EVA 2007 London Conference, 11-13 July

RESURRECTING THE DIGITISED BODY :

THE USE OF THE ‘SCANNED IN’ BODY FOR MAKING ARTWORKS.

 

Marilène Oliver
Printmaking
Royal College of Art
Kensington Gore
London SW7 2EU
United Kingdom
marilene.oliver@rca.ac.uk   
http://www.marilene.co.uk

Abstract – This paper focuses on the scanning technologies of MRI, CT and PET scanning, explaining how and why I use them to create artworks. The paper concentrates on three works, each of which has been inspired by and made from scans; Family  Portrait (a series of sculptures made from MRI scans) Dervish (a sculpture made from CT scans) and Radiant (a sculpture made from PET scans). As well as focusing on my art practice and the primary experience of working with DICOM (Digital Imaging and Communications in Medicine) data, I draw on recent experience of shadowing radiographers at the Paul Strickland Scanner Centre (PSSC).

 

INTRODUCTION

In 2001 I made a sculpture called I Know You Inside Out. I Know You Inside Out is a reconstruction of a 39-year old convicted murderer named Joseph Paul Jernigan, who, prior to his execution, was persuaded to donate his body to medical science in order to become the ‘Visible Human’, a dataset of cryosections, CT scans and MRI scans. My desire to create a sculpture of Jernigan was neither anatomical nor medical – I was fascinated by the virtuality of the Visible Human – in becoming ‘visible’, Jernigan’s body was converted from flesh to voxel: in order to create the dataset Jernigan’s corpse was frozen and sliced so finely that it disintegrated to mush, leaving only digital photographs and scans. The images of his body where uploaded onto the Internet allowing him to be viewed at anytime and any place (but never all at once), he was under constant threat of being copied or translated. I downloaded images of his body and printed them onto sheets of acrylic and then ‘put him back together again’.
After making I Know You Inside Out I was inspired to preserve as well as resurrect so sought out technologies that offered images of the inside of the living body. I soon found that the photographs of cryosections of Jernigan’s body, which I used to make my sculpture, were mimicking the axial (cross section) scan of MRI and CT scanning. After some initial research into the different scanning methods, and seeing that a dataset similar to the Visible Human could be created harmlessly with MRI, I embarked on making Family Portrait – a series of sculptures of my family made from MRI scans. In making Family Portrait my engagement with medical imaging was firmly established and I have since worked with a number of different scanning technologies to create artworks including CT (After and Iceman) and PET scanning (Radiant).
MRI, CT and PET essentially create copies of our bodies, which unlike their originals are able to inhabit virtual space and become part of it. The digital copies of bodies differ depending on the processes they have had to endure in order render them virtual. I contrast and compare the different datasets within the confines of both of conventional and unconventional software. I allow the data to move freely from medical imaging to image editing to computer aided design to data processing software. The ‘Scanned-In’ body is manipulated and processed until the body bears, as a mother bears stretch marks, scars of the virtual world. Ultimately, my major and quantitative research aim is to create new and bespoke methods for exporting the processed data as figurative sculpture in order to allow us to contemplate these copies of ourselves in an actual and physical reality where we can consider how they  affect our sense of self both physically and digitally.
This paper will look at the different processes involved in MRI, CT and PET through the making of three artworks: Family Portrait (made from MRI scans), Dervish (made from CT scans) and Radiant (made from PET scans). The paper will be both technical and theoretically speculative in that as well as explaining how the works were made it will also flag up possible implications of the work.

FAMILY PORTRAIT – MRI (Magnetic Resonance Imaging)

Family Portrait (Fig. 1)is a series of sculptures of each of my family members – my father, my mother, my sister and myself. Each sculpture is a stack of 90 sheets of clear acrylic onto which an MRI scan has been silk screen printed.  The scans when printed and stacked in correct order give the illusion of a ghostly figure, which appears and disappears depending on your view point.

