Table 3 Key data extracted from true hemianopia studies

Alberti et al (2017) [43]

Cohort study

n = 12

HH (6 right, 6 left)

Stroke (n = 8)

Tumour (n = 3) Trauma (n = 1)

Adults

N/A

Participants completed two sessions in a driving simulator pressing the horn when they detected a pedestrian

Eye movements

Gaze and pedestrian eccentricity on the BHF.

Gaze and pedestrian eccentricity on the SHF

On the blind side, the first scan to reach the pedestrian occurred later, on average, than on the seeing side (medians of 1.25 and 0.75 s, respectively).

On the blind side only 40% of scans to reach the pedestrian were made within 1 s of the pedestrian appearing compared with approximately 70% on the seeing side.

Median scan magnitudes did not differ between the two conditions (approaching: 14° (IQR 9°–15°); stationary: 13° (IQR 9°–20°); p = 0.43).

Bahnemann et al (2015) [44]

Case-control study

n = 41

HH

Ischemic infarction (n = 41)

Adults

N/A

Participants were tested in a fixed-base driving simulator and 3 driving scenarios were constructed.

Eye and head movements

Amplitude and the peak velocity of saccades, distribution of fixations, number and the duration of saccades, duration of fixations

The number, amplitude, duration and the peak velocity of head movements

The low performance (LP) group showed a significantly smaller amplitude than either the control group (p = 0.031) or the High performance (HP) group (p = 0.030). The LP group also showed a significantly lower velocity than the HP group (p = 0.031), and the control group (p = 0.055).

The LP group showed a significantly more narrow spread than either the control (p = 0.015) or HP group (p = 0.003), with the HP group showing the widest spread.

The HP group displayed the highest percentage of fixations in the BHF (p = 0.002).

No statistically significant differences were found concerning the number (p = 0.304) and the duration of saccades (p = 0.221).

No difference regarding the duration of fixations (p = 0.567) between the healthy controls and the participant subgroups.

No statistically significant differences were found regarding the number (p = 0.877), the amplitude (p = 0.696), the duration (p = 0.495), and the peak velocity (p = 0.350) of head movements between the two participant subgroups and the healthy controls.

Bowers et al (2014) [45]

Case-control study

n = 26

HH (n = 14, 8 left, 6 right)

Normally sighted (n = 12)

Stroke (n = 11)

Other (n = 2)

Adults

N/A

Using a driving simulator, the effects of HH on head scanning behaviours at intersections were examined

Head movement

Total number of scans, proportions of leftward scans, the magnitude of the leftward and rightward scans

No significant differences among the vision groups in the total number of scans (p = 0.253).

The LHH group had a significantly higher proportion of leftward scans (76%) than normal controls and RHH groups (59% and 48%, respectively).

The proportion of leftward scans was highest when there was an incoming road on the left only (74%) and lowest when there was no incoming road on that side (41%, p < 0.001).

The RHH group made a significantly lower proportion of leftward head scans (i.e. a higher proportion of rightward scans) than normal controls and LHH drivers (p = 0.008 and p < 0.001, respectively) when there was no incoming road on their blind right side.

LHH group made a significantly higher proportion of head scans to their blind left side than normal controls and RHH drivers (p < 0.001) when there was no incoming road on that side.

For all groups, the magnitude of the leftward and rightward scans increased as distance to the intersection decreased (p < 0.001).

Head scan magnitudes for the two HH groups were smaller than those of the control group (p < 0.001). For the HH groups, blind side scans were not larger than seeing side scans.

Cazzoli et al (2016) [46]

Case-control study

n = 32

Patient group (n = 24, 8 neglect, 6 quadrantanopia, 10 HH)

Healthy controls (n = 8)

Ischemia (n = 17)

Haemorrhage (n = 7)

Healthy subjects (n = 8)

Adults

N/A

Participants freely explored a traffic scene without (static condition) and with (dynamic condition) naturalistic motion, i.e. cars moving from the right or left.

Eye movements

Mean gaze position during early attentional orienting.

Cumulative fixation duration (CFD) spatial distribution (the sum of the duration of all visual fixations).

The mean gaze position during early attentional orienting was significantly rightward-deviated in participants with neglect and VFD in comparison to all other groups (i.e. participants with VFD only, participants with right-hemispheric lesions without neglect or VFD, and healthy subjects).

