IntelliSpace ECG

Serial Comparison Algorithm Reference

 IntelliSpace ECG
SERIAL COMPARISON ALGORITHM REFERENCE

Revision B
December 2015
Notice Authorized EU-representative:
Philips Medizin Systeme
About This Edition Böblingen GmbH
PN 453564373101 Hewlett Packard Str. 2
Revision B 71034 Böblingen
Printed in USA German

Edition History
Revision A, June 2012
Revision B, December 2015

Applicable to IntelliSpace ECG

B.01and later.
Philips Medical Systems shall not
be liable for errors contained herein
or for incidental or consequential
damages in connection with the
furnishing, performance, or use of
this material.

Copyright
Copyright © 2015
Koninklijke Philips Electronics
N.V. All rights are reserved.
Andover, MA 01810-1099 USA
(978)687-1501

Warranty
Philips Medical Systems makes no
warranty of any kind with regard to
this material, included, but not
limited to, the implied warranties or
merchantability and fitness for a
particular purpose.

CAUTION
In the U.S. Federal Law restricts
this product to sale on or by the
order of a physician. Use of
accessories other than those
recommended by Philips Medical
Systems may compromise product
performance.
THIS PRODUCT IS NOT
INTENDED FOR HOME USE.

Medical Device Directive

This algorithm is a software
component used in IntelliSpace
ECG. It is not considered to be a
medical device.
 Contents

What is New? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
How the Serial Comparison Algorithm Analyzes Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Monitoring Infarct Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Overview of Serial Comparison Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Ventricular conduction delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Right Ventricular Infarct/Injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Requirements for Performing Serial Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Serial Comparison Statement Display Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Comparing ECGs Using the Serial Comparison Algorithm. . . . . . . . . . . . . . . . . . . . . . . . 18
Undoing a Serial Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Philips Serial Comparison Algorithm Reference iii

 Table of Contents

iv Philips Serial Comparison Algorithm Reference

 1 About the Philips Serial Comparison
Algorithm
NOTE Serial comparison of ECGs is not supported in IntelliSpace ECG Basic Edition, unless you have
purchased the Serial Comparison option. Contact your Philips representative for details.
Serial comparison is a method of comparing two ECGs to detect significant clinical changes
between the two. It provides more information regarding progression or regression of an
abnormal condition than reviewing a single ECG, as the algorithm detects significant changes
in rhythm and morphology relative to a previous ECG. The resulting analysis allows you to
focus on pathological ECG changes. This is especially important when trying to analyze serial
changes after an acute myocardial infarction (MI). Infarction evolution can only be described
when a series of ECGs are serially compared.
The Philips serial comparison algorithm uses a combination of measurements and statements
generated by the Philips DXL12/16 Lead Algorithm as the basis for comparison of the current
ECG to a suitable previous ECG that meets all of the specified requirements (some of which
are configurable and site specific).
The IntelliSpace ECG system allows you to serially compare ECGs manually via the
IntelliSpace client (using the Compare button) or automatically (for example, upon entry
into the database) via workflow rules configured at your site by the administrator.
NOTE Serial comparison of ECGs is not supported in the IECG Anywhere editor.
If serial comparison is invoked through workflow and there is a previous ECG that is a
suitable candidate for serial comparison, then the current ECG is serially compared to the
previous ECG. However, if serial comparison is workflow invoked and the previous ECG is
not a suitable candidate or there is no previous ECG, then serial comparison will indicate
there is no suitable previous ECG.
If serial comparison is manually invoked (via the IntelliSpace client), the current ECG is
serially compared to the previous ECG if the previous ECG is a valid candidate for serial
comparison.
After a manual serial comparison is performed and you save the changes, the current ECG is
saved to the database with the serial comparison statements and quality statements that have
been added and/or statements that have been deleted. Serial comparison markings are
displayed in every view in the IntelliSpaceECG client (unless you elect to hide statements, or
serial comparison modifiers and/or symbols are hidden at your site). They may also be present
on the printed report, depending on the settings at your site.