MRI is a non-invasive scanning technology, which works at an atomic level in the body in order to create images with very high soft tissue definition. Generally the two extremes of contrast in MRI are fat and water. As it’s name suggests, MRI  involves altering the magnetic field surrounding the body so the patient / subject has to go inside a large magnetic coil. The scanner works by applying a hydrogen specific radio frequency pulse through the coil. The pulse causes the protons in the part of the body being scanned to absorb energy, which makes them spin, or precess, in a different direction (called resonance). When the radio frequency pulse is turned off, the hydrogen protons begin to slowly return to their natural alignment within the magnetic field and release their excess stored energy. When they do this, they give off a signal that the magnetic coil picks up and sends to the computer system. What the system receives is mathematical data that is converted into a grayscale image. Depending on when the computer hears back from the unwinding protons the image either shows fat as dark and water bright (T1 recovery) or fat as bright and water as dark (T2 relaxation).
In clinical scans patients normally only have small sections of their body scanned at a time, such as their brain or knee.  For Family Portrait I needed a set of full body cross sections scans at 20mm intervals which was a challenge for the radiographers as it meant they had to take four sets of scans and make sure they overlapped correctly. It was equally challenging for us being scanned as it meant the scans were very long - my mother, the first to be scanned was in the scanner for two hours. Although the scan was long, it was not an altogether uncomfortable experience. The radiographers both at Queen’s Medical Centre in Nottingham where we had the Family Portrait scans done, and at PSSC where I have recently been shadowing radiographers did everything they could to ensure the patient / subject was in comfortable inside the scanner. At PSSC they ask patients to bring a CD of music to listen to whilst inside the scanner (to disguise the noisy clanging of the magnets) and there is a mirror above their head reflecting a picture of dolphins for them to meditate on.
The reason the process takes so long is not only because the scanning takes a long time (although now scanners have stronger magnets so scans can be acquired faster) but also because all scans are bespoke – the radiographer alters the scan parameters to each patient/subject.  Like with flat bed scanners the MRI scanner first takes a ‘preview’ or ‘scout’ scan which allows the radiographer to scan the exact area needed and at the correct angles.

Fig. 1 Family Portrait 2003, silk screen on acrylic, each sculpture 50x70c192cm

A few weeks after the scanning I was able to download the sets of scans from the radiologist’s website. For each of my family members I have approximately ninety images or slices. In order to be able to print these images onto acrylic sheets and then reassemble my family I first had to translate them using the computer.  For each family member I worked with screen based slices of their bodies for over a week. I imported them into Photoshop, sized them back to true body size, altered contrast levels so that each body was equal and isolated them from the black background.  I then had to prepare them for screenprinting by applying a halftone to each image. After having translated my family’s slices I then screen printed each slice of my family onto an individual sheet of acrylic. The whole process took approximately two months for each figure and throughout my family’s bodies were fragmented and either computer or silk screen based. Only after more than half a year of screen printing, and then a three day long installation process was I able to ‘put them back together’ again as life size, standing, space occupying figures.
My parents divorced twelve years before I made Family Portrait and at the time of the scanning my sister lived in Beijing, my mother in France, my father in Essex whilst I was living in London. My father often used to express sadness at not having a portrait of ‘our family’. This in turn made me sad, I knew that both politically and geographically it would be impossible for us to have a family portrait made. On reflection I realise I made Family Portrait (somewhat clandestinely) to satisfy this sadness.  Although what you see when you stand in front of Family Portrait is a row of reconstructions of my family members which operate as a ‘family portrait’ we were never together in its making. We all entered the scanner individually and at different times. My mother stretched out her hands in the hope ours would join hers. Through digital processes and my artistic invention it may seem we are a family, but in reality it is only the digital copies of our bodies that star in the portrait.
Digital media has the intrinsic ability to perform tricks seamlessly; in a relatively short space of time we have come to accept that we can never believe what we see. From the model on a cover of a magazine to a feature length Hollywood film the computer splices, layers and fuses images (both moving and still) together with no trace of a cut mark. Family Portrait also boasts these characteristics both in its content and in its visual effect. The bodies are hollow, they have 20mm gaps in the them yet from many angles they look whole and complete. Our eyes accept the trickery very easily, I have overheard viewers saying the Family Portrait bodies are made of smoke and it is most common for them to notice the gaps last and not first.
Family Portrait sculptures have in their intention not only a desire to unite our fractured family but also to preserve it for eternity. Ancient Egyptians believed that if they preserved the corpse its soul would return and find it. Cryo preservation technology claims to ‘suspend’ the body until a future when technology will find a way to bring the body back to life.  In Weird Science [1] two teenage boys, Wyatt and Garth, create their perfect woman by scanning in various images (such as Kelly Le Brok’s face and Einstein’s brain) and connecting to the mainframe of American computing. If Wyatt and Garth can do it by scanning in images of the exterior of the body, surely images of the inside of the body would be even better.  If MRI is able to convert the body to data, how more far fetched than cryo preservation is the possibility that a future technology might be able to convert the data back to body. Converted to data, the body is freed from its physicality, its aging, its vulnerability. The data body is eternal.
      

