Electronic Proceedings:

The forth workshop of the series Microscopic Image Analysis with Applications in Biology was held in September 2009 at the National Library of Medicince at the NIH campus in Bethesda. The goal of this workshop is to bring together researchers that work on the emerging interface of engineering and biological research. This is why we were particularly grateful that the National Libraray of Medicine provided the Lister Auditorium for the venue. In particular we would like to thank Dr. Terry Yoo (NIH NLM) for making the local arrangements. In addition we would like to thank Dr. Raghu Machiraju for coordinating the paper reviews and Dr. Stephen Lockett for organizing the panel discussion.

The workshop program included a panel discussion, 16 oral presentations, as well as 12 posters. As in previous years we encouraged both full paper as well as abstract submissions. In addition the workshop included a session on invited talks on Atlas Based Imaging which was organized by Dr. Hanchuan Peng. The focus of the panel discussion was the topic robust imaging. Problems such as the inherent variation of biological specimen and their quality make the problem of quantitative imaging particularly challenging. Overall the group broadly acknowledged the need for annotated standardized data sets for algorithm validation and testing.

The convergence of technical developments that enable automated microscopy imaging at higher resolution and throughput and life sciences applications that require the analysis of a very large number of samples radically changes the traditional role of microscopy. As opposed to analyzing a very limited number of samples manually, it is now possible to automatically analyze a large number of biological samples at the cellular and sub-cellular scale and monitor their dynamics over time. We are seeing more and more enterprise level applications where specialists in biology, biomedical imaging, and bioinformatics work closely together. Such projects will not only provide biologists with an unprecedented amount of quantitative information, they will also allow the investigation of the inherent variation of biological systems of interest.

Algorithms that allow the automatic analysis of such data sets are becoming a crucial component of microscopy workflow. These datasets pose a number of challenges that are very distinct from conventional clinical imagery in their size and abundance, the detail of relevant features, and their statistics. Sophisticated algorithms are necessary to process such imagery and extract biologically relevant features and information.

While certain applications of high-throughput microscopy of very simple biological model systems are already established, biologists are still exploring the potential of this automated approach. The study of complex model systems, the analysis of whole organisms of small critters, and in-vivo models will pose novel challenges that need to be addressed. The challenges increase many-fold when animal systems for disease and cancer are considered. In 2007 the workshop was held as an independent one-day event in Piscataway, NJ. A one-day workshop was held in conjunction with MICCAI in New York last year.

For the organizers

Jens Rittscher
Tolga Tastizien