Take a Byte Out of Time

hca summer 2012 exhibit poster

The exhibit brings together examples from archival collections covering 100 years at Oregon Health & Science University.

These examples include early xrays from the Hilda Drum Collection, glass lantern slides from the Laurence Selling Collection, research data from the Earl Palmer Papers and images of microarrays and gene-expression from the files of the David Grandy Papers.

In addition, the exhibit touches on some of the challenges that arise in archives when preserving research data. The challenges include everything from the fragile nature of early slides to the complex series of files used to create more modern datasets.

Soviet Doctors

Research results and presentations of Soviet knowledge via the medium of vinyl LPs. This is not a case of research data being saved in a certain format necessary to its rendering at a later date, or its re-examination. Rather, these LPs highlight the challenges of transmitting data across cultural barriers and preserving that transmission for future researchers.

The Soviet Doctors LPs contain lectures from premier Soviet scientists in the fields of cybernetics, modern science and organ transplantation. The main issues surrounding the keeping of these materials and their availability for future generations is two-fold in this case. The equipment necessary to listen to the LP is a record player with 33/3 RPM speed (not uncommon among record players). The second major challenge is in the form of a language barrier. Currently the combination of those two forces make the LPs for the most part in-accessible. As long as the vinyl is stored in a cool environment, standing vertically, then with the right equipment it should consistently render for a long time, even longer in many cases than cassette tapes or cdroms.

Laurence Selling Collection

Scanned glass lantern slides. Example slide contained EEG (electroencephalograph) information, the other two slides show possible physical issues with the medium (breakage and de-layering)

The glass lantern slides in this example were used by Selling in lectures and in the displaying the results of tests. In the examples given we see the charting of an electroencephalograph (EEG) as an example of what the read-outs look like and most likely as teaching aid for how to read the results.

The second example shows the common issues that are run into when preserving glass lantern slides. firstly, they are very fragile. Secondly, the various layers of materials involved make the structure of the item prone to disintegration. Glass lantern slides were commonly used starting in the mid to late 19th century up until 1915-1920 and were composed of glass layers, a binding agent, an information recording medium (dyed starch grains and a silver halide emulsion), then another glass plate. Common issues include peeling of one of the layers, chemical disintegration of the binding or simple glass breakage from poor handling/accidents.

Harold B Meyers

The featured data is from the blood pressure of a rabbit used in papers on the toxicology of iodine. Meyers probably used a Van Basch’s type of sphygmomanometer to collect data on the blood pressure of the rabbits in his experiment. The read-out of the data is represented by a graph showing the peaks and valleys of the animal’s biological rhythms. The data was then used to determine at which point biological functions slowed due to a drop in blood pressure because of the effects of the iodine injections.

Similar to the other cases physical media used in experimentation, the data from the experiment can be examined by future researchers and re-interpreted. Additionally, the results can be re-tested using modern methods.

Hilda Drum Collection

Contains 3 large animal skull xrays. Drum was an early xray technician and these xrays (technically photographs of xrays then scanned into jpeg) represent early tests. In addition, these x-rays highlight some of the challenges in preserving this type of media. The potential issues are based on the type of photographic film used to produce the xray print initially with additional concerns for the size of the xrays (increased weight) and potential to become brittle or off-gas depending on the type of film used. Acetate film shows degradation after 20 years stored in office-like conditions. Ideal storage is at 35-55 F and 50%RH, stored vertically, lignin-free, buffered envelopes. For Nitrate -based, copying and disposal of originals is advised due to off gases from the material as it degrades. Polyester-based (last 30 years) are less of a challenge in terms of preservation.

David Grandy Collection

Dr. Grandy’s research into the effects of LSD and MDMA on neurology, specifically which neurotransmitter proteins are involved and to what extent in the mediation of these drugs’ effects. The example figure from Dr. Grandy’s research paper show “deduced amino acid sequences for the rat and human trace amine receptors aligned with other homologous G protein-coupled receptors” (from Dr. Grandy’s paper “Amphetamine, MDMA, LSD, and Metabolites of the Catecholamine Neurotransmitters Are Agonists of a Rat Trace Amine Receptor).

Other examples from Dr. Grandy’s papers include several microarrays found on a ZIP disk and stored in .PCT and .PIC format. These formats are Macintosh specific Quickdraw-based formats. Typically only viewable on Apple machines, these files can be rendered via Adobe Photoshop. The .PCT and .PIC formats are considered deprecated, having been eclipsed by PDF. The formats in and of themselves in this case represent second-generation data. The original data having been constructed and rendered using microarray analysis tools that generate complex file formats such as MeV, tav, .dat, .cel, .chp, .rpt and some .txt files. Depending on the information the researcher wants to present, elements are arranged and image files constructed from those elements. These files are typically in .jpeg, or .tiff but as we have seen in this case, sometimes they can be in a format unreadable by the technology at hand. At which point, it is either necessary to acquire that technology or migrate the data to another more common format. Migration from one format to another has its own challenges, the potential for loss of data or context with migration is very real. Files, divorced from their relationships within an electronic format can become unstable, or can risk becoming a file that does not contain all of the data the original creator intended to maintain.

Other challenges can be found in the way that data is stored, lost, stored again in a different format, or migrated to a format for purposes other than preservation. These Dotter xrays represent one such challenge in that they were original created as physical items, but at some point along the way those physical items were digitized into a proprietary software format and the originals not heard from since. These particular xrays were stored through a software suite called eFilmLite. The output of these files can only be read through the software contained on the cdrom which also contains the images. This effectively makes the cdrom an access point that must be used each time. Furthermore, the software used to store the xray images overlays a certain level of metadata important to the software creators, but irrelevant to the original xray. The metadata provided covered the original image as can be seen and did not contain necessary contextual information. From what could be observed from going through these files, it appeared as if the patient numbers had no real correlation to Dotter’s own patient records. In fact they seem to be a simple alpha-numeric consecutive counting scheme. The remaining metadata appears to be default field data or data related solely to how eFilmLite perceived the files’ properties.

The tactic used to ensure that we store the most robust, information-rich and accessible format led to the stripping away of the metadata from each image by using the software tools to remove the layering, therefore allowing a .tiff to be exported. The .tiff is then stored on the digital archive networked drives where access can be allowed without the need to go through the software barrier presented by eFilmLite.