Investigating the limits of sensory discrimination with applications to decision-making in hearing

Sai Vemula (1), Kenneth Norwich (1), Willy Wong (2)

1. Institute of Biomaterials and Biomedical Engineering

2. Edward S. Rogers Sr. Department of Electrical and Computer Engineering


The study of intensity discrimination is concerned with how signals from the environment are coded at different levels of sensory systems. In the human senses, evaluations of intensity discrimination, by means of sensory thresholds, are valuable tools in assessing the health and performance of sensory systems. The sensory differential threshold, or the just-noticeable-difference (JND), is an important psychophysical metric used to explore sensory perception. 

In audition, a differential threshold refers to an observer’s ability to either detect a sound from a background sound signal (a Type II experiment), or to discriminate a sound signal from another (a Type I experiment). These thresholds are expressed by the Weber fraction ∆I/I, relating the minimum stimulus change perceived to the background sound signal. However, not all reported measures of the differential thresholds are equal; for the same sensory signals, studies conducted with the contrasting experimental methods yield differences in the Weber fraction curves and psychometric functions. 


This study aims to address these objectives through both experimental and modelling approaches. The experimental approach consists of psychoacoustic Type-I experiments in healthy-hearing participants, which will probe the participants’ thresholds over a range of intensities and pulse durations. Results from pilot experiments suggest that there is no significant difference between the Weber fraction curves acquired at two different pulse durations (200ms and 500ms). Psychometric functions were also found to be invariant in slope as a function of intensity. 

The modelling approach focuses on the entropy theory of sensation, which explicitly relates the physical sensory signal at periphery to the resulting response in the sensory system (perception). While the entropy model already provides a theoretical derivation of the differential threshold, key improvements are needed before it can be fully applied to the Type-I paradigm. 


The characterization of sensory JNDs advances our understanding of sensory systems and guides the development of sensory prostheses and assistive devices. 

Neural SystemsiARC