Data processing: speech signal processing – linguistics – language – Speech signal processing – Synthesis
Reexamination Certificate
1998-10-07
2001-09-11
Korzuch, William (Department: 2641)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Synthesis
C704S271000, C381S320000, C434S185000
Reexamination Certificate
active
06289310
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates in general to the field of auditory testing of humans. More specifically, this invention relates to an apparatus that uses an acoustic processing profile derived from a computer program that universally screens individuals for auditory discrimination problems associated with spoken language.
2. Description of the Related Art
Modern research indicates that up to ten percent of humans have language-learning impairments (LLI) resulting from the inability to accurately process short duration acoustic events at rates that occur in normal speech. Their trouble distinguishing among elements of speech is neurologically based and has far reaching consequences: academic failure, emotional and disciplinary problems, and possibly diminished lifelong achievement and self-image. No bracket of intelligence, race, gender or economic level is immune from this problem.
More specifically, people with LLI have difficulty detecting and identifying sounds that occur simultaneously or in close proximity to each other—a phenomenon known as “masking.” Because of masking, people with LLI require sounds that are as much as 45 decibels more intense than preceding or subsequent masking noises to distinguish and understand them. In addition, people with LLI are consistently poorer at detecting a brief tone presented with a masking noise, particularly when the brief tone is turned on immediately prior to the masking noise. This phenomenon is called “backward masking.” Similarly, when the brief tone is turned on immediately after the masking noise a similar decrease in detectability can occur. This phenomenon is called “forward masking”. For a tone to be detected by a person with LLI in the presence of a masking noise, the tone must be separated in time or frequency from the masking noise.
The inability to accurately distinguish and process short duration sounds often cause individuals to fall behind in school. Since the individuals can't accurately interpret many language sounds, they can't remember which symbols represent which sounds. This deficiency causes difficulties in learning to read (translating from symbols to sounds), and in spelling (translating from sounds to symbols). In fact, it is common for an individual with LLI to fall two to three years behind his/her peers in speech, language and reading development.
One way individuals develop such auditory processing problems is from middle ear infections when they are young and beginning to develop the oral representations of language in the central auditory nervous system. For example, when a child has an ear infection, fluid can build up and block or muffle the sound wave entering the ear causing intermittent hearing loss. Even if the infection doesn't permanently damage the ear, the child's brain doesn't learn to process some sounds because it hasn't heard them accurately before, on a consistent basis. This typically occurs during a critical period of brain development when the brain is building the nerve connections necessary to accurately process acoustic events associated with normal speech.
Researchers believe that the auditory processing problem is essentially one of timing. Vowel sounds like /a/ and /e/ usually last at least 100 milliseconds and typically have constant frequency content. Consonants, on the other hand, typically have modulated frequency components, and last less than 40 milliseconds. Individuals with LLI cannot process these faster speech elements, especially the hard consonants like /t/, /p/, /d/ and /b/, if they occur either immediately before or after vowels, or if they are located near other consonants. Rather than hearing the individual sounds that make up a particular phoneme, individuals with LLI integrate closely associated sounds together over time. Since the duration of vowels are typically longer than consonants, the modulated frequency portions of consonants are often lost in the integration, an affect that may also hinder the resolution of the vowel, particularly short duration vowels.
This problem of abnormal temporal integration of acoustic events over time is not limited to individuals with LLI. Rather, the problem extends to stroke victims who have lost the neurological connections necessary to process speech, as well as to individuals raised in one country, having one set of language phonemes, and attempting to learn the language of another country, having a distinct set of language phonemes. For example, it is known that an individual raised in Japan is not often presented with phonemes similar to the English r's and l's, because those consonants are not common in the Japanese language. Similarly, there are many subtleties in the sounds made by a speaker of Japanese that are difficult to distinguish unless raised in Japan. The phonetic differences between languages are distinctions that must be learned, and are often very difficult. But, they are clearly problems that relate to the temporal processing of short duration acoustic events.
The above described temporal processing deficiency has little if anything to do with intelligence. In fact, some LLI specialists argue that brains choosing this different route by which to absorb and reassemble bits of speech may actually stimulate creative intelligence, but at the expense of speech and reading problems.
Recent studies have shown that if the acoustic events associated with phonemes that are difficult to distinguish, such as /ba/ and /da/, are slowed down, or that the consonant portion of the phonemes are emphasized, that individuals diagnosed with language impairments can accurately distinguish between the phonemes. In addition, if the interval between two complex sounds is lengthened, individuals are better able to process the sounds distinctly.
Heretofore, the solution to the processing problem has been to place individuals with language impairments in extended special education and/or speech therapy training programs that focus on speech recognition and speech production. Or, more commonly, repetitive reading programs, phonic games, or other phonic programs are undertaken. These programs often last for years, with a success rate that is often more closely associated with the skill of the speech and language professional than with the program of study.
Another problem associated with abnormal temporal integration is one of detection. That is, modern hearing tests are not designed to evaluate whether an individual has one of the above-described masking, or integration problems. Rather, hearing tests typically determine whether an individual can hear particular frequencies, at particular amplitudes. The tests do not determine whether the individual can process short duration acoustic events in the presence of masking acoustic events. If tests indicate that an individual cannot hear particular frequencies, hearing aids may be recommended. However, hearing aids typically just amplify acoustic events within a particular frequency range, without regard to the content of the acoustic events. That is, equal emphasis is provided to all signals within a given frequency range, while acoustic signals outside of the given range (background noise for example) are eliminated.
Alternatively, tests used to determine whether an individual is language learning impaired are often provided in the form of reading tests, rather than aural tests. However, as hinted at above, failure to perform well in school, or more specifically, to properly process phonemes common in spoken language, have more to do with the processing of acoustic events than with reading. Thus, reading tests are inadequate in determining whether an individual properly processes acoustic events common in spoken language.
What is needed is a method and apparatus that acoustically screens individuals to determine whether they properly process acoustic events that are common in spoken language. More specifically, what is needed is a program that can be easily executed by individuals, of all ages, ge
Miller Steven L.
Peterson Bret E.
Protopapas Athanassios
Huffman James W.
Korzuch William
Scientific Learning Corp.
Storm Donald L.
LandOfFree
Apparatus for enhancing phoneme differences according to... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus for enhancing phoneme differences according to..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus for enhancing phoneme differences according to... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2544388