Participants with neglect and VFD showed significantly lower percentage CFD values than healthy subjects.

Participants with VFD showed significantly higher percentage CFD values than all other groups.

Fourtassi et al (2016) [47]

Cohort study

n = 19

9 patients (3 right HH, 2 left HH, 2 quadrantanopia, 2 no deficit). Two had HH + neglect; 2 had neglect but no VFD10 healthy subjects

Stroke (n = 9)

Healthy subjects (n = 10)

Adults

N/A

Gaze positions were recorded during memory recall of French towns in an imagery task, a non-imagery task (verbal fluency), and a visually guided task.

Eye movements

Spatial consistency between gaze positions, spatial distribution of the gaze throughout the task, saccades’ directions and amplitudes.

Gaze was constantly shifted with respect to their body midline, contralesionally for all HH participants without neglect and ipsilesionally for the two HH participants with neglect.

Participants performed a similar number of saccades of similar magnitude in both directions as did control subjects (rightward and leftward).

No differences between the number or the mean magnitudes of rightward and leftward saccades, either in imagery or in non-imagery tasks (p > 0.05).

Gbadamosi et al (2001) [48]

Case-control study

n = 34

HH (n = 14, 7 left, 6 right, 1 bitemporal deficit)

Healthy controls (n = 20)

Stroke, trauma, operation

Adults

N/A

Each subject was tested during a standardised routine that consisted of one stimulus presentation phase and three subsequent imagery phases. Six different stimuli that were all bordered by a reference frame were used.

Eye movements

Number of fixations, fixation duration, saccade amplitude

The number of fixations was lower in the imagery phases than the viewing scanpaths in both groups.

Fixation duration was significantly higher in the imagery scanpaths than in the viewing scanpaths.

The median saccade amplitudes were lower during the imagery phases in both groups.

Grasso et al (2016) [49]

Before and after study

n = 10

HH

Ischemic (n = 8)

Haemorrhagic (n = 1)

Arteriovenous malformation (n = 1)

Audio-visual training for 4 h daily, over a period of 2 weeks.

Eye movements were recorded during several visual tasks before and after audio-visual training

Eye movements

Exploration time, number of fixations, saccadic speed

Significant improvement in scanning efficiency. Number of fixations reduced: 80.9 to 73.6 to 70.2. Mean saccadic speed faster: 50.45d/s to 64.81d/s to 64.00d/s. Mean exploration time lower: 27.2 s to 23.5 s to 23.2 s.

Hardiess et al (2010) [50]

Case-control study

n = 24

Hemianopia (n = 10, 2 left, 8 right) Quadrantanopia (n = 2)

Healthy controls (n = 12)

Ischemia (n = 10)

Brain surgery (n = 1)

Haemorrhage (n = 1)

Adults

N/A

All participants were tested in two tasks, i.e. a dot counting (DC) task requiring mostly simple visual scanning and a cognitively more demanding comparative visual search (CVS) task.

Eye and head movements

Task performance (response time), scanpaths, gaze performance, head movements

Search time significantly increased for the two participant groups (p < 0.001).

No differences were found regarding the scanpaths between the normal subjects (230° DC, 1050° CVS) and the high performance group (240° DC, 1100° CVS).

Low performance group performed with a highly increased number of fixations (34 DC, 52 CVS) repetition of fixations (9 DC) a higher proportion of fixations towards the impaired visual field (65% DC, 58% CVS) compared to controls (25 DC, 37 CVS / 52% DC, 53% CVS respectively). The scanpath appeared rather unsystematic and time-consuming (340°) compared to controls (230°). Participants also displayed significantly increased fixation duration (290 ms DC, 260 ms CVS) compared to controls (270 ms DC, 240 ms CVS ).

The high performance group showed an organised scanning pattern similar to that of the unimpaired normal subjects.

In the free head comparative visual search task all subjects performed maximum head movements in a range of ±3° and ± 20°.

The average maximum amplitudes were larger for LP participants (14.13 ± 8.0 mean ± SD), while for HP participants they were within a range (8.89 ± 4.78 mean ± SD) similar to the one of normal subjects (9.12 ± 5.0 mean ± SD).

