Dr Christopher Rynn (Educator)
MSc Forensic Art course convenor, CAHID, University of Dundee. Craniofacial anthropologist: specialist in forensic facial reconstruction and forensic facial image analysis.
Location University of Dundee
Activity
-
:-D Great questions!
There are some areas along the jawline where the skin attaches more firmly to the bone. You can draw a pentagon from the nose, down the nasolabial creases to just outside the corners of the mouth, then straight down the "marionette lines" from the corners of the mouth to the jawline, and a kind of strap of connective tissue across... -
Dr Christopher Rynn (Educator) replied to doug b
Hi Doug,
In forensic cases there is sometimes hair found at the scene, in which case it can be represented in some manner. If none is found at the scene and it's a male skull, we crop the top of the head off the image to avoid misrepresenting hairline. If it's a female skull then we try to represent the hair as vague as possible, so it's difficult to tell... -
Hi Mary Ann,
Ageing in the adult face is predominantly textural, until the age of around 50,so it's mostly taken care of in the application of texture in photoshop after the sculpture is done in Freeform.
The effects of gravity and loss of elasticity in the older face becomes more structural, so is considered in the "skinning" of the 3D model.
-
Hi Richard,
We can gauge the general size and projection of the ears, but not much at all about specific shape of the auricle, dictated by cartilage. This specific shape is a good biometric when matching high resolution facial images to one another (except when the angle is different), but it plays a relatively small role in facial recognition. -
Hi Sara, Hi Eve,
Essentially, the method will estimate the nose before it was broken, then we have to try to replicate the individual break, so I'd expect lower accuracy due to individual interpretation. -
Hi Daniel,
Thanks for the comment :-)
Which projections do you mean? -
Dr Christopher Rynn (Educator) replied to Wilma P
Hi Wilma, Hi Deborah,
The database is essentially a shortcut to speed things up. Originally a set of muscles were produced which fit a given skull. Once these are imported into the system, they can essentially be stretched and inflated to fit any skull, which is much quicker than sculpting them from scratch.Skulls which have a bigger temporalis muscle...
-
Hi Jane,
Once you get used to it, using the haptic feedback armature does feel very similar to sculpting with a tool.
Yes, it is always preferable to have access to the actual skull (or close up photos) for thorough analysis, but for forensic casework, it speeds things up a lot to do it virtually and the end product is a non-physical image anyway.
If it... -
Yes - the lips should change to follow the dentition.
From personal experience - I had 4 premolars removed and a fixed-brace to straighten my remaining teeth as a young teenager. So compared to my older brother who had no orthodontic intervention, my palate, chin and mouth all ended up smaller, and my nose also appears slightly larger, because it's all... -
Hi Caroline,
The data has always been grouped using sex, age, ancestry, but interestingly, there seems to be a much greater effect from functional patterns, like dental occlusion, i.e. Type 2 (overbite) has deeper tissue below the mouth, as the muscles act to keep the lips closed in life. Conversely, Type 3 ("underbite") has deeper tissue above the mouth,... -
Thanks Marie,
Yes, the face is pretty much there, with all the muscles and features in place, before any subjective decisions are made at all. -
We used to take a plaster cast of the skull, use wooden kebab sticks (cut to length with scalpels and glued into holes, drilled into the plaster cast skull) for the tissue depth markers, then sculpt the muscles and face in actual clay.
I kinda miss getting my hands dirty, but I've managed to find other ways to do that too! ;-) -
Hi Sara,
Interesting point,
If we found dentures with the body, I'd presume the person wore them in life and would be recognisable with them (whether they belonged to them in the first place, or were made to fit someone else!) so I'd definitely use them in the facial reconstruction.If there are no dentures or teeth, it becomes problematic to set the...
-
The database of muscles / facial features is just a starting point, really, in that once a feature is imported into the software, it becomes "virtual clay" and can be manually stretched, squashed or deformed into any shape.
As we build the database bigger, it speeds things up because it means less 'sculpture' - less manual distortion, if there is e.g. a nose... -
Hi Jennifer,
In the last couple of succesful cases we've had, it has proven to be all important where the image was circulated.
