Absolute Identification Research Articles

Perception of multiple electrode stimulus patterns: Implications for design of an electrotactile speech processor

Identification of multiple-electrode stimulus patterns was evaluated in nine adult subjects, to assess the feasibility of providing additional speech information through the tactual display of an electrotactile speech processor. Absolute identification scores decreased from 97.8% for single electrodes, to 61.9% for electrode pairs, and, to 31.8% for electrode triplets. Although input information increased with paired-and triple-electrode stimuli, information transmission scores were not significantly increased for either electrode pairs (2.99 bits) or triplets (2.84 bits) as compared with single electrodes (2.84 bits). These results suggest that speech coding strategies using stimulus patterns of electrode pairs or triplets would provide little improvement beyond that found for the present single-electrode scheme. However, higher absolute identification scores (73.6%), and an increase in information transmission to 3.88 bits, were recorded for test stimuli containing all combinations of paired and single electrodes. Based on this finding, two stimulus sets using a restricted number of combinations of paired and single electrodes were evaluated. The two stimulus sets simulated the spatial patterns of paired and single electrodes arising from use of alternative speech coding schemes to increase consonant voicing information. Results for the two stimulus sets showed higher electrode identification scores (79.7% and 90.4%), as compared with paired-electrode stimuli. Although electrode identification score was not as high as for single electrodes, information transmission was increased to 3.31 bits for the VF2 stimulus set. Analysis of the responses also showed that scores for identification of simulated voicing information conveyed by the two stimulus sets were 99.4 and 90.4% correct.(ABSTRACT TRUNCATED AT 250 WORDS)

Electron Microscopic Definition of Intestinal Endocrine Cells: Immunogrold Localization and Review

Intestinal mucosal biopsy specimens processed during the past 25 years were used to example the ultrastructural characteristics of intestinal endocrine cells. The cells were defined on the basis of morphologic criteria and, when feasible, with specific antisera and immunogold staining. The hypothesis was that each endocrine cell, once well defined, should be identifiable on the basis of standard morphologic criteria not requiring specific immunostaining. This was not the case, D, G, EC1, EC2, ECn, D1, and intestinal gastrin cells have characteristic secretory granules and, when sufficient granules are present, can be identified consistently on the basis of morphologic criteria. Absolute identification of D, G, IG, and TG cells requires staining with specific antisera, a condition easily obtainable only for D, G, and IG cells. D1, EC1, EC2, and ECn cells must be identified morphologically until secretory products specific for each of these cells are identified. I, L, N, and K cells are remarkably similar in appearance and must be distinguished by specific staining. Mo, S, and P cells were not identified by either morphologic appearance or immunostaining. It is suggested that a cell similar to the D1 cell but with exceptionally small granules may be the P cell. Absolute identification of intestinal enteroendocrine cells by electron microscopy requires specific staining. The characteristic appearance of the secretory granules of many of these cells (D, G, EC1, EC2, ECn, D1, and IG) permits morphologic identification when numerous secretory granules are present.

Connectionist Models of Orientation Identification

Abstract We have used connectionist simulations in an attempt to understand how orientation tuned units similar to those found in the visual cortex can be used to perform psychophysical tasks involving absolute identification of stimulus orientation. In one task, the observer (or the network) was trained to identify which of two possible orientations had been presented, whereas in a second task there were 10 possible orientations that had to be identified. By determining asymptotic performance levels with stimuli separated to different extents it is possible to generate a psychophysical function relating identification performance to stimulus separation. Comparisons between the performance functions of neural networks with those found for human subjects performing equivalent tasks led us to the following conclusions. Firstly, we found that the ‘psychometric functions’ generated for the networks could accurately mimic the performance of the human observers. Secondly, the most important orientation selectiv...

Unmasking the magnitude estimation response.

