Surgery – Diagnostic testing – Detecting brain electric signal
Reexamination Certificate
2000-01-07
2002-01-29
Nasser, Robert L. (Department: 3736)
Surgery
Diagnostic testing
Detecting brain electric signal
C600S559000
Reexamination Certificate
active
06343230
ABSTRACT:
RELATED APPLICATIONS
This application is related to the co-pending and commonly assigned U.S. Patent Application entitled “Hearing Evaluation Device with Patient Connection Evaluation Capabilities,” Ser. No. 09/479,559 filed by Matthijs P. Smits, Vineet Bansal, Abraham J. Totah and Bryan P. Flaherty and the U.S. Patent Application entitled “Hearing Evaluation with Noise Detection and Evaluation Capability,” Ser. No. 09/479,548 filed by Matthijs P. Smits and Bryan P. Flaherty, the disclosures of which are hereby incorporated herein by reference.
FIELD OF THE INVENTION
The present invention relates to devices and methods that use electroencephalographic responses to auditory stimuli to evaluate the hearing of a subject, and that are capable of quickly and accurately determining when a subject is unlikely to pass the hearing loss test.
BACKGROUND OF THE INVENTION
In the past, hearing impairment in babies and children was often not detected until after it was observed that the baby or child did not respond normally to sound. Unfortunately, it often took months or even years for the parent to observe the impairment, and by that time the child's language and learning abilities were negatively and often irreversibly impacted. Indeed, recent studies indicate that the vocabulary skills of hearing impaired children markedly increases if their hearing loss is detected early. The optimal time to evaluate hearing loss is thus immediately after birth, both because early detection allows for early treatment, and because parents often fail to bring their infants to later appointments. As a result, a number of states have implemented programs to evaluate newborns for hearing loss.
However, babies, especially newborns, cannot participate in traditional hearing tests, which require the subject to indicate if he or she hears the auditory stimulus. Thus, devices and methods have been developed to objectively determine hearing loss, without the voluntary participation of the subject. One such method involves analysis of the involuntary electroencephalographic (EEG) signals that are evoked from a subject in response to an auditory stimulus. It has been found that when a subject is able to perceive a sound having particular characteristics, a specific EEG waveform known as an Auditory Brainstem Response (ABR) is generated. This ABR response signal is typically small in magnitude in relation to general EEG activity. Therefore, statistical and signal processing techniques have been employed and developed to help detect, to a pre-defined level of statistical confidence, whether an ABR response has in fact been evoked. ABR testing is especially applicable to evaluation of infants, but can be applied to any subject.
The ABR that is evoked in response to the auditory stimulus may be measured by use of surface electrodes on the scalp or neck. As a practical matter, the electrodes will also detect noise signals from neural activity (besides the ABR), muscle activity, and non-physiological environmental noises.
Especially for subjects who do not pass the hearing evaluation, ABR testing can be time consuming. It has been found that in many cases a subject's failure to pass can be adequately predicted after a short period of time. This information can be conveyed to the test operator, who may then decide whether to continue with a full length test or not.
The present invention represents a major advance in the art because it allows for early and accurate prediction in many cases that a subject will not pass the evaluation. The resultant time savings may not only allow for more tests per unit time, but may also make such tests more attractive to parents and health care providers.
DESCRIPTION OF THE PRIOR ART
Under prior art, a subject will “pass” the evaluation as soon as enough data have been collected to conclude, within a statistical level of confidence, that an ABR waveform is in fact present. If the data are insufficient to make this conclusion, then the evaluation continues until the total number of individual response signals—known as “sweeps”—exceeds a preset threshold. Prior art devices have sed a threshold of 15,000 sweeps. If testing concludes because the number of sweeps equals the maximum, then the subject would typically be referred for further testing to determine if in fact he or she suffers from hearing impairment.
Given the special difficulties in hearing evaluation of infants, it is important to minimize the time required to conduct a hearing evaluation. It is also important to minimize the time required for evaluation given the goal of universal evaluation. However, under the prior art, a non-hearing subject would need to undergo an entire, full length evaluation session, regardless of whether the data were ambiguous, or whether they showed that it was very unlikely that the subject would pass. Full-length evaluations may last as long as an hour.
OBJECTS AND SUMMARY OF THE INVENTION
The present invention provides a device and method for use in analyzing the EEG signal evoked in response to the auditory click stimulus, to determine if the subject suffers from hearing loss. Broadly, the invention is directed to devices and methods that are capable of predicting a failure to pass a hearing evaluation, based upon an analysis with normative data, and using certain statistical techniques.
In one embodiment of the invention, evoked EEG responses to auditory stimuli are collected, and organized into “sweeps,” with each sweep containing the response signal for one auditory stimulus. The sweeps are organized into B blocks, with each block b containing a number of sweeps N
b
.
The response signal for each sweep is digitized and converted into a series of binary numbers corresponding to whether the amplitude of the response signal is positive or negative at various points in time. The digitized, binary waveform is compared to a benchmark ABR waveform to determine if the ABR is present. To make this determination, a polarity sum is calculated, which represents the sum of the polarities of the response signals within all blocks at each measured point in time. Statistical techniques are then used to determine if an ABR is present, relying upon the expected distribution of polarity sums in the absence of an ABR. This expected distribution is developed theoretically. A “Pass” is triggered if the observed polarity distribution, as represented in a specifically defined test statistic, indicates that the likelihood that an ABR is present is above a predetermined threshold. After a certain number of blocks have been completed, evaluation will cease if a “Pass” has yet not been triggered. Under such circumstances, the subject will be referred for further testing to determine if he or she in fact does suffer from hearing loss.
In accordance with the present invention, the polarity distribution for each block is evaluated independently, and compared with normative data for a hearing subject. These normative data reflect the expected polarity distribution for a hearing subject, given certain testing conditions. In particular, the normative data account for the prevailing noise conditions.
If the average difference between the expected and the observed polarity distributions for all blocks is too great, a “predicted refer” indication is given. The operator may then continue testing (knowing that a predicted refer has been generated), or may terminate the evaluation. The present invention therefore improves upon the prior art because it provides an accurate way to predict a failure to pass, therefore conserving evaluation times and resources, without significantly sacrificing evaluation accuracy.
Although the embodiments described here are directed towards evaluation of newborn hearing, it is believed that the present invention can be applied to any evaluation, whether of hearing or not, in which evoked potentials are analyzed.
As described below, the present invention makes extensive use of normative data. These normative data were derived from analysis of clinical data, and from computer simulations repres
Coppin Christophe M.
Smits Matthijs P.
Nasser Robert L.
Natnithithadha Navin
Natus Medical Inc.
Orrick Herrington & Sutcliffe LLP
LandOfFree
Hearing evaluation device with predictive capabilities does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Hearing evaluation device with predictive capabilities, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Hearing evaluation device with predictive capabilities will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2852786