Kasneci et al (2014a) [51]

Case-control study

n = 40

HH (n = 10, 4 right, 6 left)

Glaucoma (n = 10)

Healthy controls (n = 20)

Stroke (n = 10)

Glaucoma (n = 10)

Adults

N/A

Eye movements were recorded during a 40-min driving task

Eye, head and shoulder movements

Driving performance, horizontal gaze activity (HGA), horizontal gaze distribution (HGD), head and shoulder movements

Participants who failed displayed significantly reduced scanning activity in comparison to controls who passed the test (p < 0.05).

Participants who failed the driving assessment performed significantly less exploratory head and shoulder movements than control subjects and participants who passed the driving test (p < 0.001).

No difference was found regarding the horizontal gaze activity between the participant subgroups.

Participants who passed the driving test performed significantly more glances towards their visual field defect than participants who failed (p < 0.05).

Kasneci et al (2014b) [52]

Case-control study

n = 20

HH (n = 10, 4 right, 6 left)

Healthy controls (n = 10)

Ischemia (n = 8)

Brain surgery (n = 1)

Cranio-cerebral trauma (n = 1)

Adults

N/A

Eye movements were recorded while participants were asked to collect 20 products placed on two supermarket shelves as quickly as possible.

Eye movements

Horizontal gaze activity (HGA), glance proportion in percentage (GPP), glance frequency (GF)

No difference was found in HGA between the participant subgroups.

No significant difference between the subject subgroups regarding the proportion of glances beyond the 30° VF (GPP30c).

A significantly higher GPP-30° in participants with a right-sided defect compared with their controls.

A significant difference in GPP beyond 60° (GPP-60c) between participants with left HH who passed (HHp) and participants with HH who failed (HHf).

HHp performed significantly longer glances towards the visual field periphery beyond 60° than HHf (p < 0.05).

Keller et al (2010) [53]

RCT

n = 20

HH (n = 13, 7 right, 6 left)

Quadrantanopia (n = 7)

Stroke (n = 18)

Tumour (n = 1)

Trauma (n = 1)

Adults

Audio-visual exploration training (AV) versus visual exploratory training (VT).

Each participant received 20 therapy sessions (each session lasting 30 min) over 3 weeks.

Eye movements were recorded after training

Eye movements

Number of saccades, amplitude of saccades

Comparisons between the two forms of training revealed a significantly greater improvement for all outcome variables for the AV group. In particular AV stimulation significantly increased the number (56 AV and 37 VT) and amplitude of saccades (52° AV and 44° VT).

Lévy-Bencheton et al (2015) [54]

Before and after study

n = 14

HH (8 right, 6 left)

Ischemic stroke

Adults

The training consisted of three different anti-saccades (AS) training: one with (adaptation) and two without (delayed shift and no shift) saccadic adaptation. Each training lasted around 15–20 min

Eye movements were recorded before and after training

Eye movements

Mean horizontal saccadic amplitude, mean fixation duration.

No significant increase of saccade amplitude during the pop-out exploration task in the group of participants who demonstrated an increase of anti-saccade amplitude during the adaptation training (p = 0.70).

Amplitude of saccades significantly decreased following the delayed shift training (p = 0.04) and following the no shift training for the group in which the adaptation training was inefficient (p = 0.02).

For the treated-side- a significant increase of the amplitude of return leftward saccades for the LHH participants with positive slopes following the adaptation training only (p = 0.001).

A decrease in the number of leftward saccades in LHH participants following the adaptation training.

Iorizzo et al (2011) [55]

Case-control study

n = 12

HH (n = 6)

Healthy controls (n = 6)

Stroke (n = 6)

Adults

Eye and head movements were recorded while subjects were asked to visually detect and track flying basketballs as they appeared in the virtual world.

Eye and head movements

Fixation durations and numbers, fixation distribution, degrees of head movements

HH participants made shorter fixations (227 ms ± 19) than controls (316 ms ± 12) when sitting, but not walking (161 ms ± 15; 158 ms ± 10 respectively).

When sitting, hemianopic subjects made more fixations (211 ± 97) than controls (89 ± 33) over a 2-min period.

When walking, they made a similar number of fixations (184 ± 48) over the first 5 paths as controls (118 ± 35).

Fixation bias into the BHF when sitting was greatly exaggerated (4.7 ± 2.1°) relative to that in sitting controls (1.1 ± 0.8; p = 0.002).