The fuzzy judgement model of Ward (1979) predicts an inverse relation between the amount of stimulus information available to subjects and the magnitude of sequential dependencies on previous stimuli and responses in psychophysical scaling tasks. Ward confirmed this prediction for magnitude estimations of interdot distance for previous responses but not for previous stimuli, although the inverse relation has been repeatedly reported for both the previous stimuli and responses in absolute identification (e.g., Mori, 1989). This paper further explores this seemingly puzzling contradiction. A magnitude estimation of loudness experiment was conducted in which the amount of stimulus information available to subjects was manipulated by a modified version of informational masking (Watson, 1987). An absolute-identification-with-feedback experiment was also conducted to check the effectiveness of the informational masking in reducing the amount of stimulus information. The results of the magnitude estimation experiment show a striking similarity with those of Ward and generalize the failure of sequential dependencies on previous stimuli to vary inversely with stimulus information. An additional assumption that judgement strategies are altered under low-information conditions is necessary to explain this result.

The influence of duration on the perception of pitch in single and simultaneous complex tones.

The influence of duration on the virtual pitch of complex tones was measured using an absolute identification paradigm. If performance with two-tone complexes is expressed in terms of a single central frequency-coding noise function, this function is found to depend on duration in about the same way as the pure-tone difference limen function. The function is further found to be a reasonably good predictor of pitch identification performance with multitone complexes. Another experimental finding was that subjects tend to switch to the analytic mode of pitch perception when complex tones are shortened (i.e., they tend to hear the spectral pitches instead of the virtual ones). A third finding was that with simultaneous complex tones the degradation of each pitch percept depends not only on duration and harmonic order of the tone but also on the harmonic order of the other tone.

Absolute Pitch Identification: Effects of Timbre and Pitch Region

Pitch identification by absolute-pitch (AP) possessors was tested with three different timbres. Twelve notes per octave in the equally tempered chromatic scale were presented randomly over a seven-octave range. AP subjects identified the note quite accurately but showed a considerable number of octave errors, indicating that they indeed responded primarily to pitch classes. By contrast, subjects lacking AP gave widely scattered responses, indicating that they could not perceive pitch classes at all but responded solely on the basis of the approximate pitch region (pitch height). The performance of AP subjects was best for piano tones and worst for pure tones, and responses were most accurate in the middle region of the musical pitch range. Of 12 pitch classes, white-key notes on a musical keyboard (diatonic tones in the C major scale) tended to be identified more correctly and rapidly than were black-key notes (nondiatonic tones in the C major scale). All of the AP subjects had received training on the piano since a young age. It is suggested that AP can be acquired through such early musical experience.

Analysis of a synthetic Tadoma system as a multidimensional tactile display

The Tadoma method is a means of speech reception based on tactile monitoring of the articulatory process. A "synthetic" Tadoma system, involving an artificial face with six facial actions, has been developed as a first-order approximation to the natural Tadoma system. Experiments were conducted to explore the information-transmission characteristics of the synthetic Tadoma system in terms of the four facial movements it incorporates: upper lip in-out, lower lip in-out, lower lip up-down, and jaw up-down movements. Discrimination experiments showed that the just-noticeable difference associated with each movement is about 9% of the reference displacement. One-dimensional (1-D) absolute identification experiments produced, on the average, 1.6 bits of information transfer. Four dimensional (4-D) identification experiments produced information transfers in the range of 3-4 bits. Of the four dimensions considered, performance on the lower lip up-down movement was most affected, and performance on the jaw up-down movement was least affected, by simultaneous roving movements on the other dimensions. As a result of the interaction among the movement channels, the sum of the 1-D information transfers exceeds the 4-D information transfer. However, the sum of the 1-D information transfers obtained from tests with roving parameters is approximately equal to the 4-D information transfer (possibly exemplifying a "generalized information-transfer additivity law"). In general, both the discrimination and identification results appear unexceptional and, hence, the reception of facial movement information by itself does not appear to account for the extraordinary success of the Tadoma method.

Boundary conditions on parallel processing in human vision.

