From shulea@waag.org Fri Sep 11 00:39:54 1998 Received: from home.actlab.utexas.edu (home.actlab.utexas.edu [128.83.194.11]) by A01.waag.org (8.9.1a/8.9.1) with ESMTP id AAA00799 for ; Fri, 11 Sep 1998 00:39:53 +0200 Received: from cynbewynbe (scts1-45.znet.net [207.167.86.45]) by home.actlab.utexas.edu (8.8.7/8.8.7) with SMTP id RAA12943; Thu, 10 Sep 1998 17:39:33 -0500 Message-Id: <199809102239.RAA12943@home.actlab.utexas.edu> X-Sender: sandy@home.actlab.utexas.edu X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0.1 Date: Thu, 10 Sep 1998 15:32:31 -0700 To: shu lea cheang , tabody@waag.org From: Allucquere Rosanne Stone Subject: Re: Jose's digital cadavers In-Reply-To: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" I'd like to add a few notes to the coming discussion about digital cadavers, especially as they relate to the Visible Human Project (VHP). This is in the interest of complexifying the discussion, and in pointing out how VHP constructed its origin myth in such a way as to render the program's origins and underpinnings invisible both to most researchers and to the public in general. It is not meant in any way to criticize Jose's (or Lisa's) work, but to nuance it. Since this is background material of interest mainly to the social and cultural studies industries, I think it should be shared prior to the actual public event, whose focus is slightly different. I stopped work on this research some time ago because I fell in love with and later married the project's software designer and lead programmer, which presented me with an ethical problem as well as a serious distraction. Recently I've been thinking about taking it up again, since it seems clear that I live in a country in which ethics no longer exist. VHP is the commercial exfoliation of a project begun in the 1960s at the University of Washington in the department of Biological Structure. At that time John Sundsten, the department chair, began to make crude visualizations of human organs using a flying spot scanner. This work petered out due to lack of any way to process the images, but the project was revived in 1978 with Sundsten's formation of the Graphics Lab within the department and the hiring of a task group dedicated to producing a database of the human body. By then there were powerful (for their time) computers available, and the lab acquired a Northstar Horizon and flatbed plotter to produce the earliest images. Sundsten hired Jeffrey Prothero, a brilliant young programmer with wide-ranging interests that included medical imaging, to design the visualization engine, which was later named Skandha. (The word refers to a Sanskrit term incorporated into Buddist practice, referring to the five levels of illusion before true vision. The current version is Skandha4, so true vision must be close at hand.) The Graphics Lab and its evolution into the Digital Anatomist Project (DAP) was funded by a seed grant from the National Library of Medicine (NLM), and its work was later incorporated into Al Gore's High Performance Computing Challenge (HPCC). The DAP's Visible Human, before it was the Visible Human, is still used as part of an online quiz system for medical students. The online CT brain scans remain those of the lead programmer's brain, because he was handy, and he loved to lie still for the long times it took to produce a whole-brain CT image because it gave him uninterrupted time to think. When the group needed bodies, they went not to the prisons but to the hospitals and the police, seeking people recently dead from blows to the head, sudden fever, or drug overdose. In particular, young ODs were the prize, considered the hen's teeth of the project. Dan Graney was the head of the Willed Body Program, and it was his task to keep close contact with the hospitals and police, as well as the close order drill of medical science such as persuading older people on their deathbeds to "leave their assets" (i.e., bodies) to the Medical School. But nothing was as good as a nice young overdose fresh from the ER. By 1984 Prothero had developed the visualization engine to the point where it was capable of modeling more complex data than could be obtained by the slicing methods then available. Sundsten went out into the grapevine, and after a while was referred to Wolfgang Rauschning, a brilliant and talented Swedish orthopedic surgeon who was planning a second career in medical visualization. Rauschning specialized in knee surgery, and as anyone with a serious sports injury knows, orthopods badly needed better understandings of the interior workings of the knee and similar complex structures. As part of his research Rauschning had created a breakthrough in imaging technique. Instead of attempting to slice through the body sample with a microtome, he delicately and evenly abraded the surface, removing layers only a fraction of a molecule thick. Then he polished the new surface thus exposed, and photographed it. The effect was the same as if he were using a microtome capable of slicing tissue to a thickness of a few molecules. A side effect of this is that one obtains a huge number of "virtual slices". In fact, once the project got rolling Rauschning produced such enormous volumes of data that the DAL couldn't process it all. He