While walking, subjects placed most of their fixations in the side of their BHF (p = 0.002).

Head movements were generally small, with a standard deviation averaging just above 0.2° in the sitting condition and close to 1° in the walking condition, with no significant differences between controls (0.73 ± 0.17) and hemianopes (0.84 ± 0.29).

Machner et al (2009a) [56]

Case-control study

n = 18

HH (n = 9)

Healthy controls (n = 9)

Infarction (n = 7)

Haemorrhage (n = 2)

Adults

N/A

Participants performed an exploratory visual search task and eye movements were recorded.

Eye movements

Number of fixations, Saccades in the BHF (%), amplitude of saccades towards the blind/ intact hemifield, item refixations in the BHF (%)

Significant differences were found on the number of fixations: LHH participants (57 ± 3), RHH participants (104 ± 11) and controls (28 ± 1; p < 0.001).

In both LHH and RHH participants, the number of saccades did not differ significantly between the blind and the intact hemifield.

In LHH participants saccades towards the (blind) left side were slightly but significantly smaller than the rightward saccades (5.8 (±0.5) /6.3 (±0.4), p < 0.05). This was true for healthy controls (6.3 (±0.2) /7.0 (±0.3)).

RHH participants’ leftward saccades were slightly larger than the rightward saccades, but the difference was not significant.

The mean saccadic amplitude differed significantly (p < 0.01) between controls (6.5° ± 0.1), LHH (6.0° ± 0.2), and RHH participants (4.6° ± 0.2).

Item refixations increased for RHH participants (d = 31.3% ± 4.2, p < 0.001) and LHH participants (d = 17.0% ± 3.9, p < 0.001), when compared to controls.

Machner et al (2009b) [57]

Case-control study

n = 27

HH (n = 9)

Simulated HH (n = 9)

Healthy controls (n = 9)

Infarction (n = 7)

Haemorrhage (n = 2)

Adults

N/A

Eye movements were recorded while subjects searched for targets among distractors and indicated target detection by clicks.

Eye movements

Fixation duration, item refixations, number and amplitude of saccades

For participants with simulated HH, mean fixation duration (263 ms ± 6) was longer (p < 0.05) than in controls (230 ms ± 6) and HH participants (243 ms ± 6).

HH participants made more saccades (89 ± 7) than controls (29 ± 7, p < 0.001) and simulated HH participants (43 ± 7, p < 0.001).

Participants’ mean saccadic amplitude was significantly smaller than in controls (d = 1.2° ±0.4, p < 0.01) but not than in virtuals (d = 0.6° ± 0.4, p > 0.05).

HH participants showed a higher rate of refixations than controls (d = 23.3 ± 3.2, p < 0.001). Simulated HH rates of refixations were not just lower than in HH participants (d = 41.3 ± 3.2, p < 0.001) but also than in the controls (d = 18.0 ± 3.2, p < 0.001).

In both groups, HH participants and simulated HH, there were no significant differences between saccades which landed in the left or right half of the screen concerning total number, amplitude or fixation duration.

No significant differences were found when comparing rates of item refixations between the two absolute hemifields. For all three groups, numbers of saccades towards either side were evenly distributed.

Mannan et al (2010) [58]

Before and after study

n = 31

HH (n = 18, 13 left, 13 right)

Infarction (n = 22)

AVM (n = 6), Temporal lobectomy (n = 1), Glioma (n = 1), Cystic teratoma (n = 1)

Adults

Visual search training. 40-min sessions, 5x per week totalling 20 sessions.

Participants searched for randomly positioned target presented for 3 s (line, square or triangle) among 40 distractors

Eye movements

Proportion of initial saccades made into the BHF, mean duration for fixations made in BHF only, mean duration for fixations made in SHF only

Participants made exactly the same proportion initial saccades (50%) before and after training.

BHF: 54 (3), 51 (2), 51 (3), 56 (3) (visit 1,2,3,4 respectively).

SHF: 49 (3), 48 (2), 48 (3), 52 (3) (visit 1,2,3,4 respectively).

The proportion of fixations made in hemianopic hemispace also remained unchanged after training (47%).

Participants made a significantly greater proportion of fixations in hemianopic hemispace after training when the target was located on this side.