DERVISH – CT (Computer Tomography)

Unlike MRI, CT scans cannot be performed for non-clinical research as the body is exposed to radiation during the scan. CT uses the same technology as the x-ray, a beam of light with a very small wavelength (smaller than UV and invisible to the eye). When x-rays travel through the body they are absorbed by the tissue and bone (different kinds of atoms); the harder the tissue (the bigger the atom) the more is absorbed. In a CT scanner an x-ray tube emits a fan of x-ray beams. This x-ray tube rotates around the subject of the scan. Outside the rotation sphere of the beam is a fixed ring of detectors that measures how many of the x-rays are being absorbed by the body. Rather than spin around the body, CT scanners now spiral around the body allowing dense and continuous datasets to be acquired. In recent years there has been a rapid development in CT scanners and the latest 64 slice machines boast being able to scan at sub millimetre intervals.
CT scans are much faster to acquire than MRI scans and as a result many more people can be scanned in a day. Although this is a positive for NHS targets it does not make for a pleasant scanning environment. Whereas in MRI the radiographer has the time to ensure the patient is comfortable and relaxed, the radiographer in CT needs to make sure they scan their patients as quickly as possible so as not to get behind on their long lists of patients. Because CT takes such high-density information of the body there is no need for the radiographer to spend too much time setting up the scans – they just need to make sure the body is in the scanning range and then they perform the scan, which takes approximately two minutes. The data is post processed in order to give the radiologist/consultant the images of the body they want. To make organs more visible almost all patients are asked to drink a contrast agent before they have the scan as well as an injection of iodine during the scan.
As it is not possible for me to commission a CT scan I have had to look for existing CT datasets to work with.  Recently I have started to work with medical imaging software called OsiriX (a free download for Macintosh), which provides a number of datasets. Having worked primarily with single MRI images to make sculptures I have found the full body CT datasets astonishing. For Family Portrait I was supplied with jpeg versions of scans meaning I was unable to use the scans for any other purpose than the one I had originally intended. With the CT datasets I have downloaded from OsiriX I can slice the body in any direction I chose and as many times I chose. I can also create 3D volumes which can be animated and export 3D surfaces to be rapid prototyped. Although much of these abilities are thanks to the OsiriX software it is only with high resolution CT that its functions are really able to perform. I have put MRI datasets through the same processes but because the data isn’t so dense the images are distorted and inconsistent.
After observing the CT scanning at the PSSC I felt I really wanted to test the CT data to see what it could do, so I spent a lot of time working with a CT dataset of a full body which has been anonymised and called MELANIX. MELANIX is a woman who was scanned with her arms above her head. I do not know who she is, where she is or when she was scanned. MELANIX has been severed from her original physical body and now exists, much like Jernigan the visible human, waiting to be manipulated, sliced, rendered and surfaced.
One tool in OsiriX that I have found particularly engaging is 2D Multi-Planar Reconstruction. 2D Multi-Planar Reconstruction allows you to choose axis points in the body to pivot around. This function allows the body to spin around different points; it can pivot around its nose, its heart, its stomach, its brain. When working with this function I started to think of other times I had experienced this type of vision or encounter of information and it struck me that is was very much a disembodied vision:  the 360 degree pivot function exists in computer games, in 3D modeling software, even when navigating the Internet we choose (normally via Google) our point and then look around. Then we jump to another pivot point and look around that point. In the increasingly popular, and many would argue significant, Second Life [2] (an increasingly popular 3D virtual world users can interact with and help create) it is the same – avatars (your on screen double which you create by modifying one of a number of standard bodies) can teleport to a location and then spin around before deciding whether to stay and enjoy some shopping, conversation dancing or even cybersex before teleporting (via a text entry box) to a different point in the virtual world.
I believe the 2D reconstruction tool highlights an emerging way of viewing and encountering information. The Panoptic allows us to see all from a hierarchical vantage point. It is a God’s (or Big Brother) eye view of the world. This new structure (which I shall call the ‘Pivotoptic’) allows us to enter the information/world at ground level and spin around inside it. Laws of perspective are replaced by density of information around the axis. The more we pivot, the faster we spin, the more information we acquire (and also the dizzier we get). The Panoptic is not relevant in virtual space – it is impossible to get an overview of data, the only way to find the information we want is to go inside it. It is impossible to stand outside of the Internet and look in – the obstructions and distractions are too dense. The same is true of the DICOM body – a radiologist has go inside the body to see what they want to see, from outside looking in there is too much other information blocking the view. The Pivotoptic is establishing itself in the real world too: Merseyside Police’s anti-social behaviour taskforce AXIS (even the name enforces its alliance), supported by the Home Office will launch drones [3], first used in military reconnaissance missions into the air in June 2007. Drones are little helicopter like devices, which can be remotely controlled to go into areas that normal cameras and police officers couldn’t or wouldn’t go into.

