Proper Evaluation Of Blindsight Patients
Kyle Weiss
Claremont McKenna College
Morten Overgaard, Katrin Fehl, Kim Mouridsen,
Bo Bergholt, and Axel Cleeremans
When a patient gets injured in the V1 area of their primary visual cortex they can have impaired vision, or they could go completely blind. However, some patients who lose their vision can still show behaviors that are visually guided.. Blindsight allows blind patients to unconsciously detect stimuli in their blind field even though they are often not conscious of it. Blindsight was originally believed to be fabricated and so the history of it is short.
Blindsight was first seen in monkeys and it was shown that they could tell different shapes, colors, and contrasts in their blind visual fields. At first, researchers were surprised that humans shared the same impairment but that the humans saw themselves as blind. When researchers realized that patients did not perceive objects, they started doing similar tests and developed a correlation between eye movements and stimuli being presented in the blind field. Additionally, studies showed that even though patients would not be able to consciously perceive the stimuli, if asked to guess whether or not an object had appeared, they would oftentimes guess at a higher rate than the assumed 50% (Weiskrantz et al. 1974) This showed that using guesswork which they assumed was random, the blindsight patients were being accurate, even comparable to the levels of the monkeys previously tested, when trying to discriminate stimuli in their blind field. Research than began trying to figure out what exactly blindsight is.
On the visual pathway map, the V1 area is at the beginning of both the dorsal and ventral pathways and therefore necessary for vision. The V1 area is known as the primary visual cortex and is located in the occipital lobe. This translates to area 17 on Brodmanns map. The cells that are in the retina of the eye project to the visual cortex but in blindsight patients, this area is either completely or partially destroyed. However, the other pathway goes to the subcortical areas. Although the V1 pathways are much larger and are the main pathways used in vision, they are not the only pathways used in the visual process. The subcortical pathways seem to have a function that is still used even while the V1 pathway is destroyed.
Blindsight has traditionally been shown to occur with no actual visual awareness in the patients damaged fields. (Weiskrantz et al. 1974) This means that although patients may perceive an object in their visual field, they are not aware that they are actually perceiving it. For example, one of the most famous blindsight patients, (Benedict 2008) TN had a completely destroyed visual cortex, yet could walk down a hallway filled with objects and completely avoid all of those objects. However, when asked, TN was not aware that he had avoided the objects.
To further develop theories about blindsight, researchers began using much more accurate tests to test detection. Looking at the dilation of the pupils in blindsight patients was shown to be the most accurate way to test patients residual function. Patients would be presented with stimuli in their intact field and then something would be flashed in their blind field. The pupils would dilate different amounts which would tell researchers how accurately detected the object was (Barbur J. L., Forsyth P. M. 1986).
Using this pupil method of studying patients, blindsight patients were shown to have much more extensive abilities in their blind fields then previously assumed. The abilities of each patient are unique but as a generalization, blindsight patients have been able to detect and then discriminate simple shapes, color, line orientations, and motion among other things. (cf Sahraie et al., 2003) Additionally, blindsight subjects could guess the emotions shown on a person in their blind visual field at a rate higher than chance. Also, when subjects see a particularly fearful site in their blind field, their amygdala(which is sent stimuli from the senses and then sends the proper response signal) is activated. (DeGelder et al., 2001). Still however, this research has not given a definitive answer to blindsight so many theories still exist.
Developments in blindsight have led to much advancement in the theories surrounding the phenomenon. A large reason for this is blindsight’s relationship with consciousness. Since subjects are not actually aware that they are perceiving objects, those involved in philosophy are interested in whether thought needs to be conscious in order to affect us. (Rosenthal, 1993) These theories also gained prevalence due to the invention of transient magnetic stimulation or TMS. TMS allows researchers to temporarily simulate blindsight in non-damaged subjects.