1
About the Philips Serial Comparison Algorithm

The following 2 statements apply only to an ECG that was serially compared on a
TraceMasterVue C.03.xx system (or older):
 Serial comparison markings are displayed in every view in the ECGVue application
(unless you elect to hide statements, or serial comparison modifiers and/or symbols are
hidden at your site).
 Serial comparison markings may also be present on the printed report, depending on the
settings at your site.
NOTE The support for serial comparison modifiers and/or symbols was discontinued in IntelliSpace ECG
A.01.
NOTE You can also perform a visual comparison between two ECGs by displaying them simultaneously on
the screen in Waveform view. This method does not involve the serial comparison algorithm. For
details, see the section on using the Waveform view in the client application Instructions for Use.
On a site-wide basis, the administrator can configure some display options for serial
comparison modifiers. See “Serial Comparison Statement Display Options” on page 18.

What is New?
In 2014 Philips introduced a new version of the DXL ECG analysis program(PH 110C) that
improves detection of acute myocardial infarction when ST confounders are present. The
following serial comparison algorithm modifications were made for compatibility with the
new Philips DXL algorithm.


 When the serial comparison algorithm recognizes a significant change in rhythm, the
previous rhythm is called out in the summary.

How the Serial Comparison Algorithm Analyzes Data

The serial comparison algorithm currently in use has evolved from the original algorithm
developed in 1978 as a collaboration by a group of cardiologists at Duke University and
University of Florida1. It has been used in clinical settings at the Heart Stations at hospitals
affiliated with these universities for many years, and has been further refined and enhanced
over time.
The program compares two ECGs at one time, using a combination of measurements and
statements generated by the Philips DXL 12/16 Lead Algorithm as the basis for comparison of
the current ECG to a suitable previous ECG.
If serial comparison is manually invoked (via the IntelliSpace client) by clicking the Compare
button on the Action bar or through a workflow rule, the current ECG is serially compared to
the previous ECG if the previous ECG is a valid candidate for serial comparison.

1. Dr. J. Greenfield, Duke University; Dr. L. Crevasse and Dr. M. Ariet, University of Florida

2 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

For the rhythm categories, the algorithm uses a “statement only” approach, and does not
take measurements into account outside of heart rate. Similarly, in those few cases where
measurements are not available in the previous ECG, only statements are compared for the
diagnostic categories.
Although the comparison is only between two ECGs, the results of comparisons are
cumulative, so long as there are no significant measurement changes that would justify the
removal of a statement and statements are not deleted by an editing cardiologist. This allows
a diagnosis to be retained indefinitely.
The serial comparison algorithm compares interpretive statements and the waveform
measurements for each ECG to determine whether any change in statements is significant.
After performing the serial comparison between two ECGs, the comparison program adds
and/or deletes statements, as appropriate. The complete set of statements is saved with the
ECG.
NOTE You can undo the comparison at a later time, for example, if the comparison ECG was invalid, or a
more representative ECG was acquired that you want to compare to. For details, see “Undoing a
Serial Comparison” on page 18.
The comparison program is designed to compare abnormal conditions by category, which
correspond to the ECG diagnostic categories such as Basic Rhythm, Left Ventricular
Hypertrophy (LVH), Inferior Infarct , and so on. The program only looks for changes of
condition in those categories that correspond to the statements found in the interpretation
of the ECGs being compared.
The rhythm and morphology categories are analyzed as follows:
 For rhythm categories, in addition to rhythm statement, heart rates are compared to
determine whether, for the same rhythm statement in the current and previous ECGs,
the heart rate has changed by 20 bpm or more.
NOTE This threshold can be configured by the user. See Configuring IntelliSpace ECG for more
information.
 For the morphology categories, the program uses the concept of significant/insignificant
change. A significant change represents a change in a measurement variable that exceeds
the predefined limit. The change can be positive or negative, indicating a move toward a
higher or lower degree of severity within a category. For key threshold values used, see
Table 1.
Serial comparison is particularly helpful for tracking the evolution of myocardial infarction.
The program automatically includes the date of analysis in any infarct statement that has the
modifier “acute” or “recent” to indicate the date of the infarct. When serial comparison takes
place, if the date of the previous ECG is within 14 days of the current date, the program looks
for significant ST, T, or Q wave changes in the leads relevant to the infarct location. If the
changes are significant, the infarct is marked as Evolving. If changes continue on the next
ECG, the statement reads Evolution Continues.