Fig. 3 Dervishes 2007, inkjet on film and acrylic, each sculpture 50x30x200cm.

Dervishes (Fig.3)is a series of five sculptures made using the Multi-planar Reconstruction tool. I made Dervishes in an attempt to expose this new vision and to offer a new body politic. As explained briefly above, the 2D Multi-Planar Reconstruction tool allows you to set an axis point in the body and then pivot around it. OsiriX allows you to export the 360 degree rotation as an animation from which I devised a way to export 360 still images. Each Dervish is made up of 36 slices of the body around a certain axis – every tenth frame of my animation / every 10 degrees in the body. Each image is then printed onto a clear film and bound at the centre like a book before being attached to a hanging system which allows each page / scan to fan out and from a distance give the illusion of a figure and from close up offer an entrance to the core of the body.
Dervishes comprises of five sculptures, one has its pivot point through the centre, one through its spine, one through its belly button, one through its left side and the final through its right side. The resulting sculptures offer differing encounters with the body. The central axis is where the information is most dense and information gaps or openings increase (equally) the further from the axis point you are.
I see this work as being a model of disembodied knowledge and encounter of information to such an extent that as its creator I have become blind to the body that facilitates it. At first I felt uncomfortable using dataset of a person that I didn’t know but now I feel it is the most appropriate situation when I am post processing the scans so heavily. Freed from an emotional link to the original (or ‘the truth’ as one radiographer once referred to the scanee) I have been able to invest myself and my ideas into it. I identified MELANIX as having the qualities I wanted from a digital body (it was more or less whole and high resolution) and from then on I used it for my own goals. I was not interested in trying to create sculpture which represented a body, I wanted to create a sculpture that explores how it feels to work DICOM data, a sculpture that exposed not the body itself but that exposed the vision that is seeing that body.

