC or server is the central hub of data collection from processing algorithms and based on the results may
multiple WBANs of multiple patients. The PC aggregates instruct the BFE device to execute certain self-
data from multiple patients, processes the data, stores management tasks. For example, the PC may analyze
them in specific formats and also makes the data available the ECG signal for the presence of baseline drift and
to multiple remote terminals for display. The PC also may send a message to the ECG BFE device to
makes the patients’ data available to the doctor at any restore the ECG baseline to ground by re-charging an
place and at any time by transmitting important patient AC coupling capacitor in the BFE device hardware.
data to hand-held devices like mobile phones and PDAs. Similarly, if very small amplitude ECG signals are
The PC also has the capability to issue control commands being sensed by the electrodes, it can instruct the BFE
to any BFE device in its network to change its operating device to increase the amplification of the BFE
parameters. The PC also manages the shared resources device’s programmable gain amplifier hardware.
like printers, fax equipment, etc., which are connected to
it. For a NiBP BFE device the PC may instruct the
device to take blood pressure measurements either at
Since the PC is the central hub of information aggregation programmed intervals or autonomically when certain
from BANs, user inputs from GUIs and information from thresholds are exceeded or under manual control.
remote terminals, it is better equipped to make self-
management decisions for the entire health-monitoring The PC can also perform autonomic power
framework. Also, since the PC has tremendous management of the devices connected on its network.
computation capacity it is better suited to run complex The PC can turn ON or turn OFF the BFE devices or
decision trees to arrive at the best management decision. keep them in SLEEP state. The PC can wake the
sleeping devices at appropriate times, instruct them to
A single PC can be used to cater to the processing needs take a measurement and put them to SLEEP again.
of multiple patients in a ward of the hospital. Multiple Also, the PC can extend the battery life of a device by
wards of the hospital can be connected to a server to dynamically changing the device parameters and
further aggregate data from the entire hospital and to settings. For example, the PC can instruct an ECG
send/receive messages to hand-held devices. BFE device to reduce the number of leads to be
The self-managing capabilities of the PC/server, shown by sampled or reduce the sampling rate per lead,
the APM layer in Figure 3, can be as follows. depending on the quality of ECG required.

• Network management. The PC forms a network of the • Autonomic GUIs. The PC can autonomically adjust
AGG devices and dynamically changes the network the screen space to display a number of vital signal
topology as patients get transported in and out of the waveforms and numerical data from a number of
wards for tests and operative procedures. The PC is patients. The screen can be automatically partitioned
also aware of the devices in the BAN of each AGG and re-adjusted to accommodate the data from new
device. The PC enables automatic device discovery, devices that join the network, and the screen space
device association and disassociation algorithms, and can be made free and utilized in a better manner when
automatic device-to-patient binding. Each patient is the BFE devices leave the network. When the screen
associated with a unique PAN-ID, and all devices in space is limited, and displaying all parameters from
the BAN of the same patient have the same PAN-ID. all patients is not possible, the PC should give weight
The PC binds the real-time data of each patient to the to critical parameters such as ECG readings.
patient’s respective backend database. • Managing alerts. The PC is the central computing
Apart from managing the networking of AGG and resource for the entire health-monitoring framework
BFE devices, the PC also manages the shared and needs to process alerts coming from almost all
resources on its network like printers, scanners, fax devices attached to the network.
machines, etc. Broadly, the alerts can be classified as alerts
• BFE device control. Since critical patients have generated by the patient’s physiological status, device
different monitoring requirements as compared to status, and communication errors.
non-critical patients in recovery, the PC can re- • Alerts due to patient’s physiological status. The PC
configure the BFE devices in different operating processes the physiological signals and parameter
modes and change the device settings depending on data from the patient and checks if any of the
the level of monitoring required.

