Stereopsis Slides PDF

Stereopsis Advanced Clinical Skills Dr Sheila Rae 1 Stereopsis Depth can be perceived or judged in two forms Binocular disparity (stereopsis) Monocular depth cues Relative size of objects Texture gradient Motion parallax Aerial p...

Stereopsis Advanced Clinical Skills Dr Sheila Rae 1 Stereopsis Depth can be perceived or judged in two forms Binocular disparity (stereopsis) Monocular depth cues Relative size of objects Texture gradient Motion parallax Aerial perspective Linear perspective Accommodation for near objects Overlapping contours 2 Depth perception: monocular cues 3 Stereoacuity Angular measurement of minimal resolvable binocular disparity which is necessary for the appreciation of depth Measured in seconds of arc (“) Smaller numbers mean better stereoacuity Greater the disparity the greater the depth effect Better stereoacuity means smaller amounts of disparity can be detected Smaller depth effects 4 Stereoacuity 5 Stereopsis vs. stereoacuity Stereoacuity is a measurement of the limits of disparity detection Not often done in the clinical setting Stereopsis is a clinical assessment of disparity detection Usually assessed in bigger increments Not a precise threshold 6 Reasons to assess stereopsis Children to demonstrate the presence of BSV Indicator of normal development Screening test to detect strabismus, anisometropia and amblyopia Reduced stereopsis indicates Failure of motor fusion (strabismus) Failure of sensory fusion or simultaneous perception (anisometropia and amblyopia) Monitor improvements due to treatment Strabismus surgery Occlusion therapy 7 Reasons to assess stereopsis Adults in diagnosis of decompensating latent deviations Shows system is struggling to cope Struggling to maintain motor and sensory fusion Adults in management of decompensating latent deviations Improvements show management is working Normal adult levels should be at least 40" 8 Factors affecting stereoacuity Monocular blur is more detrimental to stereo acuity than binocular blur Anisometropia and amblyopia will reduce stereopsis Should be assessed with patients spectacles on Optimal correction for the test distance Stereopsis will be reduced or absent with a failure of motor fusion or sensory fusion 9 Stereopsis theory Stereopsis is the perception of the relative depth of objects on the basis of binocular disparity Inputs from right and left retinas combined and compared Images at corresponding points (or with Panum’s areas) compared There will always be a slight difference in the images from each eye due to the inter pupillary distance Each eye is seeing the same object from a slightly different angle Brain compares the inputs from each eye and matches detail that it recognises as being the same The disparity between the matched details is then used to judge depth 10  Stereopsis theory Local stereopsis Global stereopsis Contour based Disparity of clusters within a Contours in each image are relatively large stereoscopic matched pattern (random dot) Disparity identified Requires more visual Requires less visual perception perception Gross tests effective Tests need to be sensitive Affected more by Titmus, Frisby, Randot (in parts) strabismus TNO, Lang, Randot (in parts) 11  Local vs. global stereopsis Local Global 12 Stereopsis Tested clinically Usually tested at near only Qualitative Evidence of stereopsis Quantitative A measure of stereoacuity Graded response Calibrated for a specific viewing distance Assumes an average IPD 13 How stereopsis tests work Disparate images presented to each eye Slight displacement of the target seen by each eye If this is disparity is detected by the brain (which require sensory fusion of the two images) then the targets would be seen in depth Tests need some method of presenting different targets to each eye Most tests use horizontal disparity to simulate depth Some tests use disparity created by different viewing distances in free space 14 Stereopsis test results In the clinical setting, putting the patient in a bracket is usually sufficient, rather than determining a precise end point Want a test to have good sensitivity and good specificity for picking up strabismus and anisometropic amblyopia No stereopsis Gross stereopsis > 300” Reduced, but present stereopsis 80-300” Normal stereopsis 60” or better 15 Dissociation varies between tests No dissociation (real depth) Frisby Lang two pencil Howard Dolman Partial dissociation Polarisation Titmus (Wirt) Randot Mallett unit Red & green filters TNO Computerised tests Optical Lang 16 Crossed and uncrossed disparity Tests can be designed with the horizontal disparity either crossed or uncrossed Most tests crossed when used in the correct way with the correct filters Crossed disparity tends to give slightly higher stereo values in normal and exo Px Crossed disparity matches the latent deviation position Eso’s tend to do worse Holding the test upside down or reversing the filters swaps the disparity from crossed to uncrossed 17 Tests suitable for pre-schoolers No filters required Frisby Lang Filters required Stereo smile 18 Lang two pencil test Test for gross stereopsis, qualitative Patient attempts to place pencil on top of pencil held by examiner Patients with stereopsis finds it easier binocularly, than monocularly Patient with reduced stereo will find both difficult 19 Frisby Stereotest Can be held at variable distances to give a range of disparity Random shapes on on transparent perspex sheets Sheets are 6mm, 3mm and 1.5mm in thickness to give different depth values Shapes forming a square printed on one side, a circle printed on the other side to simulate depth in one of the four squares 20 Frisby Stereotest Measures stereopsis from to 340 to 55” Good for young children as no filters needed Eye movements can be observed Sounds to enhance Moncular cues (parallax) can show depth without stereopsis if plates held flat against a surface Standard test distance 40cm As you double the distance, you quarter the disparity 21 Frisby near stereo test Disparity changes with changed test distance 22  Lang Panography dissociation High powered cylinders Separate images in strips 23 cyls per cm 46 strips, 23 per eye No filters needed Random dot