Prior to do load analysis on the bones, you will need to segment them from the DICOM images (I assume you are using CT-scans) and reconstruct them as 3D models.  Some of the following viewer also have basic segmentation capability:

http://technologyadvice.com/blog/healthcare/5-dicom-viewers/

https://sourceforge.net/projects/mito/?source=directory

Once segmented, a Marching Cube algorithm is usually used to reconstruct the 3D model usable for FE analysis.  You might need to improve the triangulation to get good results though.

In order to convert DICOM images to Iges file, you need to

use either Mimics software or softwares similar to that. Even these

softwares do not convert Dicom images directly to Iges file . They only

convert imgaes to STL file, then you might need to use software like 3Matic

to convert STL file to Iges file. Hope this helps.

The most straightforward way to do that is by using the open source Image (https://imagej.nih.gov/ij/).

It reads DICOM very well and can convert to other formats such as TIFF. There is also a lot of analysis that you could do with ImageJ

For a slightly more advanced solution you can also use ImageMagick,

you should convert DICOM file into STL format with MIMICS software then try to figure out a IGS or STP or maybe a 3DMODEL format by using Solidwork or Catia sofware. I am not sure but sometimes I heard that Ansys Icem could mesh a STL format. If it was correct you could jump a step. what is important for you is making an meshable format file. If you need more help inform me. (a late response just to new one surfing for it!)

where can i access dicom dataset (free and large enought)?