How do I use MedScan to read a document I have on my local drive?

You can use MedScan to extract relationships for an individual file or from all files found in a directory. Files can be html or text or *.pdf documents (note: some older *.pdf documents contained scanned image files of documents and cannot be read by natural language processing technology).

Launch the MedScan reader from the import menu. Select “Relations from Literature via MedScan” as shown in Figure 1.

To extract relations from a single file, such as a full text article, select the “Extract text from a single file” option (as shown in Figure 2),select the browse button (red arrow), and point to the appropriate file on the hard disk and press start.

To extract relationship information from more than one file at a time, put all the files in one subfolder and then select “Extract text from all files in a folder and its subfolders” (Figure 3). MedScan will “read” the designated files and extract a list of identified entities and identified relationships. This information will be added to your local ResNet database and will automatically be displayed in a pathway if the “Open the pathway automatically when import is finished” box is selected.

Select a location for the resultant pathway to be stored, provide a name for the pathway and choose start to run the MedScan text-mining process on the designated files.