The Attachment Server contains clipped images, drawings, chemistry files, full page images, and other non-textual files delivered by patent offices. Most of the links to the files in the Attachment Server are embedded in CLAIMS Direct XML files. There is a simple web service available to access this content (see Attachments for more information). This document describes the content of the Attachment Server, explains how image files are embedded in the XML files, and discusses how to interface with the server.

Raw Data

Image content from patents is available from data sources in different formats, which include the following:

• Full page images
  These images are facsimiles of the original documents delivered as TIFF files or PDF documents, which can be a single page (one file per page) or multi-page (one file per document). Note that additional PDFs may also contain information such as sequence lists and reference tables.

• Embedded images
  Embedded images are images that are part of the document. Unlike full page images, they are not the entire document.

• ST35, mixed mode
  Mixed-mode format is an old format delivered by some patent offices. This format includes both text and binary data. The text blocks and binary blocks are bytes of data lined up in a file. There is no separate image file; the image data is just another "in-line" block in the file.

• Referenced images
  Images are delivered as separate files in TIFF or JPEG format. They are named systematically, and correspond to references within the XML or SGML documents. This is the most common current practice for delivering images. Multi-page TIFFs are indicated when multiple id attributes point to the same file. These are usually only found in JP records. A special tool or viewer is required for proper handling of multi-page TIFFs. A list of viewers can be found here.

Like the images, other non-textual content can also be provided in separate files and stored in special formats such as chemical structures, gene sequences, or mathematical formulae.

Attachments in the CLAIMS Direct Patent Database

Images and other files containing non-textual content (collectively "the attachments") are stored in the Attachment Server. Files containing embedded images, as well as mixed mode files, are pre-processed before being loaded to the CLAIMS Direct Patent Database. Links to the attachments are generally referenced in the CLAIMS Direct XML files.

IMG Element

The most common way to find embedded/referenced images in the full text of Alexandria XML files is by using the img element.
Depending on the data source, different parameters can be found in this element, including the following most relevant attributes: 

The img element may be found in the abstract, description, or claims sections of the XML document:

Abstract Section

<abstract mxw-id="PA87583406" lang="EN" load-source="patent-office">
  <p num="pa01">This invention relates to conduits (12) that allow communication of fluids
                from one portion of a patient's body to another; and, more particularly, to a
                blood flow conduit to allow communication from a heart chamber to a vessel or
                vice versa, and/or vessel to vessel
    <img id="iaf01" file="imgaf001.tif" wi="78" he="89" img-content="drawing" img-format="tif"/>
  </p>
</abstract> 

Description Section

<p num="p0078">For screening purposes a GST and 6 x His fusion of the LBD (from amino acids 155
                of hLXRalpha to 447) of human LXRalpha is constructed by first cloning a Gateway
                cassette (Invitrogen) in frame into the Sma I site of the pAGGHLT Polylinker (Pharmingen) [...]
                Primers used for Amplification are:
  <img id="ib0004" file="imgb0004.tif" wi="165" he="15" img-content="dna" img-format="tif"/>
  <!-- [...] -->
</p>

Claims Section

<claim-text>A phosphoramidite reagent of the formula I
  <img file="00270001.tif" id="img-00270001" he="32" wi="91" img-format="tif" img-content="cf"/>
    wherein Y and Y' each independently are selected from optionally substituted C<sub>1-6</sub>-alkyl
    or Y and Y' together with the nitrogen to which they are bonded form a non-aromatic [...]
</claim-text>

Drawings Section

Attachment references can be embedded in the img element as described above but they can also be in a drawings section of the XML document. Pages published at the end of the original patent document and containing drawings are also referenced in this section.

<drawings mxw-id="PDW3055834" load-source="patent-office">
  <figure num="1">
    <img file="00380001.tif" id="img-00380001" he="228" wi="180" img-format="tif" img-content="dr"/>
  </figure>
</drawings>

A figure id can be used to reference a drawing in the full text as follows:

<p num="p0031">Referring now to
  <figref idrefs="f0001">FIGURES 1A and 1B</figref>, a coronary artery bypass is accomplished by disposing a
    conduit 12 (<figref idrefs="f0001">Fig. 1B</figref>) in a heart wall or myocardium MYO
    of a patient's heart PH (<figref idrefs="f0001">Fig. 1A</figref>). [...]
</p>
 
<drawings mxw-id="PDW10967064" load-source="patent-office">
  <figure id="f0001" num="1">
    <img id="if0001" file="imgf0001.tif" wi="140" he="230" img-content="drawing" img format="tif"/>
  </figure>
</drawings>

Chemistry Section

Chemical formulas can be found in the img section with "chem" or "cf", along with "drawing" values in the img-content attribute. Another way to encode a chemistry-specific embedded image is through the use of the chemistry element:

<claim-text>Use of a compound according to formula (I), or pharmaceutically acceptable salts or solvates thereof,
  <chemistry id="chem0011" num="0011">
    <img id="ib0071" file="imgb0071.tif" wi="74" he="50" img-content="chem" img-format="tif"/>
  </chemistry>
 wherein [...]
</claim-text>

The USPTO provides chemical structures in ChemDraw (CDX) and MDL (MOL) formats. References to these special files can be found in the attachment element in the chemistry section:

<chemistry id="CHEM-US-00001" num="00001">
  <img id="EMI-C00001" he="19.64mm" wi="54.36mm" file="US07307149-20071211-C00001.TIF" alt="embedded image" img-content="table" img-format="tif"/>
  <attachments>
    <attachment idref="CHEM-US-00001" attachment-type="cdx" file="US07307149-20071211-C00001.CDX"/>
    <attachment idref="CHEM-US-00001" attachment-type="mol" file="US07307149-20071211-C00001.MOL"/>
  </attachments>
</chemistry>

Other Sections

In the same way that chemical structures are referenced in the special "chemistry" section, other attachments also have specific elements in the XML files: sequence-list, math, megatable-doc, and table-external-doc.

Search Report Pages

Search report pages published by some patent authorities are frequently distributed as full page images. References to the image files can be found in the search-report-data container under the doc-page element:

<search-report-data>
  <doc-page id="srep0001" file="srep0001.tif" wi="154" he="233" type="tif"/>
</search-report-data>

File Naming

There is no consensus in naming referenced files. Every publishing authority follows its own rules, which also change with the years. The following tables demonstrate some representative examples of attachment listings associated with patents from different publishing authorities.

Attachment list for: EP-2207108-A1-20100714

Attachment list for: WO-2010072727-A1-20100701

Attachment list for: WO-2009081462-A1-20090702

Attachment list for: US-7307149-B2-20071211

Attachment list for: JP-2005100000-A-20050414

Attachment list for: KR-100920729-B1-20091007

Data Coverage

The availability of attachments differs depending on your subscription level. The attachment collection can be queried through our shared service API to get exact file counts for your level.