Monitoring Infarct Evolution

One of the most critical functions of the serial comparison algorithm is to describe the changes
that occur after an acute myocardial infarction (MI). The algorithm labels the date of the ECG

Philips Serial Comparison Algorithm Reference 3

About the Philips Serial Comparison Algorithm

in which the infarct is first identified as acute, and carries this date through successive ECGs.
The age of the infarct impacts how the data are compared.
1 First, there must be an infarct statement in the current ECG in a specific category of
interest. The previous ECG must also have an infarct statement in the same or related
category. That is, anterior MI (AMI), anterior lateral MI (ALI), lateral MI (LMI) or
inferior MI (IMI), posterior MI (PMI), lateral MI (LMI).
2 The evolution logic looks for significant measurement differences (SMD) in the category
of interest. The differences are generally an increase or decrease in the involved leads in
one of the following measurements:
– ST onset of +/-0.1 mV
– T amplitude of +/-0.2 mV
– Q duration of +/-20 ms
3 If the infarct evolution logic indicates there are SMDs and there is less than 15 days
between the current ECG and the date when the infarct was first identified as acute, and
the previous infarct is not evolving, the Evolving Infarct statement (MI-97) is generated.
If the previous infarct is evolving, the Evolution Continues (MI-98) statement is
generated.
If the previous infarct statement is evolving and there are no SMDs, the No Further
Evolution (MI-99) statement is generated.
4 Once the infarct age reaches 15 days, evolution logic is discontinued. If there is an
evolution statement in the previous ECG, the current infarct statement is used. The
exception is if an ST-onset elevation of 0.1 mV remains in the involved leads, then the
ST Elevation Persists (MI-93) statement is generated. After 60 days, this statement is
replaced by the current infarct statement.
5 If at any time there is no current infarct statement in the category or related category,
the infarct evolution statement is removed.
In an acute/evolving phase, it is likely that the ECG tracings of infarct location correspond to
a single occlusion in the coronary arteries. In a series of serial comparisons, the location of the
infarct may appear to change due to minor measurement variations that trigger different
statements due to exceeding different algorithm thresholds. The algorithm minimizes this
variability by grouping some infarct locations to represent the same occlusion. For example,
a statement of acute inferior infarct in an analysis can be followed by a statement of posterior
infarct in the next ECG. When this happens, the algorithm reports the inferior infarct (acute or
evolving) and adds the statement, Consider Posterior Wall Involvement (CPWI).
The algorithm has been tested extensively in clinical settings against cardiologist edits to the
interpretive statements. In general, it has been found that the number of editing changes
cardiologists need to make to the interpretive statements generated by the analysis algorithm
is greatly reduced when ECGs are compared versus when individual ECG analyses are
reviewed.2 This reduction occurs in part because the serial comparison algorithm benefits
from any edits made to the previous ECG and because SMDs need to be present for most
statements to be changed.
2. Gregg RE, Deluca DC, Chien CH, Helfenbein ED, Ariet M. Automated serial ECG comparison improves
computerized interpretation of 12-lead ECG. J Electrocardiol 2012; 45: 561-5.

4 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

Philips Serial Comparison Algorithm Reference 5

About the Philips Serial Comparison Algorithm

Table 1 Serial comparison algorithm threshold values to determine significant

change in key morphology categories

Category Compared Relevant Leads Involved Values Qualify as

Measurement “Significant Change”

Conduction Disturbance

AV Conduction Delay PR interval All leads   ± 30 ms

Axis QRS axis Limb leads   ± 20 degree
Intraventricular QRS duration All leads   ± 20 ms
conduction defect
(IVCD)
Bundle Branch Block QRS duration All leads   ± 20 ms
(BBB)
Wolf-Parkinson- QRS duration All leads   ± 20 ms
White Syndrome
(WPW)
Hypertrophy

