Scientific Visualization and Computer Graphics

Scientific Visualization and Computer Graphics > BI > FSE > RUG * Print * Search

Project Proposal: Skin Imaging Computational Database

Background

Modern skin imaging modalities produce a wealth of image data regarding skin lesions (e.g. dermoscopic images, confocal microscopy 3D image stacks, and histopathological images). Together with these images, doctors store meta-data concerning patients, diagnostics, treatment plans, and patient questionnaires. Furthermore, images in such databases are manipulated by various image processing tools (e.g. segmentation, contour extraction, tumor quantification, etc).

Medical specialists need a way to organize and work with such information. Ideally, we need a system where medical users can easily

  • add new images and associated meta-data
  • add relationships between all entities in the system (images, patients, diagnostics, treatment plans, questionnaires)
  • apply image-processing operations from a predefined toolset to the available images
  • visually query the available information along any desired dimension (e.g. patient name, imaging modality, extracted image metrics, dates, diagnostic outcomes, etc)

Goal

In this project, we aim to design a flexible, generic, and easy-to-use computational database for skin imaging. The requirements of this system are as follows:

  • all data in the system is modeled in terms of entities and relationships, using an ER database model
  • the database schema is easy to extend/modify to add new properties and relationships
  • a visual front-end is developed by means of which users can pose complex queries on all available ER attributes
  • the results of the above-mentioned queries are displayed visually by means of tables and actual images
  • predefined image processing operations (available e.g. as third-party executables) can be interactively applied to images in the database
  • the system is easy to use by a medical specialist (non-programmer)

Technical requirements

To ensure the integration of the system with third-party software and infrastructure, the design should take the following into account:

  • the system is based on an underlying database (MS SQL Server techology)
  • entities and relationships are developed using the Entity Framework provided by the .NET 4.0 platform
  • an existing C# framework that allows visual design of computational processes using the Entity Framework is provided (see image below)
  • the entire system will be developed in C#
  • development will take place on a Windows Server 2008 platform, using Visual Studio

Provisions

The following items are available at the project start and during the execution:

  • collection of images with meta-data already gathered by medical specialists
  • visual designer framework (C#) for creating visual representations of computational processes
  • computational platform with pre-installed database and programming tools
  • support from medical specialists to validate the development

Deliverables

  • a working system (software source code and demonstrable executable on a provided platform). The system
  • a report describing the architecture and design of the system
  • a short manual describing the operation of the system

Evaluation

The project will be evaluated as follows:

  • quality of provided solution (ease of use, quality of documentation, coverage of functional and non-functional requirements): 60%
  • quality of provided technical documentation (design, architecture) and code: 20%
  • end presentation: 20%

Participants

This project can be done by one or two students.

Contact person

This project is supervised by prof. Alexandru Telea.