Uses Julesz principle Lang I (cat, car, star) At 40 cm: 1200” to 550” Lang II (elephant, car, star, moon) At 40 cm: 600” – 200” 23 Lang I &II tests Random dot stereogram and cylindrical gratings Should not be used to measure stereo thresholds Lang I – 1200”, 600”and 550” 0 – 36% strabismic children fail this test Lang II – 600”, 400” and 200” Low sensitivity, high specificity, regardless of age Better at detecting strabismic than anisometropic amblyopia 24 Lang I &II tests Lang I – 1200”, 600”and 550” 0 – 36% strabismic children fail this test Lang II – 600”, 400” and 200” Low sensitivity, high specificity, regardless of age Better at detecting strabismic than anisometropic amblyopia Monocular cues? Children are more likely to miss elephant 25 Stereo Smile test (PASS test) PASS test: Preschool Assessment of Stereopsis with a Smile Useful screening test with young children Not seen often in UK Does require filters to be worn Stereo Smile II One demo card, one blank card and four test cards Used at 40cm 26 Stereo smile test (PASS test) Can use as preferential looking in the 6-24 month age group Toddler or baby will choose to look at something interesting, rather than nothing Will look at the smile plate rather than a blank plate Observe the direction of their gaze Older children can point to the smiley face Stereo Smile II 480, 240, 120 and 60” 27 Tests suitable for school-age children and adults Filters need to be worn Local stereopsis Titmus (Wirt) Randot (part) Mallett unit Global stereopsis TNO Randot (part) 28 Mallett Unit Polarised dissociation Two rows of symbols Add 3Δ base out to shift the image from one eye Fuse the two rows of shapes Symbols vary in disparity 10’ (600”) to 30” of arc Viewing Distance 36cm Requires motor fusion 29 Titmus (Wirt, 1971) Partial dissociation Crossed polarising filters Vectographs polarised to oblique meridians 45 deg (LE) and 135 deg (RE) Relative depth Qualitative assessment Fly ~ 3000 sec of arc Quantitative Animals 400 – 100 sec Circles 800 – 40 sec of arc Ask which circle or animal from a group seems to stand out Suppression check at bottom Viewing distance of ‘fly’ 16” (approx 40 cm) R or L disappears if suppressing 30 Titmus Stereotest Young children can pinch the wings of the fly to show gross stereopsis Animals measure up to 100“ 400, 200, 100” Use these with younger children Circles up to 40" 800, 400, 200, 140, 100, 80, 60, 50, 40” Use these (omit the animals) with older children and adults 31 Titmus Stereotest Monocular cues mean the disparity can be seen without the filters Especially with larger disparities Fourth group of circles always gives weird results Top row: 800, 400, 200 Middle: 140, 100, 80 Bottom: 60, 50, 40 32 Random dot presentation Julesz, 1960 Random array of dots No form within the dots No monocular cues to depth Each square identical except Section of dot pattern displaced horizontally No stereoscopic information in any one dot Possible to make inappropriate matches Needs global processing Brain matches dots across the whole pattern Filled in 33 Randot Partial dissociation Polarised vectographs Part contour stimuli Some monocular cues Part random dot principle No monocular cues Simple geometric shapes Quantitative assessment 500 to 20” of arc Viewing distance 16” (approx 40cm) 34 Randot 35 TNO Partial dissociation Red & green filters (red before LE) Random dot principle No monocular cues 3 screening plates 1980" 3 quantitative plates 4 presentations on each plate 2 presentations at each level 480 – 15sec of arc Viewing distance 40 cm 36 TNO Random dot analglyph dissociated with R&G filters Used from 30 months old Plates I to 3 screening, ~ 2000” Butterfly and shapes Plates 4 suppression Three circles, red, white green One circle disappears if suppressing Plates 5 to 7 qualitative 480” to 15” Newer editions 5&6 only, stop at 60” Sensitivity = 47% (Ohlsson, 2001) Sensitivity 37%, specificity = 86% children age 3-6 years (Molgaard, 1984) Better at detecting strabismic amblyopia 37 TNO ‘Pie’ plates used for grading Newer test versions don’t include plate VII 38 Crossed and uncrossed disparity TNO tests in the labs The versions we have are uncrossed as the R & G filters are the wrong way round – GREEN should be on the right eye On the quantitative plates, the ‘pie’ should normally stand out from the box and the ‘missing piece’ should appear flat When uncrossed, the ‘pie’ appears to recede and the ‘missing piece’ appears to stand out Can rectify this by either swapping the filters round or holding the book upside down Make sure the instructions you are giving the patient fit with what they are actually seeing 39 Distance stereopsis tests Most tests are designed for near Not useful for BV problems at distance Not useful for other purposes such as assessing stereopsis in monovision CL wearers DV tests are available FD2 (Frisby) Randot distance Versions on computerised test charts 40 Comparison between near stereopsis tests 41 Testability, sensitivity and specificity How good a stereo test is can be judged in a number of ways Can compare the mean stereopsis values Tend to vary between tests due to test design Not too important as long as you know the norms for your test Testability is a measure of the % of people tested who can successfully perform the test Important for tests for younger children 42 Testability, sensitivity and specificity Sensitivity is a measure of false negative rates People who have a condition but pass the test Strabismus, amblyopia, significant refractive error Specificity is a measure of false positive rates People who are normal but have ‘fail’ the test Unnecessary referrals Testability sensitivity and specificity are all important when using stereotests to screen for strabismus and amblyopia 43 Normal values (Randot) 3 years 5 years 7 years Mean 100” 60” 40” Lower limit 400” 200” 60” of normal Seven year olds demonstrate essential adult levels of stereopsis Lower limit of normal is useful for referral decisions 44 Stereopsis and aging Generally agreed that stereopsis declines with age May be related to reducing VA and contrast sensitivity May be related to cognitive decline Elderly tend to perform worst where the stereo tasks are more complex 45