Right atrial P amplitude 2 of II III aV F   ± 50 μV or

enlargement (RAE) V1 V2   ± 100 μV
(if HR
Left atrial P or P amplitude V1   ±100 μV
enlargement (LAE)
Right Ventricular R amplitude/ V1   ± 0.5 or
Hypertrophy (RVH) S amplitude ratio, V6   ± 150 μV
S amplitude
Left Ventricular R amplitude aVL   ± 300 μV or
Hypertrophy (LVH) 3 of V1, V2 V5   ± 500 μV
V6
Infarcts

Inferior myocardial Q amplitude, 2 of II III aVF   ± 100 μV or

infarction (IMI) Q duration   ± 20 ms
Posterior myocardial R amplitude/ V1 or V2   ± .5
infarction (PMI) S amplitude ratio
Lateral myocardial Q duration 2 of I aVL V5   ± 15 ms
infarction (LMI) V6
Anterior myocardial Q duration 2 of V1 V2 V3   ± 20 ms
infarction (AMI) V4

6 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

Table 1 Serial comparison algorithm threshold values to determine significant

change in key morphology categories (continued)

Category Compared Relevant Leads Involved Values Qualify as

Measurement “Significant Change”

Infarcts (continued)

Right ventricular ST onset 1 of V1, V4R   ± 50 μV

myocardial infarction
(RMI)
Anterior lateral Q duration 2 of I aVL V5   ± 20 ms
myocardial infarction V6
(ALI)
Apical myocardial Q duration II   ± 20 ms
infarction (APMI)
Lateral lead involved Q amplitude, V6   ± 100 μV or
Q duration   ± 20 ms
ST-T Abnormalities

T wave/Ischemia T amp 2 of II III aVF or   ± 150 μV

2 of V3 V4 V5
V6
ST depression ST mid 2 of II III aVF or   ± 150 μV
2 of V3 V4 V5
V6
ST-T/Repolarization ST mid, T amp 2 of II III aVF or   ± 150 μV
abnormality 2 of V3 V4 V5
V6
ST elevation ST mid 3 of I II aVL aVF   ± 150 μV
V4 V5 V6
Tall T wave T amp 1 of V1 V2 V3   ± 200 μV
V4 V5 V6
Drug/QT QT duration All leads   ± 50 ms
Pericarditis T amplitude Any 3 leads   ± 100 μV

Overview of Serial Comparison Rules

Once the comparison is complete, the program adds the applicable comparison statements, as
well as one of the following summary statements, based on the results of the comparison.
The following tables provides an overview of the high-level rules used in the serial
comparison program.

Philips Serial Comparison Algorithm Reference 7

About the Philips Serial Comparison Algorithm

 Table 2 shows rhythm-related rules

 Table 3 shows contour/morphology-related rules
 Table 4 shows the serial comparison statements
 Table 5 shows summary statements

Table 2 Rhythm-related rules used by the algorithm

Rhythm Comparison
The algorithm compares primary and secondary rhythm between the two ECGs.
Primary rhythm changes. Primary rhythm changes are always labeled as Significant
rhythm changes.
Primary rhythm:
“Sinus”, and related subcategories, including among others, “sinus bradycardia”,
“sinus tachycardia”, “sinus arrhythmia”. “Atrial”, and subcategories including “atrial
fibrillation”, “atrial flutter”, and so on. “Junctional” and “Ventricular” and related
subcategories.
These changes are labeled as follows:
 Significant rate increase if the rhythms have the same condition and the current
rate is 20 or more beats per minute (bpm) greater than in the previous ECG.
 Significant rate decrease if the rhythms have the same condition and the current
rate is 20 or more bpm smaller than in the previous ECG.
 No change in rhythm.
 When a change occurs between sinus rhythm (SR) and sinus arrhythmia (SA), the
statement No significant rhythm change is added, since these are both types of
sinus rhythm.
A change in rate (20 bpm) is not labeled significant rhythm change unless the basic
rhythm statements change between normal sinus rhythm and sinus bradycardia or
sinus tachycardia.
Secondary rhythm changes. Secondary rhythm changes are always labeled as
No significant changes.
Secondary rhythm is the second rhythm statement on the report and comprises
premature complexes; pauses; AV block, AV conduction; miscellaneous arrhythmias;
pre-excitation.
These changes are labeled as follows:
 Rate increase when the rate is higher than in the previous ECG.
 Rate decrease when the rate is lower than in the previous ECG.
 No change in rhythm (= and Remains).