A Self-Managing Framework for Health Monitoring 321

Intel Technology Journal, Volume 10, Issue 4, 2006

programmed limits are exceeded. For example, the and configurable logic to heal itself in case of
PC computes the heart rate by identifying and component failure. The storage media and the
counting the number of ‘R’ waves in the patient’s Internet broadband link should be robust. The
ECG per minute. In case the heart rate crosses the aggregated patient data and the network status
programmed limits, the PC may request the NiBP information should also be copied in a central archive
BFE device to take frequent blood pressure so that in case of malfunction, a standby PC should be
measurements and may also power ON a defibrillator able to take over the monitoring functionality of the
machine in case the patient needs to be administered failed PC without loss of data.
an electrical shock to restore normal heart rhythm.
The PC can also intelligently identify irregular heart USAGE MODELS
rhythms called arrhythmias and keep a log of these
This self-managed digital health-monitoring framework
with a time stamp. In some critical conditions, the PC
can be applied in several use cases, some of which are
can automatically send the abnormal physiological
described below:
data to a doctor’s mobile phone/PDA for diagnosis
and also trigger audio-visual alarms locally.
Use Case 1: Remotely Monitoring Patients at
• Device status alerts. The PC monitors the status of Home
the BFE and AGG devices and manages the alerts Home monitoring involves monitoring of non-critical
generated due to low battery, sensor failure or patients, patients in the recovery stage after being
disconnection, component failure, calibration errors, discharged from hospital, proactive health monitoring, or
etc. For example, if the patient condition is not monitoring the health of the elderly within their homes or
critical, under low battery conditions the PC may care facilities.
instruct an SpO2 BFE device to intermittently
monitor the blood oxygen saturation at 1-minute Joe, 68 years old, has recovered from open heart surgery
intervals, rather than monitoring it continuously. This and has been discharged from the hospital. Lately, Joe has
allows the LEDs inside the SpO2 finger probe and the become forgetful and forgets even basic tasks like taking
associated circuitry to sleep between measurements, his medicines on time. Joe’s children are working and are
thereby extending the battery life. For sensor concerned about Joe’s health when they are at work. Dr.
disconnection the PC can raise appropriate audio- Smith wants to remotely monitor Joe’s recovery as Joe
visual alarms and/or page the hospital staff on duty. performs his daily activities to ensure that his recovering
cardiovascular system is not subject to a sudden stress. Joe
In the case of component failure or calibration errors is fitted with a wireless ECG device, a NiBP device, and a
the PC can compensate for the error by software Pulse Oximetry device for monitoring his blood oxygen
correction or instruct a device like the NiBP to saturation. The devices connected to Joe are miniature,
measure blood pressure using the redundant pressure lightweight, and they do not interfere with his daily
sensor. Once an error is detected and flagged by the activities.
PC, the PC also monitors whether the error has been
corrected and removes the error flag accordingly. Joe has a Wi-Fi-enabled desktop PC in his living room
Alternatively, the user may manually acknowledge which is also used by Joe to watch TV. The PC
the error and put the PC in a mute state for a certain continuously monitors Joe’s vital parameters in the
amount of time within which the user is expected to background and sends data to the hospital in real-time.
correct the error. Wireless webcams fitted in different rooms of his home
also send streaming video data to the PC. Joe’s PC is
• Fail-safe mechanisms. Since the PC is the central connected to the Internet using a broadband link.
computing resource for the health-monitoring
platform, it is necessary to build reliability into the While sitting at his laptop in the hospital, Dr. Smith
PC platform. Reliable hardware design, a reliable connects to Joe’s PC and runs the pre-designed protocol
low-latency hard real-time OS, a reliable networking for remotely monitoring open heart surgery patients like
stack and application software are necessary to build Joe. Dr. Smith runs the protocol and then leaves for an
a medical-grade PC. The PC should be immune to the urgent operation. The protocol remotely and
EMI generated by other medical equipment in its autonomically programs Joe’s ECG device to
vicinity and should also have low EMI emissions. The continuously monitor a 3-lead ECG, the NiBP device to
PC should have a battery backup to remain operative take a blood pressure reading once every hour, and the
in case the mains power supply fails. The PC should Oximetry device to monitor only in critical situations. The
have the capability of self-monitoring, preventive protocol automatically sends a reminder audio-visual
maintenance, and have redundant processing cores message on Joe’s PC screen asking him to undergo the