Participants made significantly fewer fixations compared to both visits 1 and 2. Mean durations of fixations made in the BHF did not change after training (194 (13), 197 (11), 189 (10), 188 (7) ms).

Participants made a significantly larger initial saccade after training (visit 2: 6.27°, visit 3: 6.81°). The mean amplitude of saccades following the initial saccade showed no effect of training.

Participants made 57% of their saccades in the direction of the BHF if the target was in hemianopic hemispace, while only 25% saccades were in the BHF for targets in intact hemispace.

Martin et al (2007) [26]

Case-control study

n = 8

HH (n = 4)

Healthy controls (n = 4)

Stroke

Adults

N/A

Participants’ gaze patterns were analysed while they assembled wooden models.

Eye movements

Saccade duration, saccade amplitude, saccade peak velocity, fixation duration, accuracy of fixations, look-ahead and look-back fixations

HH participants made slightly longer saccades (73 ± 1.5 ms) compared to controls (69 ± 2.5 ms). The difference was not statistically significant (p = 0.34).

HH participants made slightly larger saccades (9.32° ± 0.9), than controls (8.29° ± 0.5). Difference was not statistically significant (p = 0.33).

Saccade peak velocity was slightly higher in HH (163.2 ± 11.5d/s) than controls (155.3 ± 9.9d/s). Again difference did not reach statistical significance (p = 0.63).

HH participants exhibited slightly shorter fixation durations (510 ± 151 ms) than controls (561 ± 41 ms; p = 0.72).

Fixation durations did not differ significantly between the sighted and blind hemifields of each hemianope.

Hemianopes had an average of 5.0 ± 4.0 look-back fixations, while controls had an average of 1.5 ± 1.73 (not statistically significant).

Pambakian et al (2000) [25]

Case-control study

n = 16

HH (n = 8, 4 right, 4 left)

Healthy controls (n = 8)

Infarct (n = 4)

AVM (n = 2)

Cystic suprasellar teratoma (n = 1)

Meningioma (n = 1)

Adults

N/A

Eye movements were recorded while participants viewed 22 images of real scenes. Subjects viewed each image for 3 s, initially in a spatially filtered form in which much of the semantic content had been removed, and then in their unfiltered, original form.

Eye movements

Number and duration of fixations, mean amplitude of first saccade, mean amplitude of all measured saccades, scanpath length, proportion of fixations made in BHF, percentage of saccades made towards the BHF

Participants made significantly more fixations than control subjects while viewing filtered (8.7 (0.2), 7.7 (0.1) respectively, p < 0.0002) but not unfiltered pictures (9.8 (0.1), 9.6 (0.1), p < 0.2).

Participants’ fixations duration were significantly shorter than those of control subjects (312 ms (8), 372 (11), p < 0.0001 for filtered images) and (286 ms (6), 302 (11), p < 0.2 for unfiltered images).

The mean amplitude of the first saccade was significantly larger for participants than control subjects for both filtered (3.2° (0.1), 2.3° (0.1), p < 0.0001) and unfiltered images (3.3° (0.1), 2.6° (0.1)).

Mean amplitude of the remainder of the saccades did not reach significance.

Participants produced significantly longer scanpaths than controls while viewing both filtered (32.2° (1.4), 26.8° (0.8), p < 0.0415) and unfiltered images (39.8° (1.0), 37.2° (0.8), p < 0.0415).

Viewing filtered images, left and right hemianopes made significantly more fixations into the BHF (LHH: 51 (1.4), controls (42 (1.2), p < 0.05); RHH: 63 (1.6), controls (58 (1.2), p < 0.05).

HH participants spent significantly longer making fixations in their respective BHF compared to controls for both filtered (LHH: 71 (2.3)/42 (1.9), p < 0.05; RHH: 71 (2.5) /58 (1.9), p < 0.05) and unfiltered images (LHH: 59 (2.9)/50 (2.1), p < 0.05; RHH: 58 (2.3)/50 (2.1), p < 0.05).

Participants made significantly more saccades towards their blind side compared to controls for filtered (LHH:71 (2.1)/44 (2.5), p < 0.05; RHH:71 (2.4)/56 (2.5) p < 0.05) and unfiltered images (LHH:61 (2.6)/50 (2.1); RHH:60 (2.1)/50 (2.1), p < 0.05).

