THE STACKED ABR

A Successful Small Acoustic Tumor Screening Method

Manuel Don, Ph.D.

Betty Kwong, M.S., CCC-A
Chiemi Tanaka, M.A., CCC-A

2100 W. Third St., Los Angeles, CA 90057

Phone (213) 483-4431 Fax (213) 413-6739

HOUSE EAR INSTITUTE

 Standard ABR
Cross-section of Internal Auditory Canal
The wave V latency used in the
Facial Nerve Sup. Vest Nerve
standard ABR IT5 and I-V delay
measures is dominated by neural
activity from the high-frequency
HIGH- regions of the cochlea. Thus, unless
FREQUENCY the tumor affects these high-
FIBERS
frequency fibers sufficiently,
standard ABR latencies will be
normal.
Acoustic Nerve Inf. Vest Nerve

Large Tumor Small Tumor Small Tumor

Abnormal Standard ABR Abnormal Standard ABR Normal Standard ABR

TUMOR

HOUSE EAR INSTITUTE

 Stacked ABR
Cross-section of the Required for Successful
Internal Auditory Canal Detection of Small Acoustic Tumors
Activity from area 1 1 Facial Nerve Sup. Vest Nerve ¾ An auditory signal that will stimulate
+ essentially all frequency regions of the
Activity from area 2 2 cochlea (i.e., an appropriate click)
+
Activity from area 3 3 ¾ A method for separating out responses
+ from the different frequency regions of
Activity from area 4 4 the cochlea (i.e., derived-band ABR
+ technique)
Activity from area 5 5 2
1 3 ¾ A procedure of adding the responses
5 together to approximate total neural
Total Neural Response 4
activity (i.e., stacking method)
= Sum of Activity from
five frequency regions Acoustic Nerve Inf. Vest Nerve

Large Tumor Small Tumor Small Tumor

Abnormal Stacked ABR Abnormal Stacked ABR Abnormal Stacked ABR

NORMAL NORMAL NORMAL

1 1 TUMOR
1
2 2 2 TUMOR

TUMOR
3 3 3
4 4 4
5 5 5
TUMOR
Standard ABR missed this tumor!

HOUSE EAR INSTITUTE

The Derived-band Technique
Click stimuli are delivered in the presence of high-pass masking noise.
The cutoff frequency of the high-pass noise is lowered from one run to
the next. This process masks progressively lower frequency areas of the
cochlea. Subtracting the response for one run from the previous one
forms a derived-band response. Here, the response to clicks + 8 kHz
high-pass masking noise is subtracted from the response to clicks alone
to form the derived-band ABR with center frequency (CF) = 11.3 kHz.
Base Frequency kHz Apex
3kHz.8 4 2 1 0.5

Schematic of Cochlea

Clicks alone
(Standard ABR)
Nerve Fibers

Derived-band ABR
Cross-section of
Auditory Nerve
CF = 11.3 kHz

Base Frequency kHz Apex Base Frequency kHz Apex

8 4 2 1 0.5 8 4 2 1 0.5

Clicks + 8 kHz high-pass

Nerve Fibers masking noise Nerve Fibers

HOUSE EAR INSTITUTE

 Frequency kHz
Base Frequency kHz Apex The response to clicks + 4 kHz high- Base
8 4 2 1
Apex
0.5 The response to clicks + 1 kHz high-
8 4 2 1 0.5
pass masking noise is subtracted M M M pass masking noise is subtracted
M from the response to clicks + 8 kHz from the response to clicks + 2 kHz
high-pass masking noise to form the high-pass masking noise to form the
Clicks + 8 kHz high- derived-band ABR with CF = 5.7 kHz. derived-band ABR with CF = 1.4 kHz.
Clicks + 2 kHz high-
pass masking noise Nerve Fibers pass masking noise
Nerve Fibers
Derived-band Derived-band
ABR CF = 5.7 kHz ABR CF = 1.4 kHz

Base Frequency kHz Apex Frequency kHz

Base Apex
8 4 2 1 0.5 8 4 2 1 0.5

M M M M
Base Frequency kHz Apex Base Frequency kHz Apex
8 4 2 1 0.5 8 4 2 1 0.5

M M M M M M

Clicks + 4 kHz high- Nerve Fibers Nerve Fibers

Clicks + 1 kHz high-
pass masking noise pass masking noise
Nerve Fibers Nerve Fibers

