An Interface for the Visible Human Dataset

The purpose of this project is to build an interface for the Visible Human Dataset. This is for people who are interested in extracting subvolumns of the dataset, particularly medical researchers and computer vision researchers.

Our client is likely to be a doctor, or a computer scientist. Obviously these two have very different user profiles. For example, one is likely to be task experienced and the other system experienced. Our goal is to design an interface which compromises requirements of both.

The project is a collaboration between VISLab and Department of Gynocology at UIC Hospital who is interested in extracting female reproductive system from the Visible Female Dataset.

The major functionalities include:

1. Allow user to view an image from CD, specify a subvolumn to extract, and save the subvolumn on a local hard drive. For this, we will build a schematic view of the human cadaver, and allow user to select an image using a slide bar tool. User will be able select a subrectangle of an image using a mouse, and set up subvolumn of the dataset to extract.


2. Allow user to view a slice of the subvolumn from different orienta- tions, ie, horizontally or vertically. Detect boundary of feature of interest.


Eventually as a long term goal, it will also allow user to reconstruct 3d models of anatomical units. The prototype is built using XMotif and fully functional.

Link to The National Libary of Medicine's Visible Human Project