Multimodal temporal perception deficits in a patient with left spatial neglect

Abstract

We examined multisecond time estimation (up to 60 s) for visual and auditory events in a patient with left spatial neglect (RR), who grossly underestimated all durations in all tasks. In contrast, healthy controls and a patient with left hemisphere damage (HW) demonstrated accurate estimates of the same durations. These findings add to a growing body of literature suggesting that neglect cannot be understood simply in terms of a bias in orienting attention to one side of space. In addition, these data suggest that the right hemisphere parietal cortex may be important for the perception of time across multiple modalities.

Keywords:

• Temporal estimation
• Parietal cortex
• Neglect
• Nonspatial deficits
• Spatial attention

We are grateful to RR and HW for their time and effort through several testing sessions. This work was supported by Natural Sciences and Engineering Research Council of Canada (NSERC) Canada Research Chair and Discovery grants to JD, Heart and Stroke Foundation of Ontario Grant-in-aid to JD, NSERC Undergraduate and CGS-M Awards to CM, Ontario Graduate Scholarship in Science and Technology (OGSST) to CM, and an NSERC PGS-D award to MH.

Notes

¹Note that of the patients with right parietal damage in the Harrington et al. (1998) paper, only 2 of 18 cases were reported as having neglect.

²Healthy controls did not appear to be making categorical judgements of the temporal intervals to be estimated. In fact, they demonstrated mean underestimations at the 60 s intervals and substantial variability. Had they been making categorical judgements they may still have tended to underestimate (or overestimate) durations but would have shown far less variability. No single control (or patient) demonstrated behavior indicative of categorical judgements and in fact the variance scaled with the duration to be estimated (i.e., Weber's law) as would be expected. As one example, for the 60-s duration of the nonverbal task, the healthy participants' estimates ranged from 25–160 s. Similar variability was seen across all tasks.

³While HW's estimate is outside the 95% CI on the 60 s interval of the nonverbal TE task, it was not significantly different when compared using Crawford's modified independent samples t-test.

⁴It should be noted that with a small sample size such as the one used here and the relatively accurate temporal estimates of both the control group and patient HW, correlations between signed error values may not necessarily reach significance. For example, a relationship between signed error values will cancel out when a participant slightly underestimates durations on one task and slightly overestimates in another task.