One day in 2006, Sue Black, a professor at the University of Dundee’s department of Anatomy and Forensic Anthropology, received a phone call from a man called Nick Marsh. He was a forensic photographer who had worked with Black 17 years earlier as part of a team sent by the Foreign Office to examine the bodies of victims of war crimes in Kosovo. Marsh knew that Black had a talent for identifying people from scraps of flesh and bone. Now he had evidence of a different kind and wondered if she could help.
In 2014, 43 students were massacred. Can digital forensics help solve the crime?
The piece of evidence was an eight-second-long digital video clip. Marsh had been working on a case involving a teenage girl who had alleged that her father had been coming into her bedroom at night to molest her. When her mother refused to believe her, the girl left her webcam running all night, pointed at her bed. The camera captured a person’s hand and forearm touching her. Her father denied that he was the person in the video. “It was one of the spookiest and scariest things that I have ever seen,” explains Black. “A real sort of horror movie.”
Marsh asked Black if there was a way to identify the perpetrator. She didn’t have clue. “I’d never done anything like that before. I’d never identified anyone using a hand,” she says. But after studying the footage, Black noticed something that had escaped her before: the veins on the back of the man’s hand were visible. In the dark, the camera had reverted to infrared mode, and in those conditions the deoxygenated blood in veins shows up as black lines. Black, an expert in anatomy, knew that hand-vein patterns are unique from person to person, even in identical twins. She asked the police to take photographs of the father’s hand and forearm. The vein patterns matched.
Black appeared in court as an expert witness for the prosecution, presenting her vein-pattern analysis. It was the first time in British legal history that evidence of this kind was presented in court proceedings. When she was introduced, the judge had to stop the trial for 90 minutes to ask her to explain the principles behind her analysis. Black explained her rationale, but conceded that she didn’t have statistics showing the likelihood of the hands matching. “That research had never been done. I could say no more than everything matched, and we couldn’t say it definitely wasn’t him,” she says.
Still, it was strong evidence and the prosecuting barrister expected the father to be found guilty. However, he was acquitted.
“I asked the barrister if there was something we had done wrong or something in the science that I had not been able to convey,” Black recalls. “She said, ‘No, there was no problem with the science. The jury had just not believed the girl. They thought she didn’t seem upset enough.'”Black was dumbfounded.
Shortly after the trial of the girl’s father, the Serious Organised Crime Agency (SOCA) asked Black if she could help with an ongoing police investigation called Operation Ore. It was a long-running investigation of more than 7,000 British people suspected of downloading indecent images, after the FBI had found their details on the database of a child-porn distributor in Texas.
The operation became the UK’s largest-ever computer crime investigation, involving the arrest of more than 3,700 people, including several public figures such as The Who guitarist Pete Townshend and The Thick of It actor and writer Chris Langham.
Black was again asked if she could identify people in the images. “Operation Ore was the first time I realised these kind of cases could have such a volume,” she says. “I was naive. I thought it was all about isolated people in isolated cases.” According to Black, about a million images of child abuse are uploaded to the dark web every day. When police seize mobile phones and find indecent images, they discover, on average, about 100,000 individual images. “It is a huge problem, and the police can’t get near looking at them all, nor arresting their way out of it,” Black says.
In the end, she worked only briefly as a consultant on Operation Ore, which soon became mired in controversy when journalists revealed flaws in police methods.
Nevertheless, it was a turning point for Black. During Operation Ore, she became fully aware of a problem that she didn’t realise existed and that she might be the person who could do something about it.
But in the months after the trial, it occurred to her that she might have stumbled across a new idea. Marsh had mentioned that the police were seeing an ever-increasing number of indecent images and videos of children. Abusers often appeared themselves: “Sexual abuse of children is often about power, and the touching is a part of that,” says Black. “When a perpetrator views an image of themselves abusing a child, they are reliving the enactment. If there’s a part of them present in the image, it gives them an extra feeling of involvement.”
The problem was that, in most cases, the only visible parts of the abusers’ bodies were their hands and genitalia. Previously it had been widely assumed that such evidence was not enough to incriminate someone. But Black was unconvinced. “There was a research route that had never been fully explored,” she says. “I had been involved in crimes where the victim was dead but these cases had live victims and perpetrators. I thought there might be something we could extract from those images and use in a meaningful way. I thought, ‘We should be researching it.'”
Sitting in her 70s office with its high windows to let in light, Black looks very much the academic in a cardigan, with her plaited. Her manner is no-nonsense but affable.