8 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

Table 3 Morphology-related rules used by the algorithm

Contour/Morphology Comparison
The algorithm looks at each statement in the previous ECG and develops a list of categories
that are present, and then compares the two ECGs for each of the categories at a time,
See Table 1 on page 6 for the threshold values used for key categories.
When examining Infarct categories, if the statement modifier in the current or previous
ECG is Acute, the program determines whether there are significant Q, ST, or T
changes in the appropriate leads, and labels the changes as follows:
 Evolving when the ECG date is within 14 days of the acute infarct date and relevant
measurement changes are significant.
 Evolution Continues when the ECG date is within 14 days of the acute infarct date,
the previous ECG had the Evolving statement, and there continue to be significant
evolution changes.
 No Further Evolution when the ECG date is within 14 days of the acute infarct
date, and there are no significant evolution changes.
 ST Persists when the ECG date is more than 14 days from the acute infarct date,
and relevant measurement changes are significant.

The following table summarizes the SC05 serial comparison statement set.
Table 4 SC05 serial comparison statement set

Code Statement Reason

IMI93 Inferior MI w/ persistent ST elevation Persistent J > .10 in

II III aVF
IMI97 Evolving inferior infarct ST-T changes in II III aVF
IMI98 Evolving inferior infarct continues ST-T changes in II III aVF
IMI99 No further evolution of inferior infarct No ST-T changes in
II III aVF
PMI93 Posterior MI w/persistent ST elevation Persistent J > .10 in V1 V2
PMI97 Evolving posterior infarct ST-T changes in V1 V2
PMI98 Evolution of posterior MI continues ST-T changes in V1 V2
PMI99 No further evolution of posterior MI No ST-T changes in V1 V2
LMI93 Lateral MI w/persistent ST elevation Persistent J > .10 in
I aVL V5 V6

Philips Serial Comparison Algorithm Reference 9

About the Philips Serial Comparison Algorithm

Table 4 SC05 serial comparison statement set (continued)

Code Statement Reason

LMI97 Evolving lateral infarct ST-T changes in

I aVL V5 V6
LMI98 Evolution of lateral MI continues ST-T changes in
I aVL V5 V6
LMI99 No further evolution of lateral MI No ST-T changes in
I aVL V5 V6
RMI93 Right ventricular MI w/ persistent ST
elevation“ infarct:date
RMI97 Evolving right ventricular infarct“
infarct:date
RMI98 Evolution of right ventricular MI
continues“ infarct:date
RMI99 No further evolution of right ventricular
MI“ infarct:date
CRVI Consider RV involvement
CPWI Consider posterior wall involvement
AMI93 Anterior MI w/ persistent ST elevation Persistent J > .10 in
V3 V4 V5
AMI97 Evolving anterior infarct ST-T changes in V3 V4 V5
AMI98 Evolution of anterior MI continues ST-T changes in V3 V4 V5
AMI99 No further evolution of anterior MI No ST-T changes in
V3 V4 V5
ALI93 Ant-Lat MI w/ persistent ST elevation Persistent J > .10 in
V4 V5 V6
ALI97 Evolving anterolateral infarct ST-T changes in V4 V5 V6
ALI98 Evolution of Ant-Lat MI continues ST-T changes in V4 V5 V6
ALI99 No further evolution of Ant-Lat MI No ST-T changes in
V4 V5 V6
CRVWI Consider right ventricular wall ST elevation in V1, V4R
involvement
LVHS Repolarization abnormalities Left ST-T changes
RVHS Repolarization abnormalities Right ST-T changes

10 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

Table 4 SC05 serial comparison statement set (continued)

Code Statement Reason

LBBR1 The appearance of BBB (bundle branch Heart rate change

block) may be rate related
LBBR2 The disappearance of BBB may be rate Heart rate change
related
VCDR1 The appearance of VCD (ventricular Heart rate change
conduction defect) may be rate related
VCDR2 The disappearance of VCD may be rate Heart rate change
related
BRAT1 Significant rate increase Rate > 20 bpm over previous
BRAT2 Significant rate decrease Rate less than previous
(20 bpm
EPERI Evolving pericarditis Significant T-axis change
STTNSC Nonspecific repolarization abnormalities

The applicable summary statement is added to the report, as well.