had also developed critical ancillary techniques to perfection. For example, once a fresh layer of tissue is exposed, how clearly it shows up to the camera depends critically on how its surface is chemically treated. Also, all the chemicals known at that time produced changes in the color or texture of the samples. Rauschning found a way to delicately coat the surface of the tissue with a complex glycerol mixture, let it set for a time, and then angle the camera just so. Surfaces digitized in this way evinced a tremendous depth and brilliance, better than anything the Biostructure team had seen. And to top it all off Rauschning was a perfectionist, willing to spend hours painstakingly coating and angling each surface until the image was as clear as he felt it could become. As it turned out, Rauschning was also difficult to deal with, highly proprietary with his data, and reluctant to change his procedures. In particular it was necessary for him to add fiduciary marks to each sample, so that Skandha could line them up properly in order to be able to reconstruct the solid organs, and this took a while to negotiate. Eventually they reached agreement on all issues. The project shifted into high gear. Because as far as UW was concerned the procedure was still experimental and body parts were in demand by other departments in the university, Graney parceled them out parsimoniously. As soon as a fresh overdose case became available, Graney sliced off a limb or two and iced it down. He then had the limb flown to Stockholm in polyfoam boxes, where Rauschning further cooled it in liquid nitrogen, processed it, and sent the data back to Seattle. The results were terrific. The combination of Rauschning's brilliant abrading, coating, and photography, combined with the Skandha engine's ability to reassemble the digitized information from thousands of tissue slices, was a gratifying realization of the project's purpose. Having completed what he saw as the proof of concept phase, Sundsten went back to NLM and proposed that for the next phase, the project should digitize an entire human body. The stakes quickly soared. After years of bits and pieces of grants trickling into the project, Large Major Funding was suddenly a real possibility; and Sundsten didn't notice that agencies outside the university ambit were circling in on the smell of money until it was too late. There was no question that the VHP proposal looked better on paper than the Digital Anatomist's; even the VHP's proposal itself required more money to complete than the lab saw in a month. Sundsten was not a grant writer on the same level with VHP's. Further, the lab's proposal included Rauschning, a German living in Sweden, while the VHP proposed using only American sources, suggesting that Rauschning's techniques were indistinguishable from those developed at the University of Colorado. UC did have a large microtome (or macrotome) of sorts, but its resolution was coarse, and it didn't have much similarity to Rauschning's finicky, exquisitely precise abrading method, nor to his techniques for treating the samples before digitizing them. But in a Buy-American era, VHP's bringing its own corporate backing to the table (and indeed the fact of VHP's being situated in corporate culture), combined with VHP's assurances of the identical product from local sourcing, carried the day. It was then, and only then, that the issue of where to get the bodies was raised. VHP's proposal included the whole enchilada -- obtaining bodies, translating them into digital streams, and reconstructing the streams into human simulacra. VHP used their own software, for the most part combining off-the-shelf items (that were not available when Skandha was developed) with some glue code of their own. But VHP did not have the highly developed networks that allowed the Digital Anatomist lab to obtain their corpses. I omit some data here in the interests of speed, but the outcome was that VHP wound up going to the prisons instead of the hospitals for their working material. And this is probably my relevant point here -- that VHP went to the prisons because of a chain of events related to their ability to obtain bodies from the preferred sources, which were not prisons. Further, VHP evinced considerable skill at erasing any evidence of Rauschning's work. Analysis of this data is not complete, but it's a not unreasonable guess that it was because comparisons of Rauschning's samples to the product of UC's microtome revealed the UC samples to be so inferior to the Digital Anatomist's ("like comparing flint technology to the Space Shuttle", in the words of one outside researcher) that there was real fear within VHP that NLM might revoke the grant and/or return it to the Digital Anatomist lab. Using the outmoded microtome approach, UC had trouble producing reliable slices six millimeters thick, while Rauschning's technique could reliably and repeatably produce images only fractions of a micrometer apart. Since his name was not supposed to be revealed, Joe, the prisoner, was probably the worst-kept total secret in the history of medicine. A few notes in closing: Joe was eager to participate in the project -- " Yup, yup, sign me up" is the way some of those close to him put it. There are