A Self-Managing Framework for Health Monitoring 322

Intel Technology Journal, Volume 10, Issue 4, 2006

morning exercise on his home treadmill and also programs FHR, and the software automatically sends alert messages
the treadmill to limit the maximum speed to 2 miles per and FHR tracings to Dr. Willy and the chief nurse on their
hour and limit the exercise time to 10 minutes. The mobile phones. Dr. Willy is out of hospital when he
reminder message alerts Joe in the midst of his TV receives the message on his mobile phone. Dr. Willy
program. Joe starts walking on the treadmill and during immediately rings up the chief nurse and instructs her to
his exercise session, his heart rate increases to 130 beats be prepared for an emergency cesarean operation by the
per minute (bpm). The increased heart rate is monitored time he reaches the hospital. Judy undergoes an
by the PC, and the PC triggers the NiBP device to take emergency cesarean operation and delivers a healthy
frequent blood pressure measurements and it also baby. During delivery Dr. Willy notices the umbilical cord
automatically stops the treadmill as a precautionary entangled around the fetus’s neck resulting in partial
measure to prevent excessive stress. The increased heart suffocation and thereby decreasing the FHR. Dr. Willy is
rate condition also turns on the Pulse Oximetry device to thankful that the timely alerts by the central monitoring
monitor Joe’s oxygen saturation. When Joe’s heart rate system saved the baby’s life.
returns to normal after his exercise session, the NiBP
device again increases the BP measurement interval to 1 Use Case 3: Critical Patient Monitoring in an
hour while the Oximetry device turns OFF. Since Dr. Intensive Care Unit
Smith had set the mobile phone alert limit for heart rate to
Dr. Bill is a resident doctor currently managing the 16-bed
150 bpm, BP limits to 150/100 mmHg, and the SpO2 limit
ICU in a hospital. A patient, Jack, operated on via
to 94%, he was not alerted on his mobile phone since
angioplasty, is brought to Dr. Bill’s ICU. Dr. Bill tags
Joe’s vital signs were within the limits. When Dr. Smith
Jack’s admission time in ICU by using a RFID reader to
returns from his operation he examines the stored vital
read Jack’s RFID wrist strap and also wirelessly transfers
parameters of Joe and he is happy with Joe’s increased
a softcopy of Jack’s medical record file from inside the
stress-handling capability. Meanwhile, Joe’s son is also
wrist strap to the central station for study. Dr. Bill rubs the
able to remotely keep an eye on Joe from his office by
wireless patient monitoring devices on Jack’s RFID wrist
monitoring the streaming webcams and Joe’s vital
strap and connects them to his body. The wireless
parameters. Dr. Smith has remotely also fed the medicine
monitoring devices read Jack’s Patient-ID from his wrist-
schedule on Joe’s PC. Joe’s PC flashes audio-visual
strap to form a BAN and get associated to Jack’s backend
messages on its screen in a timely manner to remind Joe to
database on the central station. The computer screen of
take his medicines.
the ICU’s central monitoring station automatically re-
adjusts itself to display, monitor, and store Jack’s vital
Use Case 2: Wireless Fetal Monitoring signs along with the parameters of other patients. Dr. Bill
During Labor prepares a monitoring schedule and a drug administration
Judy is pregnant and has been admitted to the hospital for schedule for Jack using the central monitoring application
delivery. Dr. Willy expects a normal delivery but decides software. The monitoring and drug administration profile
to monitor Judy’s Uterine Contractions (UC) and Fetal for Jack is immediately transmitted to the hand-held
Heart Rate (FHR). The fetal heat rate variability is an devices given to nurses on duty in the ICU. The central
important parameter to asses fetal well being and to assess station performs periodic diagnostic tests using the
whether the fetus would be able to sustain the stress of attached wireless devices and also sends timely reminders
uterine contractions during delivery. Dr. Willy fits Judy to the nurses to administer drugs to patients, depending on
with a wireless ultrasound sensor and a uterine pressure individual patient profiles. Once a nurse administers a
sensor by means of a belt around her abdomen. The drug to a patient she mutes the generated reminder and
ultrasound sensor monitors the FHR while the pressure logs the event, which also gets logged in the central
sensor monitors Judy’s UCs and sends the data wirelessly station. The central monitor also runs intelligent
to the central server of the labor ward, where a number of algorithms on patient’s vital physiological signals and
patients are being monitored simultaneously. Judy is still generates alerts, sends alerts to doctors’ mobile phones,
not in labor and is able to take frequent walks around her and prints important events on locally connected printers.
room and in the ward, while still being closely monitored. Dr. Bill is relieved that the central monitoring station
Judy does not feel the subtle UCs, indicating the start of manages most of his mundane tasks and helps him in
labor. However, minute uterine pressure changes are effectively managing large numbers of patients.
picked up by the pressure sensor and are evident in the
graphical tracing at the central monitoring terminal. There CHALLENGES AND OPPORTUNITIES
is a sudden and steep drop in the FHR which goes
There are several challenges which need to be overcome
unnoticed by the busy hospital staff. However, the FHR
in order to realize such a digital health-monitoring
analysis software on the server detects the sudden drop in
framework:

