Home

Demo/Gallery

Contact Us

Documentation

FAQ

Tutorial

Publication

1, What is ImCellPhen?

Interactive mining of cellular phenotypes – imCellPhen - is an innovative computing paradigm that uses intelligent human-computer interfaces to facilitate the application of the HCS technology in biomedical research. It's a a new computing paradigm that combines unsupervised pattern mining techniques, P-VDE interfaces, and CBIR-RF techniques to boost the exploitation capacity of the HCS technology and facilitate its application to biomedical research.

2, What is CBIR-RF?

CBIR-RF is the content-based image retrieval (CBIR) with Relevance Feedback (RF) techniques to HCS image analyses. CBIR, a technique that uses images as queries to search for similar ones in an image database, has been an active research field in the last decade as a result of advances in the Internet and high-throughput digital image sensor technologies (Furht, Smoliar et al. 1995; Rui, Huang et al. 1999; Smeulders, Worring et al. 2000).

3, Why do we represent clusters by their representative images? Do we need to visualize the relations among clusters?

It is very important to represent clusters by their representative images and visualize the relations among clusters for the following
two reasons. First, clustering analysis results may not be perfect if image features are not comprehensively extracted or the assumptions made by clustering analyses do not match the reality very well. Images and the relationship among clusters can help users make correct judgments. Second, as the old proverb says: A picture is worth a thousand words, it is more effective for users to explain themselves to developers via images – the main subjects of the concern. For example, biologists can use images to illustrate which image features are biological important but are neglected by computational biologists.

4, What screens will ImCellPhen benefit?

Although our initial focus will bear on Drosophila neurons, the techniques, tools and system developed will benefit any HCS screens and thus will become valuable utility for the community to use the HCS technology to explore complicated cellular morphology on an unprecedented massive scale.