reasons to believe that he was not mentally competent to consent. And when the visualization software first showed his tatooed skin (which is also an organ and therefore part of the project) and his face, some of the programmers recoiled in disgust. "I was uneasy about (slicing up a prisoner) for a long time," one scientist said, "but once I saw his face all my misgivings disappeared. That guy had the face of a medieval executioner." --Sandy From shulea@waag.org Fri Sep 11 16:29:59 1998 Received: from RL0002.unimaas.nl (rl0002.unimaas.nl [137.120.1.2]) by A01.waag.org (8.9.1a/8.9.1) with ESMTP id QAA06515 for ; Fri, 11 Sep 1998 16:29:58 +0200 X-Confirm-reading-to: J.vandijck@LK.UNIMAAS.NL X-PMrqc: 1 Received: from josevandijck (facburfdcw0127kap2.unimaas.nl) by RL0002.UNIMAAS.NL (PMDF V5.1-10 #D3114) with SMTP id <01J1OQ77PQOI001KYS@RL0002.UNIMAAS.NL> for tabody@waag.org; Fri, 11 Sep 1998 16:22:20 WE Date: Fri, 11 Sep 1998 16:22:22 +0200 From: Jose van Dijck Subject: digital cadavers To: tabody@waag.org Reply-to: J.vandijck@LK.UNIMAAS.NL Message-id: <01J1OQ77Q3V8001KYS@RL0002.UNIMAAS.NL> MIME-version: 1.0 X-Mailer: Pegasus Mail for Win32 (v3.01a) Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Priority: normal Sandy: Your comments on the backrgound of the VHP are fascinating, and eye- opening as far as the development of the slicing-digitalization technology is concerned. I hope you understood from my paper that I'm not so much commenting on the selection of a prisoner for this project, as on the eagerness of journalists and commentators to play up his criminal record. In the public dissemination of the VHP, the historical-cultural context of Jernigan the criminal becomes part and parcel of the digital cadaver! Thanks for sharing these revealing background details! Jose From shulea@waag.org Fri Sep 11 19:23:14 1998 Received: from [194.134.18.57] (A47.waag.org [194.134.18.57]) by A01.waag.org (8.9.1a/8.9.1) with ESMTP id TAA07903 for ; Fri, 11 Sep 1998 19:23:12 +0200 Date: Fri, 11 Sep 1998 19:23:12 +0200 Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" To: tabody@waag.org From: shu lea cheang Subject: sugar coated theory pills We'll be making sugar coated theory pills in Amsterdam for the forum. collecting wise words for candy pull out. please submit if you would like to be part of the candy scheme by monday. sl http://brandon.guggenheim.org/gifts ps. Vernon, if you still can't go beyond the bigdoll, please try to go to roadtrip directly. http://brandon.guggenheim.org/roadtrip/road.html there the interface with various icons would take you to various narrative. between fiction and friction. From shulea@waag.org Fri Sep 11 20:24:34 1998 Received: from home.actlab.utexas.edu (home.actlab.utexas.edu [128.83.194.11]) by A01.waag.org (8.9.1a/8.9.1) with ESMTP id UAA08300 for ; Fri, 11 Sep 1998 20:24:32 +0200 Received: from cynbewynbe (scts1-33.znet.net [207.167.86.33]) by home.actlab.utexas.edu (8.8.7/8.8.7) with SMTP id NAA15213; Fri, 11 Sep 1998 13:24:22 -0500 Message-Id: <199809111824.NAA15213@home.actlab.utexas.edu> X-Sender: sandy@home.actlab.utexas.edu X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0.1 Date: Fri, 11 Sep 1998 11:12:09 -0700 To: J.vandijck@lk.unimaas.nl, tabody@waag.org From: Allucquere Rosanne Stone Subject: Re: digital cadavers In-Reply-To: <01J1OQ77Q3V8001KYS@RL0002.UNIMAAS.NL> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" >concerned. I hope you understood from my paper that I'm not so much >commenting on the selection of a prisoner for this project, as on the eagerness >of journalists and commentators to play up his criminal record. In the public >dissemination of the VHP, the historical-cultural context of Jernigan the >criminal becomes part and parcel of the digital cadaver! You bet! It wouldn't have mattered if they had been aware that the body of choice happened to be a criminal by virtue of circumstance rather than design, they would still have reported it as they did -- although one does get the sense that the geist is laughing up its sleeve at the accident! From shulea@waag.org Fri Sep 11 23:20:02 1998 Received: from guggenheim.org (smtpgate.guggenheim.org [205.232.24.31]) by A01.waag.org (8.9.1a/8.9.1) with SMTP id XAA09465 for ; Fri, 11 Sep 1998 23:19:57 +0200 Received: from SOHO-Message_Server by guggenheim.org with Novell_GroupWise; Fri, 11 Sep 1998 17:17:20 -0400 Message-Id: X-Mailer: Novell GroupWise 5.2 Date: Fri, 11 Sep 1998 17:16:50 -0400 From: "Jon C. Ippolito" To: tabody@waag.org Cc: Jmereness@guggenheim.org, mlavin@guggenheim.org Subject: outline of tabody event in new york Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Disposition: inline Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by A01.waag.org id XAA09472 Hello all, In case you didn't know, Matthew Drutt has asked me to help coordinate the New York presentation of TABody. While joining the group at this late date means missing out on the initial salvos of some pretty interesting discussion threads, it also gives me a distance that may be useful in finding an order implicit in the welter of musings that have been posted to date. I see my job as helping to define the arena--technological, physical, and performative--in which the TABody participants will speak. Since several