Papageorgiou et al (2012) [59]

Case-control study

n = 60

HH (n = 20) Quadrantanopia (n = 10)

Normal subjects (30)

Vascular brain lesion

Adults

N/A

Participants performed a collision avoidance task with moving objects at an intersection under two difficulty levels. Based on their performance (i.e. the number of collisions), participants were assigned to either HP or LP subgroups by the median split method.

Eye movements

Saccade amplitudes, scanpath length, mean number of gaze shifts, number and the duration of fixations, mean gaze eccentricity

Compared to LP participants, visual exploration of HP participants at both traffic densities was characterised by larger saccadic amplitudes towards the BHF (18.06° HP, 14.58° LP for 50% density, p < 0.01).

Shorter saccades to the BHF were evident for HP (18.06°), LP participants (14.58°) compared to controls (22.03°).

HP participants showed significantly longer scanpaths than LP participants (1126.21° and 793.38°, p < 0.001 at 75% density).

HP participants showed more gaze shifts (10.5) and larger mean gaze eccentricity (22.38°) compared to LP (7.16, 17.22° respectively).

No significant differences between HP and LP participants regarding total fixation number, fixation duration, proportion of fixations and proportion of gaze eccentricity to the BHF.

Both HP and LP participants showed increased numbers of fixations to the BHF compared to their SHF (59.88/40.12; 59.52/40.48 respectively).

Passamonti et al (2009) [60]

Controlled before and after study

n = 24

HH (n = 12, 6 right, 6 left)

Healthy controls (n = 12)

Vascular (n = 7)

Trauma (n = 3)

AVM (n = 1)

Craniotomy (n = 1)

Adults

Audio-visual (experimental) training vs visual (control) training. 4 h over a period of 2 weeks

Oculomotor responses during a visual search task and a reading task were studied before and after training

Eye movements

Number and duration of fixations, refixations, saccadic amplitude and duration, length of scanpath, number of progressive and regressive saccades, return sweep

Participants took more time than control subjects to explore the visual display, and also produced more fixations and refixations, resulting in a longer scanpath. Saccades were slower and shorter in amplitude (p < 0.05 in all comparisons).

After the AV training—significantly fewer fixations (LH: 95 (24) at S1 and 60 (16) at S4, RH: 100 (32) at S1 and 83 (29) at S4). Saccadic duration was markedly reduced (130 ms (30) at S1, 84 ms (15) at S4).

Mean saccadic amplitude significantly increased (LH:7.69 (0.7) at S1; 9.32 (0.2) at S4).

Length of scanpath significantly reduced in RHH and LHH participants after training LHH:624 (90) at S1;456 (72) at S4, RHH:746 (165) at S1;587 (198) at S4.

Rate of refixations (%) significantly reduced (RHH: 11 (2)–5 (2) and LHH: 12 (3)–4 (2)).

Reinhard et al (2014) [61]

Case-control study

n = 47

HH (n = 33, 15 right, 18 left)

Healthy controls (14)

Infarction (n = 12)

Traumatic (n = 6)

Haemorrhage (n = 4)

Postoperative complications (n = 7)

AVM (n = 2)

Unknown (n = 2)

Adults

N/A

Eye movements were recorded during saccadic and fixation tasks

Eye movements

Landing accuracy of saccades, Fixation stability (FS) after landing, distribution of fixational eye movements (FEMs)

The mean number of hypermetric saccades during gaze shift to the left was approximately 2.5 times higher in LHH (1.2 ± 0.2 SEM) compared with RHH (0.5 ± 0.1 SEM).

The mean number of hypermetric saccades was much higher during gaze shifts to the BHF (1.2 ± 0.2 SEM) than to the SHF (0.5 ± 0.2 SEM) (p = 0.003) and compared with control subjects (0.26 ± 0.08 SEM) (p = 0.0001).

Fixation stability after landing was lower after saccades to the BHF compared with the seeing side with the mean FS of 67.5% vs 74.5%, respectively.

Mean FS after gaze shifts to the seeing side was significantly higher than the FS after gaze shifts to the BHF (p = 0.013).

Fixational eye movements were asymmetric towards the hemianopic side. In participants with LHH, the mean was shifted to -1.312° (left) (p < 0.0001). In participants with RHH, the mean was shifted to 2.6° (right) (p < 0.0001).