The superior colliculus theory is one theory on blindsight and the one previously referenced with sub-cortical pathways. Basically, the superior colliculus has a retinotopic map which detects all movement of the eyes and then would be able to cue where the movement is coming from. Then the brain would still be able to perceive the object and not consciously know that it was doing so(Brodmanns Areas2,3,11,12). However this theory is controversial because the superior colliculus has a very broad function that cannot map out specifics such as color and line orientation, which some blindsight patients can do. Therefore it has been proposed that actually this theory only holds true for the location of an object and that other mechanisms explain other residual function.
Other theories exist such as those involving the geniculoexstrastriate pathways and partial functioning in the striate cortex also have been tested. If a combination of the three previously mentioned theories were true, that would explain the vast range of abilities that some blindsight patients have. In fact, many researchers think that the varying degrees of striate cortex damage account for the variability in blindsight behavior (Radoeva et al. 2008).
All of the tests that led to this collection of knowledge have been done using a binary method of data collection. This means that when a researcher held an object or moved an object in the patients blind visual field, the patient responded with either a yes or no answer, either the object was perceived or not. The only exception to this was the pupil dilation method but that did not measure actual perception, only whether the brain had registered the object. The problem with this is that the object might actually be perceived but not at a degree where a subject would confidently acknowledge the perception with a yes.
Full Blindsight Paper With the introduction of the perceptual awareness scale, many questions are raised about the validity of previous testing results, and the new questions that could be answered with more specific and accurate results.
CRITICAL ANALYSIS OF CASE STUDY: “Seeing without Seeing? Degraded Conscious Vision in a Blindsight Patient”
Overgaard, Fehl, Mouridsen, Bergholt, Cleeremans, 2008, examined the question of whether blindsight is conscious vision that has just been severely impaired, or whether it is actually vision that is occurring unconsciously. In addition to this question, they addressed the methodology of previous blindsight experiments (e.g., name, year)—specifically, whether previously mentioned test results were accurate in their use of binary testing methods.
The patient examined in this case study is a 31-year-old women, GR. She started having painful headaches that became increasing severe over a period of 2-3 months. One day, GR developed sudden blindness in her visual field on the right side. Her Glasgow Coma Scale number dropped to a 7. This number is on a scale of 3-15 and is a basic and very reliable way of pinpointing the consciousness of a patient using a combination of verbal, motor, and eye response. A fully awake person would be at a 15. A CT scan was taken which showed that a hemorrhage had formed in the anterior section of the left occipital lobe. This was also continued into the left lateral ventricle.
GR experienced a buildup of fluids around her brain. This is known as hydrocephalus and can be fatal She ended up having an endovascular embolization that stopped the bleeding from the hemorrhage.
After this procedure, CT scans showed that a significant amount of tissue was missing from the left occipital cortex. She woke with a decreased short-term memory and right-sided hemianopia. She was basically blind in her right visual field, which corresponds with the lesion in the left occipital lobe.
As GR recovered, she regained her short-term memory, with minor cognitive dysfunctions being corrected. However, , no progress was made on the hemianopia in her right visual field. In an examination, it was shown that the eyes were still fully functional and yet she was still found to be cortically blind in the upper right quadrant.
To assess basic visual function, individual letters were presented in different parts of a computer screen and she had to acknowledge whether she detected something. W white cross was on a black background was shown and then A, B and C were shown at random on one of fifty locations which spanned the patients whole visual field. The objects were shown for very quick increments, and GR had 3000 ms in order to acknowledge having seen the letter on the screen. A total of 150 trials were used at random in this testing. This was not testing actual sight and perception, only ability to perceive that “something” was being shown. This test had the expected results THAT proved the GR to be blind in the upper right field of vision. The border region also had an expected medium amount of impairment.
In the next experiments, GR was shown shapes instead of letters in order to mentally distinguish the tasks. A triangle, circle, and square were shown in the same manner as ABC were in experiment one but if the patient said they saw the shape, they were shown another screen which gave them pictures of the shapes so she could identify which of the shapes she had actually seen. The amount of tests were the same, but the majority of tests were done in the 11 upper right visual field locations that GR was blind in.