In one case from UK, a label on the clothes was from Lithuania, so the image was circulated to the Lithuanian national missing persons authority and recognised by them as being similar to a photo they had on file.
In another case... -
Hi Everyone!
OK, Ears...
The mastoid process looks a bit different where adherent earlobes are concerned. In lobed ears (90% - genetically dominant) the mastoid process points diagonally down/forward. In adherent ears, the mastoid is said to "change direction" to point more downwards at the tip. In practice, I would say it looks a bit like if the bone... -
Originally sculpted to fit one skull, they can be saved as separate 3D models then manually distorted to fit any other skull.
-
Thanks Everyone,
Elaine's jigsaw puzzle analogy works well to illustrate that a very common misonception about bone is that it is rigid and static. This idea comes from the fact that it is the last thing left after death - but in life, bone is also alive: 'plastic' and responsive.
So it's not just bone which shapes muscle: muscle also shapes bone, as they... -
Hi Sara,
It's fantastic being able to do a manual sculptural process in virtual reality, mostly in the amount of time which is saved over using clay/wax and plaster.
It takes several days to take a plaster cast of a skull or head, but now a skull can be CT scanned in minutes and send via the internet to anywhere, pretty much instantly so that work can begin... -
Hi Elizabeth,
In practice, with the muscles of mastication (temporalis and masseter) it's all about the bulk. Temporalis should give a smooth contour between the temporal line and zygomatic arch, then masseter should run smoothly from zygomatic arch to the corner of the mandible. We can see from the shape of the mandible and the temporal fossa whether these... -
Thanks Patricia!
It's obviously sped up a bit in the video ;-) -
Unfamiliar Facial Recongnition and Familiar Facial Recognition are quite different, in terms of brain processing.
Broadly speaking, when unfamiliar with a face, we tend to remember more holistic things, like overall shape, jawline and hairline. As we become more familiar with a face, we start to focus on and recognise the finer details and the shape of the... -
Hi Steve,
The article was referring to misidentification by visual recognition, as in, by relatives viewing the body in the mortuary. -
Hi David,
True, there are a number of methods internationally, but most modern practitioners use a combination of methods which are updated with ongoing scientific research.
The images produced by facial reconstruction are seldom recognised by witnesses who are unfamiliar with the person in life, but the appeal is more targeted towards family and friends... -
Hi Rose,
We used to represent a range of hairstyles, if case-appropriate, but a few years ago settled on a different approach whereby we zoom in a bit and the top of the head is cropped off the image, if the hairline is inestimable, as in the two cases above.
If hair or scalp is found at the scene, this can be represented on the facial reconstruction. -
Hi Jenny, here's a couple of recent forensic cases which worked:
http://www.bbc.co.uk/programmes/profiles/2nvPV27X2mhY5ljYtkbj91X/bracknell-body
-
Hi Marie, I agree. The science behind facial reconstruction provides a lot of information about face shape, but certain things like hairstyle and hair/eye/skin colour are inestimable from skeletal remains alone. A way around this is to produce a black and white passport style image, and make all these details as vague as possible.
Even regarding face shape,... -
Thanks Valerie,
In ideal circumstances, with an intact skull of fairly average appearance, from a high resolution CT scan, it's rather difficult to say: it could take anywhere between 15 and 30 hours of work. -
There probably is some atavistic, indirect functional basis, but I haven't been able to figure it out... yet ;-)
-
It's an occupational hazard ;-)
-
Thanks Angela!
Check out Willem Dafoe's face, if you want to see the zygomatic muscles in action ;-)
I was lucky enough to get a broad sample of clinical data, comprising over 100 British lateral cephalograms with about 80 CT scans of European Americans, African Americans and East Asian Americans, which provided a wide range of nasal shapes, producing more... -
Thanks Susan!
Wonderful examples of both sides of the story regarding hair!