RADIANT – PET (Positron Electron Tomography)
In 2003 I exhibited Family Portrait at Herrmann Wagner Gallery, Berlin. During the exhibition I was approached by a very enthused doctor who suggested I look to PET scanning to make a sculpture for its fantastic colours. I assured him that on my return to England that I would look into PET to see whether as he suggested it might be possible to make a sculpture from PET scans. I quickly became excited by what I read and saw about PET – the images were indeed fantastic and highly colourful. The technology involved in PET is equally fantastic: PET is a nuclear imaging system which measures the distribution of an isotope administered to the patient in the body. Prior to the scan, the patient is injected with a radioactive isotope and after a waiting time of approximately an hour the patient goes into scanner. The isotope is taken up by the body in areas that are metabolising or functioning. The radioactive isotope decays by emitting positrons. Positrons then encounter and annilhate electrons, producing a pair of gamma photons moving in opposite directions. These are detected when they reach scintillator material in the scanning device that creates a burst of light that is then detected by photomultiplier tubes. PET scans are coded in colour. There are a number of different colour scales but in general the more active the area, the more brightly coloured it is. PET scans do not show the anatomy of the body as do MRI and CT, but function such that the images can be very abstract and hard to recognise. For this reason CT is simultaneously acquired so that radiologists can match function to physiology. Often scans are referred to as Fused PET which means the PET and CT scans have been overlaid onto each other.

Fig. 4 Radiant 2005, Inkjet on acrylic, 50 x 70 x 100cm

A useful analogy for PET scanning is fireworks: imagine the isotope as a box of fireworks that is injected into the body in order for it to find a place to be ignited and explode into balls of light. Just as at a firework display we are asked to stand back, so too we are asked to stand back when a patient/subject is having a PET scan. Once the patient/subject has been scanned they are radioactive and are therefore isolated. Even when the patient needs to go in the scanner the radiographer instructs them from a different room. As with CT therefore it would not be possible for me to commission a PET for non-clinical reasons.
PET scans are most commonly used to diagnose cancer as tumours metabolise and therefore show/glow up in a scan.  It occurred to me that the more tumours or metabolically active areas in the body, the more colourful the scans would be so I wrote to a number of scanning departments asking to have access to the scans in order to be able to compare them. I was invited to a meeting at the PSSC to discuss my project but was told that I would only be able to have access to scans of healthy bodies which had been given the ‘all clear’ for they were worried that a patient might recognise themselves from the placement of the tumours. This was something I had not considered, that an illness such as cancer can become your identity. Since then and after shadowing at the PSSC I realise that this is very much the case and that datasets are more likely to be called by their pathology than by the patients name because of course bodies change; tumours grow and shrink or spread, they never stay the same, they are not constant. With PET there is the added fact that the radioactive isotope is only in the body for a relatively short period of time and that after it has been excreted the body can no longer be scanned. What the person has been doing before the scan will also affect the scans – if a patient is going to have a brain scan they are asked not to read or think too much, similarly if someone is having their throat scanned they are asked not to talk because it will impact on the results.
Radiant (Fig. 4)is made from a set of ‘all clear’ scans – a radioactive healthy body, radiant with joy.  I made the sculpture in a similar way to Family Portrait in that I asked the radiographer for specific scans through the body. I made small changes in that I asked for scans at closer intervals and printed the images onto acrylic using an UV cured inkjet process in order to keep the full range of colours in the scans. I asked PSSC for a full body scan, again to follow the structure of my MRI sculptures but when I received the scans I saw that the top of the head and the legs were missing. I sent a panicked e-mail to the PET radiographer asking him if there was a disc missing but he replied that this was indeed a ‘full body’ – in radiology a full body is the head shoulders, thorax and pelvis.
I made this sculpture in 2005 before I had OsiriX and no doubt had I had it then I would have asked for DICOM data in order to work with it myself. In retrospect I am glad this was the case. I would have been tempted to be clever with the data and create new cuts through it (PET is as dense as CT data) whereas to be sensitive to the subject and expose the poetry within it needed no clever post processing – what was important here was the message of the scans and the title. People react to this work very positively – the scans are mostly a royal blue except in the active areas, which show up in yellow, green and red. When the printed sheets are stacked together the body glows densely blue and feels very spiritual.
My experience of working with PET scans to make a sculpture was very much like my time observing at PSSC: I was asked to stand back and not to stay too long. Although I prepared the images digitally, in order to keep the full colours I was unable to be involved in the process of printing. Of the three modalities I believe PET is less about digitising the body but more about documenting a mysterious spectacle taking place inside the body. I feel PET is less relevant in a discussion of the digitisation of the body or the creation of a digital copy of the body because of its temporary, event specific nature and an artwork created using PET scans needs to be sensitive to this.