The final tests were done in the same way as the second tests but with one large difference. A different scale was used to measure perception. The second test used a binary system, the patient saw something, or she didn’t, there was no middle ground. This was the standard method for testing on the subject and most researchers had used it in the past. However the researchers on this case study decided to develop their own awareness test. They call it the Perceptual Awareness Scale or PAS. Patients with blindsight were asked to scale their awareness of shapes shown on a screen. All subjects chose a multiple point scale instead of a binary system. Furthermore, a general consensus among patients was to use a four point system which had CI or clear image, ACI or almost clear image, WG or weak glimpse, and NS or not seen. This system has continued to be used as an accurate scale. Additionally the amount of time on a response, has been shown to have a direct correlation with the accuracy of the guess. The faster people reacted and answered, the more likely they were to be correct.
The scale distinction is important to make before the analysis of the second and third tests because it could be the explanation of the interesting results that were seen. Under the binary system, the amount of shapes that were seen or not seen has been shifted more towards the not seen category. In the damaged region, GR would only say she saw the shape if she knew what it was. With the binary system no relationship was observed between the patients awareness and the accuracy of the test. However, when the third test was taken with the new system the results were rather surprising. There seems to exist a correlation between awareness and accuracy for the guess when using the four points.
In conclusion, awareness can actually be very useful in predicting accuracy in blindsight patients. This does not mean that people with a “weak glimpse” are just as accurate as those with a “clear image,” but the relationship exists between the four levels and increasing accuracy based on more clarity. Data showed that if just data taken from a point between ACI and WG to CI, then perception would almost directly correlate with accuracy.
The major result of this case study is that blindsight in many patients is not a case of unconscious vision but actually just severely impaired but still conscious vision. The new testing system was instrumental in making this determination. However, the data also suggests that the many past tests, which have been done on blindsight patients, could have potentially huge inaccuracies. For example, if we had just used binary and stuck with tests one and two in this case study, the results would directly contradict the results that came out of test three.
I agree with Overgaard, Fehl, Mouridsen, Bergholt, Cleeremans and would encourage future tests to be done using their methods. Proponents of binary method would argue that is simpler and more efficient which I agree with. However, in order to get more accurate data, the PAS method seems to be superior.
ORIGINAL EXPERIMENT: Testing Out the PAS Method in Regards to Emotion
One of the questions that needs to be reconsidered is blindsight subjects ability to discriminate emotion in their impaired field. Is the person being shown happy, sad, or excited? Also, can a patient discriminate between a happy face and a body expression suggesting anger or sadness?
To understand these questions it is necessary to examine previous studies on emotion in blindsight patients. Affective blindsight is the name given to subjects reacting to emotional stimuli with no conscious perception. The most famous of the studies surrounding this was done on patient GY who had damage to his left occipital lobe. This corresponded with his right visual field impairment. Patient GY was shown video clips of a female face showing some sort of emotional expression while repeating some sentence(which did not show any emotion) and then asked to guess what emotion the face was showing. GY showed a better than chance ability to discriminate emotions. (de Gelder, Vroomen, Pourtois, & Weiskrantz, 1999). A variety of tests were given but the abovementioned tests showed a rate of accuracy of 66% of the 333 stimuli being correctly guessed. This involved two sets of stimuli, happy/sad or angry/fearful. This means that with just guessing, a 50% success rate would be average. Clearly something is being detected but it is a very small correlation compared to the 100% accuracy rate that was obtained using GY’s non-damaged field. It is important to note that body language could not be guessed at a rate above chance.
This information shows that that there are sub cortical pathways that are able to detect some level of emotion despite it being a very weak correlation. The brain region involved is not completely known but the amygdala is most likely due to its role in emotional responses, but the sub cortical pathways have not been fully identified.
