CVI Mascot - VIC


Facial Recognition Paper

Facial Recognition Paper

In our section on Recognition and CVI we considered some of the many different reason why people have difficulties recognising faces.

This paper considers a condition called congenital prosopagnosia.

Prosopagnosia is an inability recognising people by their faces, where other causes like low vision have been ruled out.

In his book The Mind's Eye, Oliver Sacks described his own prosopagnosia

My problem with recognizing faces extends not only to my nearest and dearest, but also to myself. Thus on several occasions I have apologized for almost bumping into a large bearded man, only to realise that the large bearded man was myself in a mirror.

Oliver Sacks, The Mind's Eye chapter four, Face-Blind

Congenital prosopagnosia means prosopagnosia since birth (as opposed to acquired, for example from a brain injury). The condition can be passed from parent to child in some families. Studies show that as many as one person in every fifty may be affected by prosopagnosia.

This research is looking at the pathways, from the occipital lobes where the visual information is received from the eyes, to the temporal lobes where the memories of known faces are stored (typically the right temporal lobe), and onto the frontal lobes. The frontal lobes are the conscious control centres of our brain, where we make our decisions, so once a person is recognised as someone we do or do not know, we decide whether and how to greet them.

The pathways this study is looking at are the:

  • ILF (inferior longitudinal fasciculus) - This pathway starts at the occipital lobes conducting the ventral stream and continues through to the frontal lobes.
  • IFOF (inferior fronto-occipital fasciculus) - This pathway runs from the front to the back of the brain connecting eye movements with attention. The frontal lobes are where our conscious attention is given and decisions are made. Sometimes it is called the 'executive function' or 'control panel'. The frontal lobes have been fed information from the dorsal and ventral streams, and thinks... "I wonder who that person is over there is, I'll look and maybe watch" and so a signal is sent locally to the fast eye movement instruction centre in the frontal lobes to look, and if the person is of interest back along the IFOF to a bit of the brain deep within the occipital lobes that instructs the eyes to follow the target.

The research is using an imaging process called DTI (diffusion tensor imaging), which is a type of MRI. DTI enables the researchers to look at brain pathway activity, specifically the ILF and IFOF.

Little is known about what causes congenital prosopagnosia, but the paper notes "the fact that there is a reduction in the volume of the anterior fusiform gyrus in congenital prosopagnosia".

The anterior fusiform gyrus is an area of the brain, like a long strip that runs from the occipital lobes to the temporal lobes on both the left and right sides.

So to understand the workings of the brain, there are many different components to consider, including:

  • The grey matter on the outside that does the computing and the...
  • White matter on the inside that contains the myriad connection cables

It is like cogs and wheels and pathways, all working together in harmony, ideally! But it only takes one little bit to not be working so well (as we explained in our introduction to a paper on Object Recognition) to have a considerable knock on effect on everything else.

The brain - it only takes one little bit to not be working so well, to have a considerable knock on effect on everything else.

The research found that the people with congenital prosopagnosia showed reduced activity across the ILF, particularly to the right, and IFOF. The visual information is not being processed and getting through to the frontal lobes along the ILF, and so the pathway back to issue the instruction to the eyes to 'look there' (the IFOF) does not get the information needed, so people go unrecognised. Is the lack of activity on the IFOF pathway due to the lack of information from the ILF pathway? It is possible, but further studies would be needed to confirm this.

The right temporal lobe is associated with facial recognition, thus the relative reduction of pathway activity from the right occipital lobe to the right temporal lobe, found in this research, is consistent with challenges recognising faces.

Prosopagnosia in isolation is not currently considered on its own to be a cerebral visual impairment but it can be a feature of CVI when there are more elements, such as visual field impairment or difficulties with visual search.

So where does prosopagnosia fit in the world of visual impairments?

  • There is an impairment - people can't recognise faces, and it can be embarrassing, Think how many people every day you see and pass; in the street, at work, in the park or out shopping. Each one is processed and recognised or not recognised, and if recognised this is linked to the complex memory of that person, and your behaviour will reflect the memory and determine whether you maybe say hi, give them a hug, ask a question or choose to ignore them. Imagine if every person you encountered was not recognisable - prosopagnosia is a very real problem
  • Some people with prosopagnosia cannot see nor interpret language in facial expressions such as a smile or an angry look
  • Prosopagnosia can accompany dorsal stream dysfunction, lack of visual field or a degree of reduced visual acuity or contrast sensitivity, and this bigger picture is one of cerebral visual impairment
  • The impairment is visual - many use alternative forms of recognition to substitute, particularly voices, but prosopagnosia is an inability to recognise people by their faces- that is visually by their faces, and rely on interpretation of what is said and how it is said, to understand what is meant.
  • The impairment is from the brain - quite how, where and why is still an issue for the researchers, but there is no doubt that the challenges caused by prosopagnosia are from the brain.

There are many views on how much of the brain relates to visual processing, some estimate 40%. We think a more interesting question is how much of the brain's processing does not rely in some way on visual information?

Each person who has any problem with visual processing, whether it fits within the CVI diagnostic conditions or not, is unique. The complex nature and range of challenges they face are unlikely to fit neatly into a single definition, be it simultanagnosia, prosopagnosia or hemianopia. They all rely on the cogs and wheels and pathways to work together, and if the result is an inability to recognise faces for example, then there will be many different possible reasons why, and these reasons will not fit into a little box.

We wonder whether there is a need to create a wider umbrella term to include all impairments that are both visual and due to the brain, so that children and adults who do not meet the criteria for a diagnosis, but clearly have a visually related impairment, have some sort of route to help and support?

Please note: This paper is behind a paywall, if you click on the link you can read the abstract for free, and if wanting to read the whole paper there is a reduced fee to read it.

Paper: Reduced structural connectivity in ventral visual cortex in congenital prosopagnosia. Cibu Thomas et al, Nature Neuroscience (2009)


Your generous donations will be put to immediate use in supporting our charity...

Donate Here

About Us

At CVI Scotland we are devoted to helping people understand cerebral visual impairments, and together working towards developing the understanding of this complex condition.