Roth et al (2009) [62]

RCT

n = 30

HH

Ischemic stroke (n = 22)

Haemorrhage (n = 4)

AVM (n = 1)

Arachnoidal cyst (n = 1)

Cerebral abscess (n = 1)

Head injury (n = 1)

Adults

Participants were randomly assigned to distinct groups performing for 6 weeks either Explorative Saccade Training (EST) (a digit-search task) or Flicker Training (FT) (BHF stimulation by flickering letters). Training sessions lasted 30 min each.

Eye movements during initial fixation and subsequent free exploration of natural scenes were recorded

Eye movements

Fixation number, fixation stability

Only with EST did the number of fixations during natural scene exploration increase towards the BHF (follow-up/pre difference, 238%).

The EST group showed decreased fixation stability (post/pre, 43%) and increased asymmetry of fixations towards the BHF (post/pre, 482%).

Schoepf et al (1996) [63]

Case-control study

n = 22

HH (n = 12, 5 right, 3 left)

Healthy controls (10)

Infarction (n = 6)

Stenosis (n = 1),

Neuritis (n = 1)

Adults

N/A

Eye movements were recorded while subjects participated in two experiments: reading experiment and predictable target experiment

Eye movements

Accuracy and distribution of eye-head gaze saccades at different target frequencies, mean velocity of eye-head gaze saccades, mean amplitude of eye-head gaze saccades

Participants with RHH demonstrated a comparable distribution of correct saccades in both hemifields (M = 60.30 ± 3.2).

In LHH participants, the percentage of undershooting and correct saccades was nearly the same and stayed around 40% in the BHF (median 38.5%, range 35.9% to 41.1%).

In the better adapted RHH participants, amplitudes of eye-head gaze saccades became significantly increased and corresponding velocities faster with increasing target frequencies.

No significant differences between saccades directed either into the SHF or into the BHF.

Tant et al (2002) [64]

Case-control study

n = 45

HH (n = 29, 15 right, 14 left)

Healthy controls (n = 16)

Stroke (n = 23)

Tumour (n = 3)

Hydrocephalus (n = 1)

Head injury (n = 2)

Adults

N/A

Hemianopic condition was simulated in healthy subjects and their scanning performance on a dot counting task was compared to their own “normal” condition and to real HH participants’ performance

Eye movements

Number and duration of the fixations, number and amplitude of the saccades, length of the scanpath

HH participants displayed significantly higher number of fixations (33) compared to simulated HH (24) and normal controls (14). Simulated HH provoked higher duration of fixations (560 ms) compared to HH participants (380 ms) and normal controls (390 ms).

In both HH groups, there were more fixations in ipsilateral (sHH 12.5; HH 15.7) than in contralateral hemispace (sHH 8.9, HH 9.1 )

Ipsilateral saccadic amplitudes were smaller than contralateral amplitudes in simulated (9.08°, 10.81°) and true HH (7.00°, 8.2°).

The amplitudes of saccades into the BHF were smaller than into the SHF, for both left- and right-sided sHH.

Turton et al (2018) [65]

Cohort study

n = 9HH (n = 7, 1 right, 6 left)

Quadrantanopia (n = 1)

Partial visual field loss with macular involvement (n = 1)

Ischemic stroke (n = 6)

Haemorrhage (n = 3)

Adults

Participants received an intensive course of scanning and search training using various process activities and occupations. Training lasted 1 h, three times a week for 3 weeks

Head movement

Head yaw and pitch

Increased searching on blind side, from head movement, following intervention (for n = 3 of 9)

Wood et al (2011) [66]

Case-control study

n = 60

HH (n = 22, 5 right, 17 left)

Quadrantanopia (n = 8)

Healthy controls (n = 30)

Cerebrovascular accident (n = 12)

Tumour (n = 2)

AVM (n = 2)

Congenital brain abnormality (n = 1)

Trauma (n = 5)

Adults

N/A

The driving performance of each participant was assessed under in-traffic conditions along 6.3 miles of non-interstate driving in residential and commercial areas of a city. Eye and head movements and lane keeping were rated.