Black grew up the youngest of two daughters in a blue-collar Inverness household, and was the first of her family to attend university – she studied biology then human anatomy at Aberdeen. She began her career teaching at St Thomas’ Hospital in London. Stints of body-identification work for the police, then the Foreign Office, led to her working in Kosovo, for which she was awarded an OBE in 2001. She has since worked in conflict zones in countries such as Iraq, and in Thailand after the 2004 tsunami.
In 2003, Black took over the University of Dundee’s Centre for Anatomy and Human Identification and began developing the links between anatomy and forensic science. In 2016, in recognition of her services to forensic anthropology, she was made a Dame.
The teams that work on forensic cases are, Black says, “very close knit. At the end [of a case] we will sit and talk it through. Counselling is always available, though we haven’t needed it yet. We are very tuned in to each other, and if someone is uncomfortable we deal with it there and then. When a team is exposed to this sort of thing, which is as bad as it gets, each of you has to know that each member is not suffering themselves.”
After Operation Ore, Black realised that hand analysis would be taken seriously only if it had a genuinely scientific foundation, rather than being based on ad hoc comparisons. It was fine to show the vein patterns of an abuser and the accused matched, but if the accused contended that many people had matching veins, Black wouldn’t be able to back up her argument with any scientifically validated evidence. In other words, she would need a substantial database of hundreds of people, compiled with a minuscule budget.
False memories and false confessions: the psychology of imagined crimes
In April 2007, Black’s department won a contract to teach more than 550 police officers, coroners and legal officials about disaster-victim identification. Black asked attendees to have photos taken of their hands, forearms, feet and legs. Most agreed.
More recently, she helped a mother’s fight to prove what had happened to her son’s body. She carried out an exhumation at his burial plot in Edinburgh, and confirmed that no humans reminds were inside the coffin.
By 2008, she’d published a study confirming the validity of vein-pattern analysis. Shortly after, she was asked to help in another case. The defendants were eight men belonging to Scotland’s largest-known paedophile network. For years, they’d colluded to rape and sexually abuse children, and shared at least 125,000 images of the abuse. The details were so harrowing that during the hearing, the public and media were barred from seeing the images, and counsellors were made available. At one point, the jury heard that one of the accused had circulated a request for “porn with young Down’s syndrome or learning-difficulty kids”.
Many images featured men abusing the children of friends. A key photo – later known as “the Hogmanay image” – showed one of the two ringleaders, Neil Strachan, 41, attempting to rape an 18-month-old boy whom he was babysitting on New Year’s Eve in 2005. The only parts of Strachan’s body visible were his penis and left hand. It was this image that Black analysed.
She was aided by a mistake on Strachan’s part. His defence team ordered that photographs be taken of his thighs, their intention presumably to show that body parts could not be used to identify someone. However, when the photographer was taking the picture, he asked his subject to hold the photographic scale, which, says Black, “gave us a beautiful view of the accused’s thumbs”.
Black compared the left thumb in the picture with the Hogmanay image and found matching details, including an unusually shaped lunule, the white area at the base of the nail. “This time, I was able to go back to my database and put statistics to the data.” In October 2009, Strachan was sentenced to life imprisonment with a minimum term of 16 years, cut on appeal to nine years.
Hands can be used to verify a person’s identity in two ways. First, they pick up marks and injuries – more than 20 per cent of people attending A&E in the UK have hand injuries. Second, it has in-built morphological features which are unique to the individual – fingertip whorls, palm prints and vein patterns. When a body is growing inside the womb, cells assemble spontaneously, rather than following a pre-established blueprint. This means vein patterns are one-offs. Veins also have the advantage of being enclosed by skin and, unlike fingerprints, can’t be altered.
Black analyses mainly the backs, or the dorsum, of hands, as these tend to be predominantly visible in the footage she works with in criminal cases. She first maps a grid of 24 cells on to the hand, covering everything from fingernails to wrist. Then she analyses each cell, looking for identifying marks and studying vein patterns, drawing dark lines over them on-screen to make them more visible. The features she most commonly checks are veins, scars, freckles, birthmarks, moles, nails and skin creases on knuckles. Each one is scrutinised. For example, scars will be classified according to whether they are linear or non-linear, or surgical or accidental, and then by the direction in which they run. When she compares the accused’s hand with the database, she can use geometrical formulae to work out the chances of anyone else having the same markings and vein patterns.
Black’s database – she has now analysed 1,000 hands – throws up fascinating insights. For instance, you are most likely to get a linear scar on the tip of your second finger, or the middle of the back of your hand. No one seems to get moles on their little fingers, and if you have moles in the same places on both hands, it will be somewhere in the lower half of a triangle drawn between the knobs of your wrists and second knuckle.