Conclusion
As an artist using datasets created from these different scanning techniques I have found they lend themselves to exposing and exploring different ideas. MRI, which is a bespoke and allows bodies into the scanner for non-clinical reasons, has the most potential for subverting the cold, distanced medical gaze. By putting my family into the scanner, love and emotion entered a space normally reserved for pathological diagnosis. Family Portrait suggests that DICOM can become relic; DICOM can be used to create images and objects from our loved ones for us to cherish and covet.
CT and software designed for its post-production offers huge amounts of possibilities through small degrees of transformation. CT data is so dense and holds so much information it needs to be processed in a number of different ways before it can be understood. CT datasets demand an investment from the radiologist/ artist – information needs to be teased of the data. I started working with CT data at around the same time as I first started making my first avatar on Second Life and just as I can modify my avatar to have any colour of skin, hair, clothes, any shape nose, mouth or legs, so too my CT dataset can have any colour kidney, heart or skull, so too it can be only bones, skinless or only lungs. Until the next click of the mouse and then it might become something different. Just as my Second Life is an avatar on which I display signs so to is a CT dataset. The difference is the level to which this can be done – my avatar in Second Life is all surface: pierce the micro thin skin and it is empty whereas CT datasets are densely and exhaustingly full. In the instance of Dervishes I decided to transform my dataset into a manifestation of my ideas around the Pivotoptic,  I am currently working on a very different work yet I am still using the same dataset. CT is like my son’s playdough, infinitely reusable and malleable but it will never dry out and one version can exist independently of the other.
PET is a record of an event, a spectacle taking place inside the body. It shows how the body hosts this spectacle and does so in vivid and seductive colours. Molecular imaging is the most rapidly growing area of research within diagnostic imaging and MRI and CT are already beginning to adapt to do similar things. Functional MRI for example is being used to see which areas of the brain are active when we are think of different things. As an artist I find PET the hardest of scanning technologies to engage with as it offers very little mediation in post-production and no intervention during the scanning. The techniques of PET could be very exciting for the artist if there were some control of the isotopes or if the patient could try to control how the isotope were taken up by the body (such as thinking of a certain thing or doing a certain activity before the scan) but because Functional MRI is already working is this area is would be not be allowed in PET. This is not to say that PET cannot be used to make artworks of significant beauty and power but that it is much harder to free from its medical context.
As well as differences there are properties that unite all kinds of DICOM. All DICOM has the potential for unfaithful duplicity; all of the sculptures discussed in this paper are editioned works - there are six copies of each sculpture plus two artist copies. Once digitised the body can be endlessly copied. All the sculptures need gaps in them to work – the techniques I use to create the sculptures all rely on leaving space between the scans. If all data in the scan where included in the sculptures, the body/information would be lost.  DICOM, like all datasets, can never be perceived all at once, we need to move, switch our vantage point and filter. 
As an artist, I work with DICOM not only to discover its potential for making artworks that can be enjoyed aesthetically but also to try and make sense of an increasingly digitised world which the physical body is not allowed access to. The more time I spend working with and through computers the more I feel the way I interact with others and even myself is changing. I definitely have both a physical and a virtual reality but I am still not confident about who I am in the virtual reality – am I a digitised version of who I am in physical reality or someone very different? In a future which promises to be even more digital and when the current generations are growing up developing their on and off line selves simultaneously what will the relationship between them be? I look to DICOM, which is a direct digitisation of body to data to give me clues and suggest answers.

References
[1]  HUGHS, J. (Director and writer): Weird Science. Universal Pictures, USA.
[2]  http://secondlife.com
[3]  CARTER, H, Police force tests airborne spy camera, The Guardian, 22 May 2007, UK.