Table 5 Serial comparison summary statements for the 0A, 09, and 08 criteria
versions

Summary Statements
For each comparison performed, one of the following summary statements is added to the
ECG.
Significant Rhythm Changes Added when there is a change in the primary rhythm or
has a change of 20 bpm or more
Significant Contour Changes Added when Now Present, Now Absent, More
Prominent, Less Prominent modifiers are present
Significant Rhythm and Contour Added when a combination of the rhythm and contour
Changes statements are present.
No Significant Change Added when there is no significant change in
measurements.

Philips Serial Comparison Algorithm Reference 11

About the Philips Serial Comparison Algorithm

Table 6 SC05 Serial Comparison Summary Statements

Statement Criteria

No significant change Intervals (PR, QRS, QTc) remain normal or

close to a previously abnormal value. No new or
deleted diagnoses with the exception of normal
variant diagnoses.
Significant change in rhythm New or deleted rhythm diagnosis.
Significant change in rhythm: previously HR change >20 bpm and <50 or >100 bpm.
sinus New or deleted pacemaker diagnosis.

Significant change in rhythm: previously

junctional
Significant change in rhythm: previously
ectopic atrial
Significant change in rhythm: previously
atrial flutter
Significant change in rhythm: previously
atrial fibrillation
New or worsened ischemia or infarction Added infarction , ST-ischemia, or T-wave-
ischemia diagnosis, or worsened ST deviation
or T-wave abnormality.
New conduction abnormality Added AV or IV conduction diagnosis.
Significant repolarization change New or deleted QT diagnosis
New or deleted nonischemic ST or T-wave
diagnosis
Change in QTc
Significant axis, voltage or hypertrophy New or deleted diagnosis from Axis or Voltage,
change Chamber Hypertrophy, or Enlargement primary
statement categories or “Suggests...” secondary
statement category.
Change in interpretation without Used when a primary or secondary statement is
significant change in waveform added or removed despite no real change in the
tracing; i.e., an interpretive disagreement exists
between the readers of the first and second
ECGs.
QTc indicates corrected QT interval; HR, heart rate; bpm, beats per minute; AV,
atrioventricular; and IV, intraventricular

Ventricular conduction delay

The ventricular conduction delay (VCD) category presents particular challenges to the SC
program. Some of the statements generated by the DXL program in this category are related to

12 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

VCD which are based primarily on QRS duration, whereas others are related to pronounced
QRS axis deviation. Because some of these conditions can exist together, the DXL program
includes combination statements such as “Incomplete RBBB and LAFB” or “Borderline
IVCD with LAD.” Even more complications arise because some of these conditions may hide
others; for example, the presence of right bundle-branch block (RBBB) makes the diagnosis of
any underlying borderline incomplete VCD (IVCD) impossible. We can apply the same idea
to the axis deviation area and say that left axis deviation is “masked” by left anterior fascicular
block (LAFB).
In the SC algorithm design, we split the compound statements into their components. In the
design, the conditions are laid out in a directed graph so that as one progresses along the graph
(as in borderline IVCD to incomplete right bundlebranch block to RBBB), the later conditions
cover up (mask) the previous ones. For conditions that can coexist, such as LAFB and RBBB,
there is no path to get from one to the other.
When the comparison algorithm starts the VCD category analysis, any compound statements
are split into their component conditions. Those components are scanned to find any masking
relationships between current and previous conditions. If any masking condition is noted, the
statements will be changed according to the rules described in Table 7. Finally, the program
will attempt to recombine the results into compound statements whenever the appropriate
statement exists.
Table 7 Summary of the criteria used in the category of VCD

Masking direction SMD No SMD

(arrow direction in Fig. 3)

