The NIH/NIGMS
Center for Integrative Biomedical Computing

Computational Microscopy of Skin Simulations

Prof. Charles DiMarzio and Dr. Milind Rajadhyaksha

Prof. DiMarzio is the Director of the Optical Sciences Laboratory in the Electrical and Computer Engineering Department at Northeastern University. Dr. Rajadhyaksha is a Research Scientist working with Prof. DiMarzio and also with the Dermatology Department of Memorial Sloan Kettering Cancer Center in New York. Prof. DiMarzio's primary interest is in development of new microscopes; he was the PI on a $750,000 award from the Keck Foundation to develop an integrated multi-modality microscope that is now functioning. Dr. Rajadhyaksha is the PI on the NIBIB grant cited above and an investigator on the NCI grant, and his primary interest is in the development of new, portable confocal microscopes and associated algorithms for detection of skin cancer at the bedside or in the clinic.

As they have developed new instruments, these two scientists have become increasingly interested in the idea of "computational microscopy", as they explain in their accompanying support letter. They have begun a pilot project to build a computational model that would allow them to simulate optical propagation in a realistic skin model, including appropriate layers, variety of cell types, cell sizes, inclusions, etc. The eventual goal is to use confocal images to derive the parameters of the model, and cross-scale aggregation to derive associated larger-scale phenomena. The underlying computational engine is a finite-difference solver developed in their lab. Finally they need interactive three-dimensional "zoom" visualization capabilities that will allow the user to probe across scales in a region interactively, change fine-scale parameters, observe the effect on coarser-scale parameters, re-run simulations, and visualize the new results.

Thus this project, although different in application area from most of CIBC's applications, has several technical connections: the extraction of model parameters from images, the cross-scale nature of the simulation process, and the need for an integrated image processing-simulation and modeling-visualization environment in particular. Prof. DiMarzio and Prof. Brooks have a record of successful collaboration and Prof. DiMarzio has had an interest in BioPSE for some time.