My research work (as a Ph.D. student and then a post-doc) on computer vision (image processing, segmentation, characterization) and machine learning has mainly focused on cell characterization and classification. I have successfully created new methods of shape and texture characterization to efficiently describe cells acquired from different sources. My first major research project was to propose new statistical methods to automatically classify cells of patients affected by the Progeria disease. This work was extended and generalized to describe the mitosis evolution of cancerous cells. I have also made significant contributions in pathology pattern recognition in retinal images of patients affected with diabetic retinopathy. Since, my work has become a standard in radiomics.

After I moved to Portland in 2013, my work has mostly focused on cell segmentation and characterization from different imaging modalities. I have successfully developed a new innovative method to automatically detect the immunogold particles, providing biologists an efficient tool of quantification. I have also worked on automatic electron microscopy (EM) image segmentation to highlight the heterogeneous cell structure information provided in EM images. I have extracted texture features from breast DCE-MRI parametric maps for the early prediction of breast cancer tumor response. Since 2016, I have been developing segmentation and characterization pipelines for tissues stained with multiplex immunohistochemistry (mIHC) and cyclic immunofluorescence (cycIF).