A Self-Managing Framework for Health Monitoring 323

Intel Technology Journal, Volume 10, Issue 4, 2006

Platform reliability. Ensuring the reliability of the entire effectively. The number of errors related to paper-based
platform is the key to the success of the proposed processes is reduced significantly in an autonomically
architecture. Reliability has to be built right from the managed framework. The proposed framework leads to
electronic component level to the OS and application better resource utilization, better resource sharing, reduces
software level. It is necessary to have fail-safe and backup doctor’s intervention, and hence makes healthcare more
mechanisms to ensure that patient monitoring is not affordable.
interrupted when parts of the network fail.
Such an open standards-based health-monitoring platform
Robust wireless communication. Wireless communication would motivate more standards-based hardware and
is the backbone of the proposed health-monitoring software designs and shift the biomedical OEMs from
framework. Robust mechanisms should be developed to proprietary hardware-centric platforms to standards-based
mitigate interference issues that result when several software-centric general-purpose PC platforms.
wireless networks co-exist. Robust error correction and
error detection algorithms should be developed to build ACKNOWLEDGMENTS
the same reliability as that of a wired link in the wireless
The authors recognize the valuable suggestions given by
interfaces.
the reviewers for improving the quality and content of this
Standardization and interoperability. The medical device paper.
OEMs need to agree on a common set of communication
protocols and standard hardware interfaces. Devices from REFERENCES
different manufacturers should be able to plug into the
[1] David Culler, Deborah Estrin, Mani Shrivastava,
system seamlessly.
“Overview of Sensor Networks,” IEEE Computer
Infrastructure and ubiquity. The health-monitoring Society, August 2004, pp. 41–49.
framework needs wireless communication infrastructure
[2] J. A. Stankovic, Q. Cao, “Wireless Sensor Networks
like Wi-Fi hot spots, routers, switches, etc. which are
for In-Home Healthcare: Potential and Challenges,”
limited to a hospital or home network. Emerging
Department of Computer Science, University of
technologies such as WiMAX can address the problem of
Virginia.
seamless wireless connectivity throughout the cities and
villages. [3] Chee-Yee Chong, Srikanta P. Kumar, “Sensor
Networks: Evolution, Opportunities and Challenges,”
Intelligent algorithms. The PC needs to run intelligent
in Proceedings of the IEEE, Vol. 91, No. 8, August
algorithms to make self-managing decisions. The
2003, pp. 1247–1256.
algorithms should be continuously evolving and patient
centric. Machine learning, artificial intelligence, and
prediction algorithms may require tweaking and clinical AUTHORS’ BIOGRAPHIES
trials until they reliably self-manage a health-monitoring Amit Baxi is a biomedical engineer working as an R&D
framework. Engineer in the Corporate Technology Group, Intel
Corporation. He has been in Intel for more than a year and
Development of miniature, ultra-low power sensors and
is responsible for architecting and building the hardware
battery technology. BFE device sensors and hardware
and software bio-medical components for Intel’s
need further miniaturization to the level of a small system-
healthcare platforms. He has more than 11 years
on-chip, and the power needs to be optimized to the
experience in the design and development of medical
microwatt level so that the sensors can operate by using
embedded systems such as Cardiac Stress Test Systems,
ambient light as the power source. Battery technology
Defibrillators, Multiparameter Patient Monitoring
such as the lithium-polymer and moldable lithylene
Systems, Spirometers, ECG monitoring and diagnostic
battery technology have to evolve further to combine high
equipment etc. His e-mail is amit.s.baxi at intel.com.
capacity, small form factor, and light weight.
Nagaraju Kodalapura is an electronics and
CONCLUSION communication engineer working as an R&D Engineer in
the Health Platforms lab of the Corporate Technology
We believe that a self-managed wireless health-
Group, Intel Corporation. He has been with Intel for about
monitoring framework can significantly improve the
six years. He is primarily responsible for the design and
quality of healthcare while providing patient comfort,
development of Wireless Embedded System Software for
mobility, and continuity of care. Such a framework does
bio medical sensors for Intel’s healthcare platforms. His
not require the doctor to be in close proximity to the
expertise is in the area of wireless embedded systems,
patient; however, it still provides the same quality of care.
device drivers, debuggers, and simulators for multicore
The doctor is enabled to monitor more patients