Base Frequency kHz Apex The response to clicks + 2 kHz high- Base Frequency kHz Apex The response to clicks + 0.5 kHz high-
8 4 2 1 0.5 8 4 2 1 0.5
pass masking noise is subtracted pass masking noise is subtracted
M M M M M M
from the response to clicks + 4 kHz from the response to clicks + 1 kHz
high-pass masking noise to form the high-pass masking noise to form the
derived-band ABR with CF = 2.8 kHz. Clicks + 1 kHz high- derived-band ABR with CF = 0.7 kHz.
Clicks + 4 kHz high-
pass masking noise pass masking noise
Nerve Fibers
Nerve Fibers
Derived-band Derived-band
ABR CF = 2.8 kHz ABR CF = 0.7 kHz

Base Frequency kHz Apex Base Frequency kHz Apex

8 4 2 1 0.5 8 4 2 1 0.5

Frequency kHz Base Frequency kHz Apex

Base
8 4 2 1
Apex
0.5
M M 8 4 2 1 0.5 M M M M
M M M M M M M M

Clicks + 2 kHz high- Nerve Fibers Clicks + 0.5 kHz

Nerve Fibers
pass masking noise high-pass masking
Nerve Fibers Nerve Fibers
noise

HOUSE EAR INSTITUTE

 The Stacking Method
V
Sum of
Unmasked Shifted
(Standard) ABR Waveforms

V
Stacked
CF = 11.3 kHz ABR
V

CF = 5.7 kHz

V
Shifted to 5.7 kHz
CF = 2.8 kHz Wave V latency
CF = 11.3 kHz
V
CF = 5.7 kHz
CF = 1.4 kHz
CF = 2.8 kHz
CF = 1.4 kHz
V
CF = 0.7 kHz CF = 0.7 kHz

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
ms ms

Derived-band ABR Summary Stacked ABR Summary

¾ Neural contributions from different frequency regions
of the cochlea can be obtained using the derived-
band ABR method.
¾ Derived-band ABRs represent activity from more
specific frequency regions than moderate-to-high
level toneburst-evoked ABRs.
¾ The Stacked ABR is formed by temporally aligning wave V
of the derived-band ABRs and then summing the
responses.
¾ Aligning the derived-band ABRs eliminates phase
cancellation of lower frequency activity. Thus, the Stacked
ABR amplitude reflects activity from all frequency regions of
the cochlea, not just the high frequencies.
¾ Reduction of any neural activity due to a tumor, even a
small tumor, will result in a reduction of the Stacked ABR
amplitude.

HOUSE EAR INSTITUTE

 Standard vs Stacked ABR Measures

100
For 95% sensitivity
90
(that is, for correct identification of
80 95 out of every 100 small tumors):

70 The IT5 and I-V Delay

measures have less
Specificity

60
than 5% specificity
50 (that is, the IT5 and I-V Delay
correctly identify less than 5 out of
40 every 100 non-tumor patients),
30 But the Stacked ABR
Stacked ABR has 83% specificity
20
IT5
I-V Delay
(that is, the Stacked ABR
10 correctly identifies 83 out of every
0 100 non-tumor patients)!

100 90 80 70 60 50 40 30 20 10 0
Sensitivity

95%

HOUSE EAR INSTITUTE

 CONCLUSION

The Stacked ABR appears to have better

sensitivity and specificity than the Standard
ABR for small ( < 1 cm) acoustic tumors.

In other words, the Stacked ABR is better at :

1. detecting small tumors, and
2. decreasing the number of misdiagnosed
non-tumor patients (i.e., decreasing the
number of false-positives referred for MRI).

HOUSE EAR INSTITUTE

Perform Standard ABR SCREENING
ABR Analyses PROTOCOL FOR
(IT5, I-V, etc.)
ACOUSTIC TUMORS
Evaluate for
auditory
No Send For An No neuropathy
Normal? MRI Tumor? and/or
refer for
neurological
Yes Yes evaluation

Perform Stacked Prescribe

ABR Analyses Treatment
(e.g., surgery)

No Send For An Yes

Normal? Tumor?
MRI

Yes Observe? No
Follow?

HOUSE EAR INSTITUTE