Table 1.
Demographic data for patients with cerebellar ataxia (n = 11).
Fig 1.
Examples of the 3 saccadic reading tasks.
(A) 40 regularly spaced single-digit numbers read aloud (2nd page) (B) 40 irregularly spaced single-digit numbers read aloud (2nd page, King-Devick test) (C) 74 irregularly spaced words in sentences with semantic context read silently (natural reading, 3rd page).
Fig 2.
Reading speed in cerebellar ataxia patients and age-matched controls during the three saccadic reading tasks.
(A) Box-whisker plots of reading speed (numbers/s or words/s) in age-matched controls (solid) and patients with cerebellar ataxia (stripes). There was significantly slower reading speed in ataxia patients during regularly spaced (p = 0.02) and irregularly spaced number reading (p = 0.02). (B) Bar graph of difference in reading speed (controls—ataxia). (C) Bar graph of relative difference (%) in reading speed in ataxia patients compared with controls. White = regularly spaced number reading aloud, grey = irregularly spaced number reading aloud (King-Devick test), black = irregularly spaced word reading silently.
Fig 3.
Ataxia patients made more saccades and fixations and had greater scan path lengths while reading 120 regularly spaced single-digit numbers aloud.
Box-whisker plots showing that ataxia patients (grey) had increased saccade time (p = 0.02), fixation time (p > 0.05), saccade count (p = 0.01), fixation count (p = 0.004), and scan path length (p = 0.01) compared with control subjects (white).
Fig 4.
Saccades were irregular and abnormal in ataxia patients compared with controls while reading 120 regularly spaced single-digit numbers aloud.
(A) Main sequence relationship scatter plots of saccade amplitude (°) vs. duration (ms) and (B) saccade amplitude (°) vs. peak velocity (°/s) for controls (grey) and ataxia patients (black). In controls, there was a clear cluster of larger amplitude line-changing saccades (black arrows) that was less prominent in ataxia patients. A linear regression model (saccade amplitude vs. duration) and one phase exponential equation (saccade amplitude vs. peak velocity) was used to fit data from saccades by controls with 5% and 95% prediction bounds dotted. (C) Relative frequency power analysis of saccade amplitude (°), peak velocity (°/s), and duration (ms) for controls and ataxia patients showed that compared with that of controls, the peak relative frequency of saccade amplitude and peak velocity was smaller and slower, respectively, in ataxia patients (black arrowheads).
Fig 5.
Ataxia patients made more fixations with shorter durations and greater dispersion while reading 120 regularly spaced single-digit numbers aloud.
Relative frequency power analysis of fixation duration (ms) and x- and y-dispersion (px) for controls (grey line) and ataxia patients (black line). Compared with controls, ataxia patients had a relatively greater peak at shorter fixation duration (left, arrow), greater x-dispersion (middle, arrow), and no difference in y-dispersion (right) during regularly spaced number reading.
Fig 6.
Irregular scan paths and staircase gaze patterns in ataxia patients during regularly spaced single-digit number reading.
(A) Scan paths during regularly spaced number reading (2nd page) of a 35-year-old male control, 67-year-old female control, ataxia patient #1 (spinocerebellar ataxia type 6), and ataxia patient #2 (multiple system atrophy). Circles denote fixations, size of the circles corresponds with duration, and lines indicate saccade paths between fixations. (B) Line graph of horizontal gaze position during regularly spaced number reading (2nd page) of a 35-year-old male control (left) and ataxia patients #2–4 (see Table 1). Vertical lines indicate saccades to the right, and horizontal lines indicate fixation. Corresponding segments of scan paths are shown above. Scale: y-axis long ticks = 400 px, x-axis long ticks = 0.5 sec.
Fig 7.
Irregular scan paths and staircase gaze patterns in ataxia patients during silent word reading.
(A) Scan paths during word reading (3rd page) of a 35-year-old male control, ataxia patient #2 (multiple systems atrophy) and ataxia patient #4 (cerebellar hemangioblastoma). Circles denote fixations, the size of the circles corresponds with duration, and lines indicate saccades between fixations. (B) Line graphs of horizontal gaze position during silent word reading revealed similarly abnormal hypometric or hypermetric saccades in ataxia patients during word reading compared with number reading (Fig 6).