Back

Uncertainty revealed by delayed responses during olfactory matching

Bhowmik, R.; Pardasani, M.; Mahajan, S.; Bhattacharjee, A. S.; Konakamchi, S.; Phadnis, S.; Musthafa, T.; McGowan, E.; Srikanth, P.; Marathe, S. D.; Abraham, N. M.

2022-09-13 neuroscience
10.1101/2022.09.11.507462 bioRxiv
Show abstract

Matching of olfactory stimuli involves both sensory and higher cognitive functioning. Different decision processes such as detection and discrimination, along with holding the perceived information are involved during the matching process. Accuracy and decision times, the interdependent readouts, can define the uncertainty involved in matching of sensory stimuli. To probe sensory and cognitive functions involving olfactory system in human subjects, we have developed a novel olfactory matching paradigm using an automated custom-built olfactory-action meter. With precise and consistent odor delivery and real-time data analysis, our system automates the entire process without any intervention by the experimenter, making it suitable as a diagnostic tool for quantifying olfactory and neurocognitive fitness. In around 400 healthy human subjects, with mean detection accuracy of 90%, we observed significantly better olfactory matching performance for simple monomolecular odors, in comparison to complex binary odor mixtures. Odor matching accuracy declined significantly with the increase in odor complexity. Olfactory matching was more rapid when subjects made correct versus incorrect decisions, indicating perceptual certainty. Subjects also took longer matching time for complex odors compared to simple odor stimuli. Thus, olfactory matching that provides a combined readout of sensory and cognitive fitness, establishes a direct link between the performance accuracy and the certainty of decisions.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.