Abstract
Sensory neurons are often reported to adjust their coding accuracy to the stimulus statistics. The observed match is not always perfect and the maximal accuracy does not align with the most frequent stimuli. As an alternative to a physiological explanation we show that the match critically depends on the chosen stimulus measurement scale. More generally, we argue that if we measure the stimulus intensity on the scale which is proportional to the perception intensity, an improved adjustment in the coding accuracy is revealed. The unique feature of stimulus units based on the psychophysical scale is that the coding accuracy can be meaningfully compared for different stimuli intensities, unlike in the standard case of a metric scale.
Similar content being viewed by others
Introduction
The efficient coding hypothesis1 states that neuronal responses are adjusted, through evolutionary and adaptive processes, to optimally encode such stimulus statistics that matches the sensory environment2,3,4. The statistics of many natural stimuli differs, over short timescales, from the average global distribution and typically, the limited neural coding range does not cover the wide range of possible stimuli values5,6,7. The efficient coding hypothesis therefore predicts that neurons adapt their coding properties to the local stimulus distribution8. In particular, the coding accuracy should increase near the most commonly occurring stimuli in order to minimize the overall decoding error and to maintain the efficient representation of the environment. Such situation is reported in the auditory coding of the sound intensity5,6,9,10, of the interaural level differences11 and time differences12, but also in the neural coding in the primary visual cortex7 and primary somatosensory cortex13. The coding accuracy is commonly evaluated by means of the stimulus-reconstruction paradigm14, that is, by answering how well may the ideal observer determine the stimulus value from the noisy neuronal response. It is assumed that the inverse of the Fisher information approximates the minimal mean squared error9,15,16,17,18,19,20,21. Higher Fisher information reflects higher coding accuracy so that a more precise representation of stimuli is possible.
The goal of this short paper is to point out to a potentially problematic aspect of aligning the maximal coding accuracy with the most frequent stimuli. Our reasoning follows from the fact that the stimulus values are quantified by choosing some convenient, but otherwise arbitrary, system of measurement units. For example, the sound intensity is typically expressed as the sound pressure level in decibels (dB SPL). The same stimulus intensity can be equivalently expressed in terms of the effective pressure in Pascals (Pa)22. The seemingly arbitrary choice of stimulus scale, however, has a non-trivial and significant impact on the coding precision. As demonstrated by Kostal and Lansky23, a non-linear relationship between different units (such as between the sound pressure and the sound level) may affect the position of maximal coding accuracy. Here we demonstrate the paradoxical consequences of the stimulus scale change on both the coding accuracy and the known stimulus distribution simultaneously. We show that the match between high coding accuracy regions and most frequent stimuli regions depends on the choice of the measurement unit. Second, we attempt to resolve this problem by arguing that the natural system for stimulus quantification is given by the scale linearly proportional to the perception intensity24,25. As an illustration, we employ the classical Riesz’s psychophysical scale for the sound intensity26 to reveal the expected coding accuracy adaptation even for low pressure levels in the experimental data of Watkins and Barbour9.
Methods
The psychophysical scale describes the perceptual intensity, ψ, as a function of the stimulus intensity I24,25. The empirical finding known as Weber’s law27,28 states that the smallest noticeable increment in perception, Δψ, remains constant if the relative stimulus increment (also known as Weber’s factor) is also constant,
As suggested later by Fechner29, Weber’s law effectively sets the scale for the perceived stimulus intensity since ΔI/I is proportional to Δψ. By integrating Eq. (1) we obtain the well known Fechner’s law, stating that the perceived intensity varies as28.
where k is a proportionality factor and I0 some reference value.