I have a seasonal beard, so the only people who don't recognise me clean-shaven are my students who have been here less than a year! :-D -
It turns out that TV host Stephen Colbert has an adherent right ear but a lobed left ear! Cate Blanchett noticed it in an interview the other night :-)
-
Facial ageing in the adult face is easier to estimate after the muscle stage. The skin would be extruded or 'grown' from the skull and muscles, producing a relatively young adult face. Then the 'skin' model would be tugged around a bit, to account for age-related changes such as loss of elasticity and sagging.
Average tissue depth data is available for... -
I don't think it's standard operating procedure across the board, but I do consider it best practice: in that it makes sense to avoid strong visual statements this way, rather than producing an array of multiple potential faces, each with specific colouring.
-
Upside down? (-;
-
Surface features such as birthmarks, and soft tissue trauma might not show on the bone, that's right.
-
Thanks Tania,
The sculpture system takes a couple of weeks to get the hang of it.
Regarding facial reconstruction, it's difficult to say. I've been working at it for 15 years and I'm still learning! :-) -
When interacting with the haptic virtual sculpture system, the skull model can be 'touched' and spun around using the position of your hand in 3D space, so it feels a lot more visceral than we could simulate without the haptic interface. Close enough to the real thing to get the shape of the face, but it will never be the same as the actual skull and a lot of...
-
Not so much in muscle size and shape, but more in the other soft tissues of the face, such as fat. In emaciated faces, the musculature is simply much more apparent.
-
Estimating lips in edentulous individuals is significantly more error prone, it's true. You lose all the information about dental profile in lateral view. The gumline might still be apparent, allowing some estimation of lip shape in frontal view, but there would be a lot of subjective extrapolation involved.
-
Anything with existing facial images would come under the heading of forensic facial image comparison. It is something we also do, but in those cases you have to avoid manipulating the image at all. In facial reconstruction, it's all about manipulating the image! :-)
-
Any asymmetry in the skull will be apparent in the face, although in some areas it is amplified, whereas in others it is lessened, depending on the soft anatomy between skull and surface.
-
European is just a label, really, like Caucasoid. The process is principally estimating face shape from skull shape. Elements indeterminate from the skull, such as surface textures, colours, etc. are rendered as vaguely as possible on the final image, which is why they tend to resemble black and white photos with imperfect focus.
-
You'd be surprised: with the Thiel method of soft-fix embalming we employ, surprising new discoveries are being made regarding muscle attachment and anatomical variation.
-
Thanks Sheila,
Pauline is right, as far as we know there is no way of accurately determining hairline from the skull. In adult male facial reconstructions, we tend to crop the image to avoid making a strong statement about the hairline. -
Hi Pauline,
All true. It's not a legally accepted form of identification, per se: more of a way of generating leads for an investigation. If the family or friends of a missing person is searching for them, this is a way of prompting a response via the media. -
Thanks Nicholas!
I'd always defer to a pathologist or osteologist regarding the cause of injury, but basically we would estimate the unbroken nose using a published method, then try to visualise how it might have looked after the injury, which would be much more subjective. -
Did you try one sideways? Not flattering! ;-)
-
Thanks Tracey, thanks Allie,
We run an MSc Medical and Forensic Art, the first semester of which is focused on anatomy studied through traditional methods, plus life drawing, human dissection and anatomical sculpture. -
You could also say that it's heavier because it's denser ;-)
-
Here's another example:
In 1861 Karl Langer (and 30 years earlier, Baron Dupuytren) found that, if you made round holes in the skin of a cadaver, they were stretched into lines by the natural tension in the tissue.
Langer's lines (great images if you Google it) are now used surgically, to guide the direction of incision to minimise scarring! -
Thanks Patricia! :-)
-
Thanks Heike!
-
Thanks Jane!
I had 4 premolars removed (2 upper, 2 lower) and a fixed brace to correct minor overcrowding, which also gave me a more projecting nose and smaller jaw. Not sure it was worth it! ;-) -
"incompetent lips" is one of my favourite phrases :-)
-
Dr Christopher Rynn (Educator) replied to [Learner left FutureLearn]
Spot on Joanne!
