Biomedical subjects
Publications and source records attributed to W H McFarland.
A newborn hearing screening technique (the Crib-o-gram) originally proposed by Simmons and Russ (1974) is described. The number of babies tested has now grown to over 12,000. The report presents follow-up data on the Crib-o-grain hearing screening program at Stanford University Medical Center. Improvements in test procedure and scoring have resulted in a current detection rate of 93%.
Explore the source record for details and available documents.
Middle-component AERs (8-90 msec) to tone-pips from 10 normal-hearing adults were subjected to three objective methods of response identification. Threshold was then determined for each subject according to four different rules. The criterion score, which considers conjointly latency and amplitude values across the middle-component peaks, was developed as a single-value measure for response-identification and subsequent threshold-determination procedures. One of the response-identification methods was applied to 10 hearing-impaired subjects; the results of the threshold-determination procedures were encouraging. Further directions toward improving objective response analysis are discussed.
Tone-pips of 500, 1000, and 3000 Hz were presented at 0-, 10-, 20-, 35-, and 50-dB HL to 10 normal-hearing subjects and at 0-, 10-, 20-, 35-, and 50-dB SL to 10 subjects with conductive, sensorineural, or mixed hearing losses. Middle component (latencies 8-90 msec) averaged electroencephalic responses to the tone-pips were analyzed in terms of peak latencies and peak-to-peak amplitudes. Properties of the responses were generally the same for both normal-hearing and hearing-impaired subjects except that the hearing-impaired subjects showed slightly greater amplitudes overall. The small reduction in latencies with increasing stimulus frequency seen in the normal-hearing subjects was not observed in the hearing-impaired subjects.
The middle components of the evoked cortical response (8-50 msec) were examined under improved signal processing conditions. 512 click stimuli were presented at 5 rates ranging from 1-16/sec to 10 normal-hearing subjects. The influence of stimulus numbers of 32, 64, 128 and 512 and stimulus rates of 1, 2, 4, 8 and 16/sec was examined. Identifiable and repeatable responses were found with as few as 128 stimuli. Stimulus rate had little effect on middle component waveform or its identifiability. An enhancement of signal-to-noise ratio through improved filter conditions is suggested as one reason for the ability to identify middle components to fewer stimuli.
11 182 newborn infants have had their hearing thresholds screened by automatically recording crib motion before and after a test sound. An inexpensive motion transducer translates the baby(s) movements onto a multichannel strip-chart recorder (batch testing) or a single channel recorder (individual testing). Records are read off-line and scored positive if there is a change in movement within 2.5 sec after the test sound. "Passing or failing" a baby requires about one minute's clerical scoring time, and is therefore both economical and simple. We have detected 33 neonates with serious hearing losses and perhaps missed one on long-term follow-up. This amounts to a deafness incidence of 1:329 live births. In the Intensive Care/Premature nurseries the incidence is 1:62.
The bases for variations of the middle-component (8-50 ms) auditory averaged electroencephalic response (AER) to clicks during the averaging process were explored by examining (1) the trend of the mean variance of the averaged waveform, (2) averaged waveforms generated by successive blocks of stimuli fractioned from a total of 512 stimuli, and (3) averaged waveforms generated by successive, but partially overlapping, blocks of responses recorded for a train of 512 stimuli. These analyses indicated no systematic changes in peak latencies or peak-to-peak amplitudes as stimulation progressed. Fluctuations in middle AER waveform are more readily explained by non-stationarity of the background electrophysiologic noise. Previously reported amplitude reduction with increasing stimulus number possibly can be explained by progressive reduction of the background noise on which the consistent-amplitude middle components are superimposed.