postings have recently requested it, I am including here an outline of the event as I understand it will occur from the New York side. I welcome questions about or revisions of this outline and look forward to engaging in a more substantive discussion of the content of the panel once this foundation is secure. TECHNOLOGICAL ARENA (components) a. Local public address system (mikes, speakers, etc.) for panelists b. Streaming audio via Netphone mixed from mikes in New York, Amsterdam, and Banff. Functionally this software will simply act like a giant speakerphone. (Outside visitors to the Web site will have "listen-only" access to this audio feed as well.) c. Ordinary Web access to a password-protected Web site for the "court reporter" (Kimberly and Shu Lea). This person's primary task will be simply to highlight a button representing the panelist speaking. d. Uploaded video images. These will be simply static jpegs shot by a roving person with a videocamera and pushed to the Web site from New York and Amsterdam (and maybe Banff). e. Comments that are input as text by outside visitors to the Web site. These can be selected and turned into audible, computerized speech by the court reporter and by visitors to the Web site. f. The same interface to input/select comments will be available to local visitors to the Guggenheim via kiosks in front of the videowall. (I am assuming these visitors would not be able to trigger audible speech that would interrupt the panelists.) PHYSICAL ARENA (space for presentation) Ascii art's not my specialty, but here goes: _ | |<-visitor at kiosk _| _ | 1-> <-reporter (kimberly) 2-> <-| monitor <-AUDIENCE 3-> <-audio (justin/james) _| _ | |<-visitor at kiosk <-roving videocamera-> _| The videowall is the three boxes at left; lines of site are indicated by the arrows. 1,2,3 are the three New York panelists, who sit on a riser at a table looking out at the audience. In front of them is a computer monitor showing the same thing that is on the middle videowall panel (which is controlled by Kimberly). Behind the audience is a table where Kimberly sits with a mouse and keyboard--the videowall is her "monitor". Justin Davila and James [Jamie] Mereness sit next to her at a computer that monitors audio and video uploads. PERFORMATIVE ARENA (script) I don't remember seeing anything definitive come over the listserve on this, so I'm throwing out this proposal for consideration: 1. test Matthew and I would like to conduct a test of all the equipment with all of the New York participants present (and at least one from Amsterdam or Banff, the more the better) on Saturday at 2 pm EDT (ie, exactly 24 hours before the real thing). This is not just to make sure the technology works, but to give the moderator a chance to try out some techniques for managing the audio maelstrom of superposed voices. 2. local introductions by Jon Ippolito (New York), John Hanhardt/Matthew Drutt/DeWaag folks (Amsterdam) I would like to suggest that the local introductions happen simultaneously in New York and Amsterdam *without* Netphone broadcast. While I'm certainly planning to introduce the panelists, thank our staff, tout the Guggenheim's wonderful programming, etc, I'm not sure that the global Internet audience is tuning in to hear this from me. I think they want to hear the panelists. 3. alternating presentations: To prevent the privileging of one venue over another, I would suggest alternating presentations, beginning with the official moderator--something like this: ny person (Lisa Cartwright) jennifer gonzalez dewaag person ny person vernon dewaag person ny person jennifer terry dewaag person banff person (Sandy Stone) During these presentations, visitors to the Web site would be able to click through the corresponding images, thanks to Shu Lea who will be uploading them to Web pages in advance. (PANELISTS WHO HAVE SLIDES: PLEASE EXPRESS THEM TO ME AS SOON AS POSSIBLE. If I don't receive them by Tuesday I cannot be sure of digitizing them in time.) The court reporter will be clicking through these images on the middle screen for the benefit of the local audience. 4. local/global discussion After the presentations, the moderator tries to keep the conversation going among all 2+ venues, with the additional job of deciding when to interrupt by triggering a text-to-speech audio clip from the local or global audience. Since it is easier for Kimberly to keep track of these rather than Lisa, I would suggest that we figure out some way for her to signal her when a germane text comes across the screen. (hand signal? software cue?) 5. contingency plans Here's the way I see the conference gracefully degrading if the Netphone or Web technology fails: a. to monitor any problems, the tech crews should maintain an open chat line through the audio-video upload computer. b. if that fails, crews should have a phone handy and know the phone numbers for New York, Amsterdam, and Banff. c. if Netphone goes down but the Web interface still works, the court reporter can type in summaries of what the panelists are saying and those can be passed back and forth among venues. d. if the Web server goes down, we all have pleasant discussions in meatspace and go out for drinks afterward. I look forward to your comments or questions. jon