-
Fantastic! :-D
-
Elvis is indeed the best example of LLSAN contacting almost in isolation!
:-S -
It took me years to get my head around it ;-)
-
Hi Patricia,
You mean about the ears?
So, the shape of the tip of the mastoid process can indicate either lobed or adherent ears.
Also, the upper part of the mastoid process and the supramastoid crest above it (basically the bone behind the ear) can bulge out, which indicates more protrusive ears by proxy. Although, ears can also be protrusive in and of... -
"Psychotic Archie" - Love it! :-D
-
Thanks Snezana!
Well done indeed! :-D -
Thanks Orla!
Facial reconstruction isn't really a legal form of identification in and of itself: it's more to generate new leads for an investigation, so that hopefully an identification could be later confirmed using other methods. -
...pretty much.
Except for the facial asymmetry, which is theoretically mostly due to environmental factors. Major things, like physical trauma, but even minor things like repeated facial expressions over the years, or on which side of the face you favour chewing, and maybe even on which side you sleep! -
Thanks Emma,
Healed trauma and surgery can be taken into account if apparent on the bone. It's a bit like using the published methods to estimate the nose before the break, and then more subjectively estimating and depicting how it might have appeared afterwards. -
Thanks Sarah and Maggie!
Adherent ears (no lobe) are genetically recessive and there is about a 10% chance of having them. In some quite rare cases, people have one lobed ear and the other adherent. The mastoid process is shaped a bit differently, with the tip pointing more downwards in adherent ears, rather than forwards. -
Good point Wendy!
I don't think that would show up on the bone. -
Thanks Dan,
I'll discuss some of the sexual dimorphism in the skull and how this pertains to the face in the hangout. -
Thanks Peter and Lorna,
The relationship between the nose and skull was a major part of my PhD research.
Rynn C; Wilkinson CM; Peters HL (2010) Prediction of nasal morphology from the skull. Forensic Science, Medicine and Pathology, Vol. 6, No. 1, 03.2010, p. 20-34. -
Hi Walter,
It can be done from a CT scan of a skull, but there are some minor surface details which can't be seen other than on the actual skull, so it's always better to have the real skull for reference.
I'll discuss the automation of the process in the hangout... -
Named after its function: "Buccinator" comes from the Latin "Buccinare" meaning "to blow a trumpet". Actually, a specific type of trumpet called a buccina!
-
There are a lot more too, but the ones which don't give shape to the face aren't sculpted on a facial reconstruction.
-
If you speak Latin and Greek, Anatomy is a lot easier! Things tend to be named based on function, location or appearance.
Can anyone deduce the function of the muscle levator labii superioris alaeque nasi?Despite the long name, it's so small and thin that there is no real need to apply it in a facial reconstruction!
-
Hi Emily,
...a combination of plaster-casting and manual clay or wax sculpture. It took a lot longer to cast the skull before and the face after. Plus, there was always the risk of damaging the more fragile specimens. -
Thanks John!
I'll talk about the automation of facial reconstruction in the hangout... -
I'll do a wee demo in the hangout, to show how the sculpting tool offers physical resistance when you "touch" the onscreen virtual model...
-
Thanks Josephine,
It's a lot quicker using 3D scans than manual casting and sculpting, plus we don't risk damaging the skull this way. The actual sculpting itself takes about the same time, but the hours/days of prep and post work (such as casting the skull before and the face afterwards) is reduced or removed.I'll chat more about the training and...
-
Hi Clay,
I agree absolutely. We teach anatomy to our MSc Medical and Forensic Art group using a combination of traditional anatomical teaching methods (dissection, specimen study, textbooks) plus sculpting anatomy in wax on plastic skulls. -
Thanks Graham!
There have been lots of accuracy studies over the years, and we've carried out a few using these facial reconstruction methods on CT data of living individuals.
The most recent couple are:Won-Joon Lee, Caroline M. Wilkinson and Hyeon-Shik Hwang (2012)
An Accuracy Assessment of Forensic Computerized Facial Reconstruction Employing... -
Thanks Vivienne!