Subsequent investigations found that Eq. (1) holds neither generally nor exactly24,30 across different sensory modalities. In particular, Weber’s factor for human sound intensity discrimination was found to satisfy26,
Here I is the basal sound intensity in W/m2, ΔI is the minimum perceptible difference, S∞ is the value ΔI/I approaches at high intensities, S0 > S∞ is the value of ΔI/I at the threshold of hearing and r is a parameter, approximately r = 1/2. Weber’s factor in Eq. (3) is no longer constant, but decreases rapidly to a plateau with increasing intensity I. Since the sound intensity and the sound pressure are related by the acoustic impedance N.s.m−3 as22
the following differential equation follows from Eq. (3)
provided that the derivative dψ/dp is a good approximation to Δψ/Δp. The solution to Eq. (5) is
which is proportional to log(c + p). Setting the values26 S∞ = 0.2, S0 = 1 and arbitrary p0 = 20 μPa (the actual reference level is of little importance here) we obtain
where so that p = 0 Pa yields ψ = 0 for convenience. Eqution (7) determines Riesz’s scale (in arbitrary units) of sound pressure values, correcting the inadequate Fechner’s law in Eq. (2) for small sound intensities (pressures). In other words, the value of ψ can be used to measure the sound intensity on the scale which is linearly related to the perception intensity. The standard sound pressure level scale L (given in dB SPL) is essentially equivalent to Fechner’s law, since due to Eq. (4) it holds22
The Eqs (7) and (8) are approximately proportional to each other for sufficiently high pressure levels (Fig. 1).
Results
The coding accuracy as a function of the stimulus intensity is significantly affected by the choice of the measurement scale23. The question is whether the coding accuracy adaptation to the stimulus distribution (as observed, e.g., in the experiments5,6,9), is preserved under the change of stimulus units.
The Fisher information IF(P) as a function of the sound pressure and the Fisher information IF(ψ) for the sound intensity measured on Riesz’s scale from Eq. (7), are related as
Similarly, one may additionally use Eq. (8) to relate, e.g., IF(ψ) and IF(L). The transformation rule in Eq. (9) is well known and can be derived directly from the definition of the Fisher information by using the chain rule for derivatives15. Similarly, the stimulus probability density function f(·) satisfies31
Therefore it follows that any visual alignment between the values of the coding accuracy and the stimulus distribution depends crucially on the choice of units. Even though the square root of the Fisher information transforms analogously to Eq. (10), the potential match between the peaks of and f is not preserved under the stimulus scale change because and f are often related non-linearly. In fact, it can be shown rigorously that also the global match between the profiles of and f is not preserved under the stimulus scale change32, unless the stimulus probability density function is exactly proportional to the square root of the Fisher information (known as the Jeffreys prior33).
We illustrate how a specific choice of the stimulus units improves the experimentally observed adaptation of the coding accuracy to the stimulus distribution. We argue that the stimulus scale proportional to the actual perception intensity (the psychophysical scale) is the natural reference frame under which the coding accuracy should be evaluated.
Neurons in the auditory system are reported to adjust their rate-intensity functions in order to improve coding accuracy over high-probability stimulus regions5,9,10. The match is not perfect for low sound intensities and a positive bias of maximal coding accuracy towards higher intensities is reported. For example, in the experiment of Watkins and Barbour9, the sound level distribution was set to be uniform over -15 dB SPL to 105 dB SPL, with an added 20 dB-wide plateau of high-probability stimulus region (Fig. 2A, filled area). At every 100 ms during the experiment a new sample was drawn from the distribution to set the amplitude of a pure tone, with its frequency matching the characteristic frequency of the studied neuron (primary auditory cortex of marmoset monkey). The dynamic rate-level function was measured and the coding accuracy (the Fisher information) was calculated (Fig. 2A, solid line), see Watkins and Barbour9 for more details. The coding accuracy adaptation was determined for four different positions of the plateau, centered at 5, 25, 45 and 65 dB SPL respectively (Fig. 2A–D). The peak coding accuracy does not align with frequently occurring low sound intensities (Fig. 2A).
The same experimental data evaluated on Riesz’s scale yield far better alignment of coding accuracy with stimuli statistics, especially for low intensities (Fig. 2E). On the other hand, the existing match for high levels (Fig. 2C,D) is preserved (Fig. 2G,H) due to the similarity of both scales for high intensities (Fig. 1). The match between the stimulus statistics and the coding accuracy can be quantified by the ratio of the maximal Fisher information in the high-probability region to the global maximum of the Fisher information. For the four examined cases of the plateau centered at (5, 25, 45, 65) dB SPL we obtain the following values of this ratio: (0.46, 0.96, 1, 1) on the pressure level scale and (1, 1, 1, 1) on Riesz’s scale. Note that the non-uniform shape of the high-probability regions results from the transformation rule for the probability density function.