Eye, head and shoulder movements

Scoring of head movements, the extent of eye movements

HH participants exhibited more head movements (57.81) than quadrantanopes (51.06) and normal controls (50.17). They also had more shoulder movements (5.45) than quadrantanopes (4.00) and controls (3.37).

HH and quadrantanopic participants made significantly more head movements into their BHF than SHF.

59% of all head movements were made into the BHF compared with the SHF (33.18 vs. 22.66), with 30% more large head movements made into the BHF compared with the SHF (13.67 vs. 10.53) and 60% more small head movements made into the BHF relative to the SHF (19.51 vs.12.13).

The safe drivers made 70% more small head movements into their blind than seeing field (21.13 vs.12.30), whereas the unsafe drivers made only 20% more small head movements into their BHF than SHF.

Zangemeister et al (1982) [67]

Case series

n = 6

HH (4 right, 2 left)

Haemorrhage (n = 2)

Infarction (n = 1)

AVM (n = 1)Occipital porencephaly (n = 1)

Transsynaptic retrograde degeneration (n = 1).

Adults

N/A

Eye-head movements were recorded while participants searched for target presented in the horizontal plane within 10° of either side of centre.

Eye and head movements

Latencies of eye (EL) and head (HL) movements, head movement velocities and compensatory eye movement (CEM) velocities.

Head movements to the SHF followed eye movements with the same 50 ms delay observed in normal subjects, but there was an additional delay of 80 ms on the average in head movements to the BHF.

Eye movements were symmetrically delayed by about 110 ms more than in normal subjects.

In participants with acquired HH, the ratio of CEM velocity to head velocity (CEM gain) was small with head movements towards the SHF, great with movements towards the BHF.

Head movements towards the BHF often showed a staircase pattern. Head movements into the SHF often consisted of a single, large movement.

Zihl et al (1995) [23]

Case-control study/ before-after

n = 74

HH (n = 60, 30 left, 30 right)

Healthy controls (n = 14)

Cerebrovascular (n = 55)

Traumatic (n = 5)

Adults

A subgroup of participants (n = 14) were systematically treated with saccadic eye movement training to compensate better for their hemianopic field loss. 30 min each session (total of 16 sessions on average)

Eye movements were recorded while subjects inspected a pattern of 20 white dots randomly scattered over the screen, with a mean spatial separation of 7°.

Eye movements

Number and duration of fixations, amplitude of saccades, rates of repetitions of fixations, length of scanpath

The scanning pattern of “pathologic” HH characterised by significantly higher numbers of fixations not only in the BHF, but also in the SHF (38/27 in LHH and 45/24 in RHH) compared to controls (09/11). The repetition rates of scanpath and fixations were significantly higher in both hemifields (37/45 in LHH and 26/48 in RHH) compared to controls (02/13).

Amplitudes of saccadic eye movements and durations of fixations did not differ significantly between the groups.

Length of scanpath significantly increased in HH participants (LHH: SHF/BHF:100.7/159°; RHH: SHF/BHF 84.3/180°) compared to healthy controls (36.9/49°).

The repetition rates of scanpaths and fixations parameters significantly reduced after treatment to less than one half as compared to pre-treatment (scanpath: 17% and 8%; fixations: 44% and 22%, p < 0.0001).

Mean amplitudes of saccades were significantly larger after practice (mean increase was 0.9°), while mean durations of fixations remained unchanged.

Zihl et al (1999) [68]

Case-control study

n = 100

VFD (n = 70: 40 HH, 20 quadrantanopia, 10 HH).

Normal controls (n = 30)

Stroke

Adults

N/A

Eye movements were recorded during the inspection of a stimulus pattern. The stimulus array consisted of 20 dots.

Eye movements

Number and duration of fixations, fixation repetition, saccadic amplitude, scan path length

In the impaired groups, scanning times on average were about twice as long and scan paths were twice as large. The number of fixations exceeded that of the unimpaired group by 55–71% (M 75.9%); the rate of refixations exceeded the corresponding values by 2–28% (M 18.9%).

Mean fixation durations were sometimes shorter (0.26 s (0.04)) and sometimes longer (0.32 s (0.08)) than controls (0.28 s (0.05)).

Subjects had smaller saccadic amplitudes (5.0° (0.8) than controls (5.4° (0.7).

Subjects showed larger scan paths (295.9° (51.4)) compared to controls (137.6° (46.5)).