AIM is the Annotation and Image Markup (AIM) Project developed at the Department of Radiology, Northwestern University Feinberg School of Medicine, and the Stanford Center for Biomedical Informatics Research  (1). The AIM Project intends to create a standardized way to add information and knowledge to an image, so that future image content can be “easily and automatically searched” (2).
In medical imaging (Imaging informatics), a radiologist interprets images mainly with free text, annotations, and markups. An image annotation is the descriptive information about the pixel data, whereas an image markup is the graphical symbol placed on the image to delineate the annotation such as, arrows, circles, and caliper measurements (3). Free text can be formatted in templates, but still there is not a way to directly tie the text to the spatial location of the pixel data.
“The goal of the AIM Project is to develop a mechanism for modeling, capturing, and serializing image annotation and markup data that can be adopted as a standard by the medical imaging community” (3).
Medical imaging plays a vital role in diagnosis, prognosis, and treatment of diseases, especially in cancer treatment. Traditional methods of using unstructured free text in radiology reports make it difficult to perform comparative analyses of lesion size(s) and location(s). This is particularly important in oncology where the progress of treatment can be measured by the interval change in size of a lesion.
Images contain vast amounts of information, the majority of which is encoded in the pixels Diagnostic Imaging Workstation Display Characteristics. However, the images as perceived by human or machine observers are not captured in a form that is directly tied to the images, thus limiting the value of radiology imaging in relation to other non-imaging data (1).
Digital images in medicine are managed with the DICOM standard format. DICOM says little about the content and meaning of the pixel data. It is difficult for humans and machines to index, query, and search for images based on free text descriptions, and this project intends to create a standardized method to do that.
The current challenges to retrieving images are: 1) no agreed on syntax for annotation and markup; 2) no agreed on semantics to describe annotations; and 3) no standard format (DICOM, XML, HL7) for markup and annotations (2).
By developing methods that describe the semantic content and tools to compose and relate the ontology-based descriptions of the content, the model of medical imaging will change from just storing pixels, to storing image data and meaning (4).
An open source tools kit is available from the National Cancer Institute’s (NCI) Cancer Bioinformatics Grid (caBIG) to incorporate formal imaging standards and technical frameworks to implement annotation and markup in images (1).
From the AIM-enabled radiology Picture Archive and Communication System PACS workstation, a pick list of RadLex  (5) anatomic terms would be selected by the radiologist to identify the location and be embedded in the annotation. Likewise, an AIM observation such as mass, and characteristic such as enhancing, would be chosen from a pick list, and contain the x and y coordinates of the markup (1).
To give an example of the AIM Project as used in radiology, a query of “find all studies that contain enhancing right middle lobe lung masses that measure between 5 and 6 cm2” becomes “find all image references in AIM annotations where AIM: Anatomic Entity = RID1310, AIM: Imaging Observation = RID3874, AIM: Observation Characteristic = RID6065, AIM: Calculation = Area and AIM: Calculation Result >5 and <6 cm2.” (1)
The use of AIM in formal imaging standards will enable the re-use of imaging information for clinical, research, translational, and educational purposes. Although useful for these purposes, AIM is critical for use in clinical trials.
Follow this link from AIM 3.0 RSNA Power Point Presentation on the NCI Wiki for a presentation of how an image markup can be coded by the AIM Project (3).
1. The Annotation and Image Mark-up (AIM) Project. Channin DS, Mongkolwat P, Kleper V, Rubin DL. 3, s.l. : Radiological Society of North America, Inc., Dec 2009, Radiology, Vol. 253, pp. 590-2.
2. National Cancer Institute. Annotation and Image Markup - AIM. NCI Wiki. [Online] [Cited: Nov 22, 2012.] https://wiki.nci.nih.gov/display/AIM/Annotation+and+Image+Markup+-+AIM.
3. The caBIG Annotation and Image Markup Project. Channin DS, Mongkolwat P, Kleper V, Sepukar K, Rubin D. 2, April 2010, Journal of Digital Imaging, Vol. 23, pp. 217-225.
4. Stanford University School of Medicine. Annotation and Image Markup (AIM) Project. Stanford Center for Biomedical Informatics Research. [Online] http://bmir.stanford.edu/projects/view.php/annotation_and_image_markup_aim_project.
5. Radiological Society of North America. RadLex. RSNA. [Online] 2012. [Cited: Nov 22, 2012.] http://www.rsna.org/RadLex.aspx.
Submitted by (Rhonda Luetkenhaus)