Discussion
The described adaptation of neural coding precision to the local stimulus distribution results in a more efficient representation of the environment5,9. However, the investigation of coding strategy should also take the actual perception intensity into the account11. In all likelihood, coding precision expressed by employing the psychophysical scale (such as Riesz’s scale) is more useful and natural than when evaluated in the standard metric system (such as dB SPL). The reasoning is that Riesz’s scale is linear in the true perception intensity as described in the Methods section. Consequently, the smallest noticeable increment in perception Δψ is proportional to a fixed value on Riesz’s scale and this value is constant for all stimulus intensities. Hence the unique feature of a stimulus unit based on the psychophysical scale is that the coding accuracy evaluated in such units can be meaningfully compared for different stimuli intensities – unlike the metric scale case. Even if coding precision varies with the stimulus intensity on the metric scale substantially, these variations might be immaterial provided that the actual difference in sensation falls within the smallest noticeable increment.
Note that if Weber’s law was valid for the sound intensity perception, the dB SPL scale would correspond to the exact psychophysical scale. From this point of view the shifted-logarithm in Eq. (7) represents a seemingly negligible correction. We have shown, however, that the difference between Riesz’s and sound pressure level scales affects the coding accuracy adjustment substantially. Ries’z correction, ψ ∝ log(const. + I), to the purely logarithmic Fechner’s law in Eq. (2) has a long history and is more fundamental and general, going beyond the case of the sound intensity perception. See34,35,36 for a detailed account. For example, the equation for Riesz’s scale as a function of the sound pressure in Eq. (7) is formally identical to the psychophysical mel scale37, which describes the perception intensity for sound frequency. Both the mel and Riesz’s psychophysical scales thus follow Weber’s law for large stimuli values only.
Our message, however, reaches beyond the topic of psychophysical scales and auditory neuroscience. We argue that the coding accuracy is generally a relative quantity, with respect to chosen units, a fact whose consequences seem to have been neglected in the experimental research. The expected matching of stimulus statistics with the coding accuracy is thus not absolute and does not hold in different unit systems. The coding accuracy reflects the spread of estimated stimulus values, which is affected not only by the stochastic nature of neural responses but also by the arbitrarily chosen unit system for stimulus quantification. In addition, we believe that coding accuracy should generally be evaluated on the scale which is linearly proportional to the internal representation of the stimulus, i.e., proportional to the actual perception intensity.
Finally, it is worth noting that different ways to asses the neural coding efficiency were developed over the decades. A substantial part of the literature employs Shannon’s measure of information38 to determine the absolute scale on neuronal performance39. By treating the neuronal system as an information channel and by maximizing the mutual information between stimuli and responses, one obtains the optimal stimulus distribution, as for example in40,41,42,43,44. Under the assumption of vanishing response variability, the optimal stimulus distribution is proportional to 45,46,47,48,49,50,51, which is known to be invariant under coordinate transformations33. Heuristically, one may view this result as providing support for the idea of high coding precision matching high probability stimulus regions40. Unlike the local method of Fisher information described in this paper, however, the information theory determines the complete (global) form of the stimulus distribution.
Additional Information
How to cite this article: Kostal, L. and Lansky, P. Coding accuracy on the psychophysical scale. Sci. Rep. 6, 23810; doi: 10.1038/srep23810 (2016).
References
Barlow, H. B. Possible principles underlying the transformation of sensory messages. In Rosenblith, W. (ed.) Sensory Communication, 217–234 (MIT Press, Cambridge, 1961).
Simoncelli, E. P. & Olshausen, B. A. Natural image statistics and neural representation. Annu. Rev. Neurosci. 24, 1193–1216 (2001).
Lewicki, M. S. Efficient coding of natural sounds. Nat. Neurosci. 5, 356–363 (2002).
Kostal, L., Lansky, P. & Rospars, J.-P. Efficient olfactory coding in the pheromone receptor neuron of a moth. PLoS Comput. Biol. 4, e1000053 (2008).
Dean, I., Harper, N. S. & McAlpine, D. Neural population coding of sound level adapts to stimulus statistics. Nat. Neurosci. 8, 1684–1689 (2005).
Wen, B., Wang, G. I., Dean, I. & Delgutte, B. Dynamic range adaptation to sound level statistics in the auditory nerve. J. Neurosci. 29, 13797–13808 (2009).