A new theory of visual search is tested experimentally with simple colour patches. The essential element of this new theory is that, whatever the search materials, efficiency increases continuously with (i) decreasing similarity between targets and nontargets, and (ii) increasing similarity between one nontarget and another. Control of 'attention' (access to visual short-term memory) is seen as a competitive interaction between display elements, and the theory shows how stimulus similarities influence the outcome of this competition. One alternative view is that parallel visual processes are limited to local mismatch detection. Search is parallel if the target forms a break in an otherwise homogeneous field, but is serial when absolute stimulus identification is required. It is shown, however, that even colour identification can be parallel, providing targets and nontargets are sufficiently dissimilar. A second alternative view is that search for simple features is parallel whereas search for conjunctions is serial. Conjunction search, however, has a characteristic similarity structure: different kinds of nontarget each share one relevant attribute with the target, but none with one another. When this structure is mimicked in search for colour patches, correspondingly poor performance is obtained.

On the minimum audible angle--a decision theory approach.

The minimum audible angle (MAA) technique is a well-known psychoacoustical paradigm often used in the study of localization of sound. A difficulty with this paradigm, however, is that, in terms of decision theory, it is subject to two quite different interpretations. Although it is normally regarded as involving a discrimination task, the present work suggests that it is more likely to be an absolute identification task. Because of this difference in interpretation, it appears that previous work has overestimated the ability of listeners to localize sources of sound.

'Where' and 'what' in visual search.

A line segment target can be detected among distractors of a different orientation by a fast 'preattentive' process. One view is that this depends on detection of a 'feature gradient', which enables subjects to locate where the target is without necessarily identifying what it is. An alternative view is that a target can be identified as distinctive in a particular 'feature map' without subjects knowing where it is in that map. Experiments are reported in which briefly exposed arrays of line segments were followed by a pattern mask, and the threshold stimulus-mask interval determined for three tasks: 'what'--subjects reported whether the target was vertical or horizontal among oblique distractors; 'coarse where'--subjects reported whether the target was in the upper or lower half of the array; 'fine where'--subjects reported whether or not the target was in a set of four particular array positions. The threshold interval was significantly lower for the 'coarse where' than for the 'what' task, indicating that, even though localization in this task depends on the target's orientation difference, this localization is possible without absolute identification of target orientation. However, for the 'fine where' task, intervals as long as or longer than those for the 'what' task were required. It appears either that different localization processes work at different levels of resolution, or that a single localization process, independent of identification, can increase its resolution at the expense of processing speed. These possibilities are discussed in terms of distinct neural representations of the visual field and fixed or variable localization processes acting upon them.

A limited-capacity response process in absolute identification.

In two absolute identification experiments, the dependency of the current response, Rn, on the immediately preceding stimulus, Sn-1, and response, Rn-1, was measured by means of multivariate information transmission (see McGill, 1954). In these experiments, the amount of stimulus information available to subjects, measured as the amount of information transmission from a current stimulus, Sn, to Rn, was manipulated. The magnitude of the dependency of Rn on Sn-1 and Rn-1 was inversely proportional to that of information transmission from Sn to Rn, supporting the argument of Ward and Lockhead (1971) that the less stimulus information the subjects get, the more their responses will be likely to depend on previous stimuli and responses. Interestingly, the sum of information transmission from Sn, Sn-1, and Rn-1 to Rn was always about 2.5 bits, without respect to the variance of each term. This result could have arisen from the operation of a limited-capacity response process.

Absolute frequency identification and stabilisation of DFB lasers in 1.5 μm region

Using a scanning Fabry-Perot interferometer and a reference laser locked to an ammonia absorption line, the wavelengths of four 1.5 μm DFB lasers were simultaneously identified with an absolute accuracy of ± 0.002 nm (±250 MHz). The lasers were then stabilised to arbitrarily specified wavelengths with the same accuracy.