I personally started out as an Anatomist, who had been previously talked out of studying Art at college, by a careers advisor who maintained that there was no such job which entailed both Science and Art(!) -
Thanks Eileen!
I'll use this as a talking point in the hangout... -
Thanks Geraldina!
It wasn't anywhere near as fast as it looks on the video, though!
-
It's a shame I can only "like" this comment once!
Thanks Pam! -
Thanks Heather,
Absolutely - we can work with a CT scan now, rather than having to cast a skull in plaster and risk damage. We still sculpt physical heads to order, for museum pieces, etc. but these might be on a physical 3D print from a laser scan or CT scan of the skull, so no risk of damage to the original specimen.Also, we can hit CTRL Z to undo!
-
Thanks Sarah!
The temporalis muscle fills the temporal fossa so that it continues that overall curve of the cranium, from the temporal line to the zygomatic arch. You can only really see a dent at the temple of emaciated people, and conversely, when overdeveloped, the muscle would not bulge noticeably out of the side of the head, but rather the temporal... -
Hi Caroline,
Yes, quite a few studies, some very recent, but a few going back a century!
I'll go into more detail in the hangout. -
Hi Asit,
With difficulty! If it is on one side, then the other side can be mirrored and adjusted to fit the asymmetry as best as can be estimated.
If it is the midface, then we have a problem... -
Thanks Jane,
Will make sure I do that, -
Hi Blanka,
There is a wide range of diseases which could affect facial appearance, many of which would be apparent on the bone, but many of which would not. I would always defer to a pathologist regarding disease of the soft tissues.
One example of something which affects the facial shape, but would not be apparent on the bone (and which hindered an albeit... -
Hi Sean,
Great question! Just as with sex, if the face appears to be of mixed or ambiguous Ancestry, then the skull will be ambiguous in a similar way.
I am in favour of moving towards an overall method which can be applied to any skull, regardless of sex/race label, but we are working from international research studies spanning back over a century, so... -
Hi Nicholas,
I would say that it's a process which uses skull shape to estimate face shape. Colour is another variable altogether. There are obviously some correlations, such as one would expect a person with, e.g. a markedly East Asian skull shape to have dark eyes and hair.On a side note which you might find interesting, a friend of mine (who is your...
-
Thanks Jan!
The range-of-potential-faces approach is one which has been used in the past, but there are some issues about how far you take it: I mean, in theory, all of the variables which are inestimable from the skull would have to be presented as a range, so there might be blonde, brunette and redheaded versions of a thinner, an average and a heavier face,... -
Hi Nicholas,
Huge fan, yes! I was lucky enough to visit the lab in Moscow in 2009, where I met Gerasimov's academic descendants, Elizaveta Veselovskaya and the late Tatiana Balueva. We subsequently became friends and co-authored a book chapter together.That's a profound question, though... I might have to dwell on that and get back to you in the hangout!
-
Hi Michele,
In the skull assessment video, it shows the nasal bones as being above the nasal aperture at the bridge of the nose, so they give rise to the line which is pointing down the length of the nose.
The anterior nasal spine is the small pointed bony process below the nasal aperture, which gives rise to the lower tangent (line) which tells us the... -
Thanks Fernando,
Excellent work! -
Dr Christopher Rynn (Educator) made a comment
Thanks for the comments everyone!
I'm so glad you're enjoying it, and I'm really impressed with your analyses. -
Hi Jeanette,
"Upturned" eyes just means that the outer corner is higher than the inner corner.
Chris -
Thanks George,
True, but it's an appeal to friends and family of the missing person, who would have known them for a long time, probably both bearded and beardless.
Another potential source of error specific to unidentified male adults is the hairline, which could be anywhere or nowhere! We tend to crop the top of the head off the image in these cases,... -
Thanks Helle!
Did you know that the character in Gorky Park was based on Mikhail M Gerasimov (1907-1970): the head of the Russian school of facial reconstruction?
So the character in the novel would have used the Russian technique, but when it was made into a movie, the actor (Ian McDiarmid) used the American technique!