On average, men have 50 per cent more scars than women, but right-handed men are more likely to scar their left hands, while right-handed women tend to scar their right – no one knows why. Black is fascinated by the stories that the hands in her database tell. One of her papers quotes lines from Arthur Conan Doyle’s A Study in Scarlet: “By a man’s finger-nails, by his coat-sleeve, by his boots, by his trouser-knees,” declares Sherlock Holmes, “by the callosities of his forefinger and thumb, by his expression, by his shirt-cuff – by each of these things a man’s calling is plainly revealed.”
Sometimes a case challenges Black’s methodology. In 2014, the Greater Manchester Police asked her to work on the case of paedophile Jeremy Oketch, a 30-year-old pharmacist who had twice raped a two-year-old girl and filmed the assaults. Although it was impossible to prove, the child’s silent compliance suggested that she had been drugged. And although the police had 55 minutes of footage to examine, the only visible parts of the rapist were a hand and his penis.
The video was so distressing, recalls Black, that when judge Hilary Manley left the courtroom to view it, she returned visibly shaken. Was Black affected herself? “Images of child abuse affect everyone who views them,” she says. “I feel anxious watching video because you don’t know what’s coming next. But you have to stay objective. It’s not my place to go back to analyse the incident, it’s my job to find something of value to the investigation.”
The Oketch case presented her with two technical problems. First, he was black, “and all the people we had looked at previously had been white. I didn’t know if all the features would be as visible on black skin, but they were.” Second, a lot of the footage was clear, the matches were numerous and potential divergences almost totally absent. That sounds ideal, but such apparent certainty brings its own risks. Black takes a file from a cabinet and slips out her report on Oketch to show me (it is in the public domain, having been used in a Crown prosecution). Information is tabulated. Under “Hand” appears a long list of features: “Hand morphology”, “Thumb nail groove from asymmetrical lunule”, “Vein pattern” and so on. Under “Penis”, a similar list: “Penile morphology”, “Vein pattern”, “Lateral deviation”. Each feature is marked to show whether it’s the same on the rapist and the suspect. They all are. “And as I learned, that can be a challenge, because it makes you ask yourself if you’re really seeing everything. Part of this work is knowing how to look; asking yourself what you might not be noticing,” Black says.
In the end, the match appeared strong. When presented with Black’s report, Oketch changed his plea from not guilty to guilty; he got 15 years. That plea change was important, Black says. It meant money that would otherwise have been spent on trials was saved. It also meant the child was spared from having to give evidence in court.
Black’s team helps police forces around the world – including the FBI, Interpol and Europol – and works on 30 to 50 cases a year. In the cases Black has worked on since 2006, the percentage in which the accused have changed their plea to guilty in response to her analysis stands at 82. Black also takes on cases related to circumstances such as those in which the perpetrator has disguised their face. Grants have helped expand the database and her team have reduced the time it takes to compile a report.
When a case comes in from the police, Black administrates the project, but the client pays the university; any payment to Black’s team could be seen to compromise its objectivity. Images or video material are delivered on encrypted drives and handed to her in person. Black works in a team of three but she first views all video evidence herself, absorbing the initial shock on behalf of colleagues. “You have to view it all the first time to know what’s coming,” she explains. “Then you can narrow it down and look at parts that are more important for the job you have to do.”
After that, she shares material she thinks is important with Lucina Hackman, a senior lecturer in human identification at the department, and both women independently single out the pictures that best highlight key anatomical features. Then they agree about the offender’s important features and a photographic specialist on the team, Chris Rynn, will enhance the images digitally. Once they have established the offender’s features, they study images of the suspect, trying to establish a match.
Roughly speaking, the degree of certainty on any biometric is dictated by the size of a data set. Black’s is not yet big enough to justify stating a statistical probability, so instead she follows the system used by the judiciary, which objectively grades the possibility of a match.
Even with clear images of a suspect’s and perpetrator’s hands, it is impossible to scientifically guarantee a match, as that depends on all the anatomical features present. A suspect can be excluded with 100 per cent certainty, but a match can only carry a grade of “strong support” that the suspect and the offender are the same person. This equates to between a 1-in-1,000 to 1-in-10,000 chance that it could be someone else.
Often this is enough for the accused to change their plea as there is normally additional evidence to implicate the person. If you’re wondering why no one is investing billions to create million-strong data sets, Black says it’s because there’s no money for research into catching child abusers. In the forensic field, most research funding goes into DNA, because it’s what they know and trust and there’s a drive to do things quicker and cheaper.