Major “Enamelin” Protein in Enamel of Developing Bovine Teeth is Albumin

The major, non-amelogenin protein component (enamelin) present in EDTA or EDTA-GU HCl extracts of developing bovine enamel has a molecular weight of approximately 67 kD and an amino acid composition rich in asp, glu, ala, leuc and lys. Elution of this component from 1.5 and 3.0 mm thick strips SDS-PAGE gels and subsequent analyses show that the component selectively reacts with polyclonal antibody to bovine serum albumin. Absolute identification of this major enamelin component as serum albumin is established by an amino acid sequence of the first 40 N-Terminal amino acids which was found to be identical to bovine serum albumin. In addition to albumin, alpha-2 HS glycoprotein was also identified in the same extracts by Western blotting against a monospecific polyclonal antibody against human alpha-2 HS glycoprotein.

Absolute and relative pitch identification by absolute pitch possessors

The identification of absolute and relative pitch was investigated with absolute pitch (AP) possessors and nonpossessors as subjects. In an AP experiment, isolated tones of different pitches (including microtonal pitches) were presented as stimuli in random order. Subjects were requested to select and press a key corresponding to each stimulus on a musical keyboard. While the responses of non‐AP subjects were widely dispersed, the AP subjects could categorize the presented tones quite consistently. Specifically, they identified white‐key notes more correctly and rapidly than black‐key notes. In a relative pitch experiment, isolated melodic ascending intervals from the minor second to the octave were randomly presented. The lower tone of an interval was either middle C, F♯, or a 50‐cent higher D♯. Then non‐AP subjects performed more or less successfully and almost equally regardless of transposition. On the other hand, the AP subjects made more errors and gave longer response times when the lower tone of an interval was F♯ or the higher D♯ than when it was C. These results suggest that at least some AP possessors have difficulty in dealing with musical intervals and tend to rely upon their AP sense in the face of a relative pitch task.

Synthetic vowel studies on cochlear implant patients.

Speech perception studies were conducted on three cochlear implant patients to investigate the relative merits of six speech processing schemes for presenting speech information to these patients. Electrical stimuli, described in this article as synthetic vowels, were constructed using tabulated data of formant frequencies of natural vowels. The six schemes differed in the number of formant frequencies encoded on the electrical signal dimension of electrode position, and/or in the range of electrode position used for encoding each formant frequency. Eleven synthetic vowels (i, I, E, ae, a, c, U, u, v, E, D) were used and were presented in a single-interval procedure for absolute identification. Single-formant vowels were used in two of the six schemes, two-formant vowels in three schemes, and three-formant vowels in the remaining scheme. The confusion matrices were subjected to conditional information transmission analysis on the basis of previous psychophysiological findings. Comparisons among the schemes in terms of the analyzed results showed that training, experience, and adaptability to new speech processing schemes were major factors influencing the identification of synthetic vowels. For vowels containing more than one formant, the information about each formant affected the perception of the other formants. In addition, there appeared to be differences between the perceptual processes for vowels containing more than one formant and the processes for single-formant vowels. Taking into consideration the effects of training, experience, and adaptability, the three-formant speech processing scheme appeared, on the basis of perceptual performance comparisons among the six schemes, to be the logical choice for implementation in speech processors for cochlear implant patients.

S-4-chlorotryptophan: Its synthesis via resolution, determination of the absolute stereochemistry and identification in the crude seed protein of the pea, pisum sativum

Racemic 4-Chlorotryptophan was synthesized by an efficient route and the racemate was resolved by a chiral column into each enantiomer, the absolute stereochemistry of each was determined by reductive dehalogenation to the known chiral tryptophans. S-4-chlorotryptophan was then identified in the crude seed protein of Pisum sativum.

Negative chemical ionization mass spectrometry of phenols, and their determination in acidic fractions of coal‐derived liquids

AbstractThe use of Freon–113 as a reagent gas in the negative chemical ionization mass spectra of phenolic compounds is reported. The dominant mode of ionization is by chloride attachment producing [M + Cl]− ions. The application of this technique to the highly phenolic acid fractions from coal‐derived liquids gives spectra containing essentially only [M + Cl]− peaks. This allows relative molecular mass profiles to be obtained. The resultant complex spectra are simplified by dividing them into their homologous series components. Since this is a low‐resolution technique, absolute component identification cannot be made; but in conjunction with other analytical methods, tentative identification of series of molecules is possible. Thus, the alkyl‐substituent ranges of the different homologous series can be assessed. Comparisons between coal‐derived liquids produced under different reaction conditions are made, and composition differences detected.