A Self-Managing Framework for Health Monitoring 324

Intel Technology Journal, Volume 10, Issue 4, 2006

processors. His e-mail is nagaraju.n.kodalapura at

intel.com.
Copyright © Intel Corporation 2006. All rights reserved.
Intel is a registered trademark of Intel Corporation or its
subsidiaries in the United States and other countries.
* Other names and brands may be claimed as the property
of others.
Bluetooth is a trademark owned by its proprietor and used
by Intel Corporation under license.
INFORMATION IN THIS DOCUMENT IS PROVIDED
IN CONNECTION WITH INTEL® PRODUCTS. NO
LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL
OR OTHERWISE, TO ANY INTELLECTUAL
PROPERTY RIGHTS IS GRANTED BY THIS
DOCUMENT. EXCEPT AS PROVIDED IN INTEL'S
TERMS AND CONDITIONS OF SALE FOR SUCH
PRODUCTS, INTEL ASSUMES NO LIABILITY
WHATSOEVER, AND INTEL DISCLAIMS ANY
EXPRESS OR IMPLIED WARRANTY, RELATING TO
SALE AND/OR USE OF INTEL PRODUCTS
INCLUDING LIABILITY OR WARRANTIES
RELATING TO FITNESS FOR A PARTICULAR
PURPOSE, MERCHANTABILITY, OR
INFRINGEMENT OF ANY PATENT, COPYRIGHT OR
OTHER INTELLECTUAL PROPERTY RIGHT.
Intel may make changes to specifications and product
descriptions at any time, without notice.
This publication was downloaded from
http://developer.intel.com/.
Legal notices at
http://www.intel.com/sites/corporate/tradmarx.htm.

A Self-Managing Framework for Health Monitoring 325

Intel Technology Journal, Volume 10, Issue 4, 2006

THIS PAGE INTENTIONALLY LEFT BLANK

A Self-Managing Framework for Health Monitoring 326

For further information visit:
developer.intel.com/technology/itj/index.htm

Copyright © 2006 Intel Corporation. All rights reserved.

Intel is a trademark or registered trademark of Intel Corporation or its subsidiaries in the United States and other countries.
For a complete listing of trademark information visit: www.intel.com/sites/corporate/tradmarx.htm