Near-infrared spectroscopy of image clarity perception in the human brain

The perception of blur in humans is intrinsic to our visual system, and dioptric power can improve clarity in many cases. This was evaluated experimentally to establish the best correction with dioptric power shifts. We used Near Infrared Spectroscopy (NIRS) to measure Oxy-, Deoxy- and Total-hemoglobin concentration changes in the brain while viewing images and reading a Snellen chart. Participants were tested with their usual correction (no diopter power shift (0 D)), with a 0.25 diopter power shift (0.25 D), and with a 0.5 diopter power shift (0.5 D). The concept of Approximate Entropy (AE) was applied to quantify the regularity of these hemoglobin time series of finite length. AE computations are based on the likelihood that similar templates in a time series remain similar on the next incremental comparison, so that time series with large AE have high irregular fluctuation. We found that the dioptric power shift eliciting the highest AE indicates the clearest visual condition for subjects. This technique may impact the current way in which ophthalmic lenses are prescribed.