Masking
Current masks previous Label masking (current) Label masking (current)
(previous to current in statement “now present” and statement “insignificant
downward direction) suppress masked (previous) measurement” and suppress
condition masked (previous) condition
Unmasking
Previous unmasks current Label unmasked (current) Label masking (previous)
(previous to current in statement “remains” and statement “remains” and
upward direction) label masking (previous) suppress unmasked (current)
statement “now absent” condition

Right Ventricular Infarct/Injury

Because the system can process 12 leads with any combination of four additional chest leads,
the right precordial leads, especially ST-segment elevation in V4R, can be used to estimate the
presence of right ventricular infarct (RVI). Right ventricular infarct/injury is often associated
with acute inferior and/or posterior myocardial infarct/injury and occasionally exists by itself.
Because the likelihood of isolated RVI is low, the SC program is designed to require the
presence of either acute or recent inferior or posterior MI to indicate a diagnosis of RVI.
The RVI logic is illustrated in Figure 1 on page 15. Panel A shows the DXL interpretation
with the presence of an acute inferoposterior infarct. The SC interpretation in panel B of an

Philips Serial Comparison Algorithm Reference 13

About the Philips Serial Comparison Algorithm

ECG 16 minutes later indicates that there is an evolving inferior infarct and enough ST
elevation in V4R to trigger the RV involvement statement.

14 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

Figure 1 Panel A, The first ECG indicates the presence of inferoposterior infarct.
Panel B, The DXL interpretation indicates RV involvement in addition to
inferoposterior infarct.

Philips Serial Comparison Algorithm Reference 15

About the Philips Serial Comparison Algorithm

Requirements for Performing Serial Comparison

For serial comparison to be performed, specific requirements must be met, both for the current
ECG as well as for a “valid” previous ECG.
The requirements are divided here into two sections: system-defined (no configurable
parameters) and site specific (configurable by your administrator). For a comparison to be
performed, all of the requirements, system-defined, and site specific, must be met.
If the requirements are not met for either current or previous ECG, no comparison is
performed.
In addition, the ECGs must not meet any of the exclusion criteria.
All of the requirements are described in Table 8, “Criteria for serial comparison of ECGs,” on
page 16.

Table 8 Criteria for serial comparison of ECGs

Type of Criteria Description

Exclusion criteria ECGs that meet the conditions in this section cannot be serially
compared.
 Current and/or previous ECGs with paced rhythm or paced
complexes are not compared.
 Current and previous ECGs that are for a pediatric (age <
16) patient. (Serial change is generally not a major concern
in pediatric ECG analysis.)
System-defined criteria  Serial comparison must be enabled on the system by your
for performing serial IntelliSpace ECG administrator.
comparison
 Current and previous ECGs must have been analyzed with
a supported criteria version.
 Supported versions are 08, 09, 0A, and 0B.
 Current and previous ECGs must have interpretive
statements.
 If the current or previous EGG’s measurement matrix is not
present, the serial comparison algorithm performs a
statement-based comparison.
 If there is a mismatch in age or gender, or a significant
change is detected in height, a comparison is still
performed, but a statement is added to the current ECG
regarding the mismatch.
 Uncoded statements, remarks, and custom statements are
not compared.
 Current and previous ECG must have standard 12 or 12+
(e.g. 15-lead, 16-lead) electrode placement.
 Current and previous ECG must not have any “No further
analysis...” coded statements.

16 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

Table 8 Criteria for serial comparison of ECGs (continued)

Type of Criteria Description

Site-specific These settings are configured for your site by the IntelliSpace
(Administrator defined) ECG administrator using the System Configuration utility on
criteria for performing the server.
serial comparison
These requirements for serial comparison are in addition to
those described above.
NOTE The options below show the factory default setting. Check
with your administrator to determine whether changes were made.
 Previous ECG is or is not confirmed, as configured at your
site.
Contact your IntelliSpace ECG administrator to determine
what the setting is.
 The algorithm checks whether the previous ECG is more
than a few minutes older than the current ECG to be used in
comparison, based on the time interval configured by your
administrator.
If the time difference does not meet the requirements, the
algorithm checks the next previous ECG. By default, the
time difference is set to 30 minutes. You can decrease this
time interval, if desired, in 1 minute increments, to a
minimum of 1 minute.
 Previous ECG must be from same institution, and
optionally, facility (or not) as configured at your site.
The location-related options depend on a variety of system
configuration factors. Talk with your administrator about
the exact location requirements for serial comparison at
your site.