With this background info, the perceptual awareness scale could hopefully be used to obtain better results. I will be using a group of 50 blindsight patients with varying degrees of visual impairment. This is a large sample size given the limited amount of possible subjects. This sample population will be as random as possible and should be large enough to eliminate any possibility of not properly assessing the population of patients with blindsight. Additionally a control group of 20 subjects without any sort of visual impairment will be used to compare the main subjects to (it is predicted that these patients will score similar to what the patients would score in their non impaired areas, but given the nature of the testing, gathering non impaired data from blindsight subjects would be very difficult.)
As in the previous study with GY, this study would center around two sets of choices. Happy/Sad and Angry/Fearful. We will have both a mans and women’s monotone voice say “Well, how are you” and other uncompromising sentences that could be said in any emotion. While subjects are shown a screen in their blind field with a corresponding face with an emotion on it. The first 25 trials would be the Happy Sad Choice and the second 25 would be the Angry/Fearful.
The faces will be a random mix of race, age, etc. After five seconds of a face being presented in the blind field, we will incorporate the perception awareness scale. The four scale point would be the same, CI or clear image, ACI or almost clear image, WG or weak glimpse, and NS or not seen. Subjects will indicate how well they perceived the image before indicating which emotion was shown.
All subjects performed these tests and the compiled data is attached. The results indicate a high correlation between awareness and perception. The aggregate scores result in average of 69% across the impaired subjects in a combination of the happy/sad and angry/fearful tests. This is only a slight deviation from the mean of 67% indicated by the tests given by (de Gelder, Vroomen, Pourtois, & Weiskrantz, 1999). This could be a result of their small amount of trials. 333 trials with one subject is not very accurate so we can assume the increase to be a chance statistical error(not factoring out the possible difference).
These results indicate that in certain tests, the perception awareness scale will not lead to more accurate testing for blindsight patients. That is not to disprove the accuracy of the testing, simply to say that at times, a much simpler binary method can be used just as effectively as a PAS. In the summary filed after the testing, 87% of patients attributed the new scale as not factoring into their guesses for each trial . This means that although patients found the degree of vision of each photo to be better categorized by the PAS, in the end it had little effect on their guess which they still believed to be just that, a guess.
That said, just like in the study of patient GR, an extremely high correlation was shown between accuracy of “guess” and the PAS reading. This would support the original conclusion that blindsight is not unconscious sight, but highly degraded yet still conscious vision. This still goes against much previous blindsight research and all theories concerning blindsight should still be tested with the PAS or some non-binary method.
References
de Gelder, Beatrice (2010). "Uncanny Sight in the Blind". Scientific American (0036-8733), 302 (5), p. 60.
DOI: 10.1038/scientificamerican0510-60
Carey, Benedict (12/23/2008). "Blind, Yet Seeing: The Brain's Subconscious Visual Sense". The New York times (0362-4331), p. D5.
Schmid, Michael C. (2010). "Blindsight depends on the lateral geniculate nucleus". Nature (London) (0028-0836), 466 (7304), p. 373.
DOI: 10.1038/nature09179
Weiskrantz, L., E.K. Warrington, M.D. Sanders, and J. Marshall. 1974. Visual capacity in the hemianopic field following a restricted occipital ablation. In Brain 97:709-28.
Barbur J. L., Forsyth P. M. (1986). Can the pupil response be used as a measure of the visual input associated with the geniculo-striate pathway? Clinical Vision Sciences, 1, 107–111
Sahraie A, Weiskrantz L, Barbur JL, Simmons A, Williams SC, Brammer MJ (1997) Patternofneuronalactivityassociatedwithconsciousanduncon- scious processing of visual signals. Proc Natl Acad Sci USA 94:9406 –9411.
de Gelder, B., de Haan, E., & Heywood, C. (2001). Out of mind. Varieties of 568 unconscious processes. Oxford: Oxford University Press.
Morten Overgaard, Katrin Fehl, Kim Mouridsen, Bo Bergholt, Axel Cleeremans. Seeing without Seeing? Degraded Conscious Vision in a Blindsight Patient
PLoS ONE. 2008; 3(8): e3028. Published online 2008 August 21.
No comments:
Post a Comment