Ultrastructural aspects of skin after cryoultramicrotomy

Skin was studied by transmission electron microscopy after cryoultramicrotomy. Global and ultrastructural preservation of the tissue was obtained with satisfactory cohesion between dermis and epidermis. The main constitutive elements could be observed with good definition allowing easy recognition of all the organelles and therefore absolute identification of each cellular type encountered. The methodology provided a comparable quality to the conventional technique. Ice crystal artefacts were minimized but slight enlargement of intercellular spaces persisted. One of the main advantages of this procedure is that neither lipidic solvents nor embedding resins, are used, so that one can envisage cytochemical studies.

Psychophysical studies relevant to the design of a digital electrotactile speech processor.

Psychophysical tests were carried out to investigate the perception of electrocutaneous stimuli delivered to the digital nerve bundles. The tests provided data for defining the operating range of a tactile aid for patients with profound-to-total hearing loss, as well as the individual differences between subjects and the information that could be transmitted. Monopolar biphasic constant current pulses with variable pulse widths were used. Threshold pulse widths varied widely between subjects and between fingers for the same subject. Thresholds were reasonably stable, but maximum comfortable levels increased with time. Perceived intensity was weakly dependent on pulse rate. Absolute identification of stimuli differing in pulse width gave information transmissions from 1.3-2.1 bits, limited by the dynamic ranges of the stimuli (3-17 dB). Stimuli from electrodes placed on either side of each finger were identified easily by all subjects. Absolute identification of stimuli differing in pulse rate gave information transmissions from 0.5-2.0 bits. Difference limens for pulse rate varied between subjects and were generally poor above 100 pps. On the basis of the results, an electrotactile speech processor is proposed, which codes the speech amplitude as pulse width, the fundamental frequency as pulse rate, and the second formant frequency as electrode position. Variable performances on tasks relying on amplitude and fundamental frequency cues are expected to arise from the intersubject differences in dynamic range and pulse rate discrimination. The psychophysical results for electrotactile stimulation are compared with previously published results for electroauditory stimulation with a multiple-channel cochlear implant.

Spectral Response Patterns of Snow and Ice Surfaces for the Landsat Multispectral Scanner (Abstract)

The high-resolution imagery recorded by systems such as the multi-spectral scanners (MSSs) of the Landsat satellites has revolutionized the study of all types of surface in the polar regions. Visible and near-infra-red imagery has found a wide range of glaciological uses. There is, however, a lack of comparability within and between MSS data which may be a contributary factor to some current problems in interpretation of remotely sensed glaciological data. With the expected continuity of MSS coverage for the forseeable future, it is highly desirable to extend use of the data beyond the basic mapping and feature identification which has made it such a valuable resource. One of the most obvious developments is to investigate characteristics of the reflecting surfaces and to achieve absolute identification of snow and ice surfaces. Although conversion of digital MSS grey tones to radiances enables direct comparison with other sources, automatic identification requires detailed and extensive knowledge of the spectral and reflecting characteristics of surfaces which are to be monitored. This is often best achieved through ground-based observation. In order to provide a base line against which corrected radiances from Landsat MSS data can be compared, a spectrally gated photometer has been used to measure albedo at MSS wave bands in a wide range of conditions. The surfaces monitored in several parts of Norway include sea ice, lake ice, snow, firn and glacier ice, permafrost, and reference surfaces. A range of supporting measurements (including grain-size, surface irregularity, density, level, and free-water content) allows accurate characterization of each surface. This enables identification of spectral-response patterns for each surface category and hence the classification of their reflectances as recorded by the MSS. Examples are given of the application of such classifications to imagery of the polar regions.