Durant, S., Clifford, C. W. G., Crowder, N. A., Price, N. S. C. & Ibbotson, M. R. Characterizing contrast adaptation in a population of cat primary visual cortical neurons using Fisher information. J. Opt. Soc. Am. A 24, 1529–1537 (2007).
Wark, B., Lundstrom, B. N. & Fairhall, A. Sensory adaptation. Curr. Opin. Neurobiol. 17, 423–429 (2007).
Watkins, P. V. & Barbour, D. L. Specialized neuronal adaptation for preserving input sensitivity. Nat. Neurosci. 11, 1259–1261 (2008).
Watkins, P. V. & Barbour, D. L. Level-tuned neurons in primary auditory cortex adapt differently to loud versus soft sounds. Cereb. Cortex 21, 178–190 (2011).
Dahmen, J. C., Keating, P., Nodal, F. R., Schulz, A. L. & King, A. J. Adaptation to stimulus statistics in the perception and neural representation of auditory space. Neuron 66, 937–948 (2010).
Maier, J. K. et al. Adaptive coding is constrained to midline locations in a spatial listening task. J. Neurophysiol. 108, 1856–1868 (2012).
Garcia-Lazaro, J. A., Ho, S. S. M., Nair, A. & Schnupp, J. W. H. Shifting and scaling adaptation to dynamic stimuli in somatosensory cortex. Eur. J. Neurosci. 26, 2359–2368 (2007).
Berens, P., Ecker, A. S., Gerwinn, S., Tolias, A. S. & Bethge, M. Reassessing optimal neural population codes with neurometric functions. Proc. Natl. Acad. Sci. USA 108, 4423–4428 (2011).
Lehmann, E. L. & Casella, G. Theory of point estimation, Ch. 2, 115–117 (Springer Verlag, New York, 1998).
Seung, H. S. & Sompolinsky, H. Simple models for reading neuronal population codes. Proc. Natl. Acad. Sci. USA 90, 749–753 (1993).
Dayan, P. & Abbott, L. F. The Effect of Correlated Variability on the Accuracy of a Population Code. Neural Comput. 11, 91–101 (1999).
Seriès, P., Latham, P. E. & Pouget, A. Tuning curve sharpening for orientation selectivity: coding efficiency and the impact of correlations. Nat. Neurosci. 7, 1129–1135 (2004).
Zhang, K., Ginzburg, I., McNaughton, B. L. & Sejnowski, T. J. Interpreting neuronal population activity by reconstruction: unified framework with application to hippocampal place cells. J. Neurophysiol. 79, 1017–1044 (1998).
Harper, N. S. & McAlpine, D. Optimal neural population coding of an auditory spatial cue. Nature 430, 682–686 (2004).
Sreenivasan, S. & Fiete, I. Grid cells generate an analog error-correcting code for singularly precise neural computation. Nat. Neurosci. 14, 1330–1337 (2011).
Raichel, D. R. The Science and Applications of Acoustics (Springer, New York, 2006).
Kostal, L. & Lansky, P. Coding accuracy is not fully determined by the neuronal model. Neural Comput. 27, 1051–1057 (2015).
Gescheider, G. A. Psychophysics: The fundamentals (Lawrence Erlbaum Associates, Mahwah, New Jersey, 1997).
Sun, J. Z., Wang, G. I., Goyal, V. K. & Varshney, L. R. A framework for Bayesian optimality of psychophysical laws. J. Math. Psychol. 56, 495–501 (2012).
Riesz, R. R. Differential intensity sensitivity of the ear for pure tones. Phys. Rev. 31, 867–875 (1928).
Weber, E. H. De pulsu, resorptione, auditu et tactu. Annotationes anatomicae et physiologicae (Koehler, Leipzig, 1834).
Dayan, P. & Abbott, L. F. Theoretical Neuroscience: Computational and Mathematical Modeling of Neural Systems (MIT Press, 2001).
Fechner, G. T. Elemente der Psychophysik (Breitkopf und Härtel, Leipzig, 1860).
Masin, S. C. The (Weber’s) law that never was. In Elliot, M. A. et al. (eds) Proceedings of the twenty-fifth annual meeting of the International Society for Psychophysics, 441–446 (International Society for Psychophysics, Galway, Ireland, 2009).