Philips Serial Comparison Algorithm Reference 17

About the Philips Serial Comparison Algorithm

Serial Comparison Statement Display Options

The IntelliSpace ECG administrator can define the following site-wide options for displaying
serial comparison statements.
The administrator can configure whether to:
 Prefer current statement over previous for similar statements, regardless of severity.
When disabled, and a previous ECG and current ECG both have different statements that
are in the same diagnostic category and no significant measurement differences are
detected, the more severe statement is displayed (whether in previous or current ECG).
 Not display previous statements with insignificant measurement change.
When enabled, the current statement is displayed regardless of relative severity.

Comparing ECGs Using the Serial Comparison Algorithm

NOTE Automatic comparison of ECGs is not supported in the Basic Edition of IntelliSpace ECG A.01.
You can perform a comparison between the current ECG and the most recent ECG that meets
the configured criteria. IntelliSpace ECG can also be configured to automatically perform
serial comparison when a certain action takes place, for example, when specific ECGs enter
the system. For details, contact your system administrator.
You can also perform a visual comparison between two ECGs by displaying them both on the
screen at the same time in Waveform view. For details, see the section on using the Waveform
view in the ECGVue or PharmaVue Instructions for Use.

To perform a manual serial comparison between two ECGs

1 Select an ECG to review for a specific patient.
This is the ECG that IntelliSpace ECG will compare to the most recent previous ECG
(that meets the comparison criteria) for this patient.
2 Click Compare in the Action bar.
IntelliSpace ECG searches for the most recent previous ECG that matches the configured
criteria, and performs the serial comparison. The serial comparison algorithm also adds
and/or deletes statements, as appropriate.
Note that the button label changes to Uncompare. You can undo the comparison, if
necessary, as described in the next section.
3 To save the serial comparison statements with the ECG, click Save in the Action bar.

Undoing a Serial Comparison

There may be occasions when you wish to recompare an ECG that has already been serially
compared. For example, the ECG to which it was compared by have been invalid, or not the
ECG you wanted to compare to, or a more recent, more representative ECG was acquired that
you wish to compare to.
The Compare button in the Action bar changes to Uncompare when an ECG is selected that
was serially compared.

18 Philips Serial Comparison Algorithm Reference

 How the Serial Comparison Algorithm Analyzes Data

When undoing a comparison, the system removes any statements that were added to the ECG
during serial comparison, as well as any edits that were made to the statements after the serial
comparison was performed. The ECG is returned to the state it was in immediately prior to
comparison. After saving the changes, you can perform a new serial comparison on the ECG,
if desired.
NOTE If an ECG has undergone multiple serial comparisons, any changes prior to the immediately previous
serial comparison are not modified.

To undo a serial comparison

1 Display the ECG to “uncompare”.
2 Click Uncompare in the Action bar.
The application reverts to the interpretive statements that were present prior to the serial
comparison, and makes the ECG available for recomparison.
3 To save the changes to the ECG to the database, click Save in the Action bar.
The Compare button is enabled in the Action bar. You can perform another serial
comparison on the ECG, if desired.

Philips Serial Comparison Algorithm Reference 19

About the Philips Serial Comparison Algorithm

20 Philips Serial Comparison Algorithm Reference

Philips Medical Systems Authorized EU-representative:
3000 Minuteman Road Philips Medizin Systeme
Andover, MA 01810 USA Böblingen GmbH
Hewlett Packard Str. 2
Copyright © 2015 71034 Böblingen
Koninklijke Philips Electronics N.V. Germany
All rights are reserved.

IntelliSpace ECG ECG Management System

Serial Comparison Algorithm Reference

Printed in USA

453564373101
Revision B
December 2015

*453564373101*
*B*