“We’ve looked at vein patterns on the right and left hands of all individuals on the database and we haven’t been able to find any two that match,” Black says. “We have expanded the database many times since we began, but we need much bigger databases to establish greater degrees of certainty. We think we might get to something that’s as good as fingerprinting.” Black is attempting to automate the process of searching for repeated patterns, creating algorithms that are able to extract the features from millions of stills or video images. “We’ve done the pilot project, which shows that we can extract vein patterns and pigment patterns. We’re now looking at whether we can do skin-crease patterns on knuckles,” Black says. “When you layer all these features and patterns, you increase the probability of identifying the right individual to the fingerprint level, or even perhaps the DNA level of certainty. It could allow us to identify and look for the first-generation producers. It would also mean reducing the strain that these images places on officers. They take a terrible toll.”
When asked about the possibility that, as forensic hand analysis becomes more common, paedophiles will start wearing gloves, Black is adamant: “They won’t. Most people who commit crimes aren’t very bright. They think they’ll never get caught.”
Case study: Dean Lewis Hardy
During a trip to Thailand in 2004, Kent-based Dean Lewis Hardy took indecent photos of four girls aged eight to ten years old, including images of his hand touching them. Five years later, he was found guilty of indecent assault after being identified through an analysis of the images of his hands. He received a six-year sentence. Prosecutors said it was the first case to use hand analysis. Black found Hardy’s scars matched that of the suspect, along with his freckle pattern and thumb skin creases. “Scars and creases are accidental,” Black explains. “Freckle patterns are random, but their presence indicates a genetic predisposition to freckle formation. Therefore, we had features of different aetiology.”
The left index finger of the offender is on the right, and that of the suspect (Dean Lewis Hardy) on the left. It highlights the freckles and a four-point punctuated scar.
The index finger of Hardy is on the right and the offender on the left. A filter has been used to make the freckles more obvious, then grouped into patterns that can be compared between the suspect and the offender.
Both images feature the thumbs of the suspect. The creases of the skin, nails and lunule – the crescent-shaped marking – have been outlined to assist the comparison with the images of the offender.
In June 2016, Black was asked by Kent Police to work on the case against Richard Huckle, one of the worst predatory paedophiles in British history. Between 2006 and 2014, Huckle had groomed and abused up to 200 Malaysian children, including babies, in Kuala Lumpur, while masquerading as an English teacher and philanthropist. Images and videos of his rapes and assaults had been shared with paedophiles on the dark web.
In December 2014, National Crime Agency officers arrested him when he arrived at Gatwick Airport to spend Christmas with his parents, and found 20,000 indecent pictures and videos on his laptop. Officers from the NCA’s Child Exploitation and Online Protection (CEOP) division viewed every picture and film clip. The material was deeply disturbing: although 23 children would be identified in the charges, the number of victims was believed to be far higher because detectives found on his computer a ledger on which he awarded himself “pedopoints” for 15 levels of abuse rated from “basic” to “hardcore”. He had also compiled a 60-page manual, “Paedophiles and Poverty: Child Love Guide”, which focused on selecting deprived victims without being caught, and was found on his laptop. He’d planned to publish it online and wanted to create a paedophile wiki guide. “I’d hit the jackpot,” he wrote, “in a three-year-old girl as loyal to me as my dog, and nobody seemed to care.”
When Black analyses the backs of hands in footage she maps a grid of 24 cells, then looks for identifying marks and highlights in the vein patterns
CEOP officers selected material they felt was clearest and passed it to Black. “Some of it was quite old, so it was degraded, but we didn’t need to study that,” she says. “Advances in camera technology mean that paedophiles are taking clearer pictures these days. It can make them easier to identify.” Even looking at this selection took her team a long time. “It took us about four days to view it all, seeing what we could use, isolating the parts to be used.”
In the end, Black’s team were able to present evidence to show that Huckle was likely to be the perpetrator, and as the evidence mounted against him, as with Oketch, he changed his plea to guilty. This resulted in the conviction of a man who judge
Peter Rook QC said had almost certainly blighted the lives of his
victims and caused them severe psychological harm.
“The significant thing about that case was the scale of the sentencing”, Black explains. “He was given 22 life sentences for 71 offences, which was a way of the courts saying, ‘We are serious about this, we are not going to take it lying down.'” Black doesn’t dwell on the horrors of individual cases but prefers to talk about what can be done to stem the sharing of child-abuse images online. “Can’t our phones recognise parts of a body and stop the image being taken?” she asks. “That’s the challenge I want companies such as Apple to take up, to stop technology being a mechanism by which our children’s innocence is being stolen. Because, you know, the statistics say that one in six people have had unwanted sexual attention as a child. One in six. I cannot think of a crime that is more important. Can you?”
Richard Benson is a London-based journalist and author. He wrote about Asem Hasna’s 3D-printed prosthetics for refugees in issue 07.17