Papoulis, A. Probability, random variables and stochastic processes, Ch. 5, 92–94 (McGraw-Hill, New York, 1991).
Kostal, L. Stimulus reference frame and neural coding precision. J. Math. Psychol. in press (2016).
Jeffreys, H. An invariant form for the prior probability in estimation problems. Proc. Roy. Soc. A 453–461 (1946).
Hecht, S. The visual discrimination of intensity and the Weber-Fechner law. J. Gen. Physiol. 7, 235–267 (1924).
Norwich, K. H. & Wong, W. Unification of psychophysical phenomena: The complete form of Fechner’s law. Percept. Psychophys. 59, 929–940 (1997).
Dzhafarov, E. N. & Colonius, H. The Fechnerian Idea. Am. J. Psychol. 124, 127–140 (2011).
Stevens, S. S., Volkmann, J. & Newman, E. B. A scale for the measurement of the psychological magnitude pitch. J. Acoustic. Soc. Am. 8, 185–190 (1937).
Shannon, C. E. Communication in the presence of noise. Proc. IRE 37, 10–21 (1949).
Rieke, F., de Ruyter van Steveninck, R., Warland, D. & Bialek, W. Spikes: Exploring the Neural Code (MIT Press, Cambridge, 1997).
Laughlin, S. B. A simple coding procedure enhances a neuron’s information capacity. Z. Naturforsch. 36, 910–912 (1981).
de Ruyter van Steveninck, R. R. & Laughlin, S. B. The rate of information transfer at graded-potential synapses. Nature 379, 642–644 (1996).
Ikeda, S. & Manton, J. H. Capacity of a single spiking neuron channel. Neural Comput. 21, 1714–1748 (2009).
Suksompong, P. & Berger, T. Capacity analysis for integrate-and-fire neurons with descending action potential thresholds. IEEE Trans. Inf. Theory 56, 838–851 (2010).
Kostal, L. & Kobayashi, R. Optimal decoding and information transmission in Hodgkin-Huxley neurons under metabolic cost constraints. Biosystems 136, 3–10 (2015).
Bernardo, J. M. Reference posterior distributions for Bayesian inference. J. Roy. Stat. Soc. B 41, 113–147 (1979).
Brunel, N. & Nadal, J.-P. Mutual information, Fisher information and population coding. Neural Comput. 10, 1731–1757 (1998).
McDonnell, M. D. & Stocks, N. G. Maximally informative stimuli and tuning curves for sigmoidal rate-coding neurons and populations. Phys. Rev. Lett. 101, 058103 (2008).
Ganguli, D. & Simoncelli, E. P. Implicit encoding of prior probabilities in optimal neural populations. In Lafferty, J., Williams, C., Shawe-Taylor, J., Zemel, R. S. & Culotta, A. (eds) Advances in Neural Information Processing Systems (NIPS), vol. 23, 658–666 (MIT Press, Cambridge, Massachusetts, 2010).
Yarrow, S., Challis, E. & Seriès, P. Fisher and Shannon information in finite neural populations. Neural Comput. 24, 1740–1780 (2012).
Kostal, L., Lansky, P. & McDonnell, M. D. Metabolic cost of neuronal information in an empirical stimulus-response model. Biol. Cybern. 107, 355–365 (2013).
Wei, X. & Stocker, A. A. A Bayesian observer model constrained by efficient coding can explain ‘anti-bayesian’ percepts. Nat. Neurosci. 18, 1509–1517 (2015).
Acknowledgements
We are grateful to Dennis Barbour for helpful discussion and to Ehtibar Dzhafarov for pointing us to the relevant literature. We also thank two anonymous referees for their comments, which helped to improve the manuscript significantly. This work was supported by the Institute of Physiology RVO:67985823 and by the Czech Science Foundation project GA15-08066S.
Author information
Authors and Affiliations
Contributions
L.K. and P.L. prepared the manuscript. All authors reviewed the manuscript.
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
Kostal, L., Lansky, P. Coding accuracy on the psychophysical scale. Sci Rep 6, 23810 (2016). https://doi.org/10.1038/srep23810
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/srep23810
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.