Semiautomated Image Analysis of High Contrast Tissue Areas Using Hue/Saturation/Brightness Based Color Filtering

Introduction: Quantification and morphological assessment of various tissue elements have numerous applications in fundamental and clinical research. Digital morphometry, in contrast to other morphologic methods, uses personal computers and specific software, to perform precise and highly reproducible results. Additionally, it delivers results in mathematical format. The aim of our study was to develop an open access digital morphometry method for measuring different parameters of various high contrast tissue elements and to elaborate a general work-around for digital morphometry study and data management.
Materials and methods: We used three different types of tissue samples and staining procedures: (1) Diffuse Large B-cell Lymphoma specimens, (2) various stage liver fibrosis specimens and (3) transversely sectioned skeletal muscle tissue to develop a digital morphometric analysis. Image analysis was performed using ImageJ software.
Results: We developed an intuitive and easy to use work-algorithm that fits generic demands. We split the algorithm into three phases, each requiring a different approach and workaround. Using the presented method we were able to quantify the proportion of CD34 positive areas in the DLBCL specimens, the vascularity of this type of lymphoma may be quantified; similarly, this method is optimal for determining the extension of fibrosis in liver specimens; and finally, morphometric analysis of striated muscle fibers was achieved.
Conclusions: We conclude that the use of ImageJ with semiautomatic color segmentation is a reliable and practical way of performing various morphometric measurements. In addition, we are confident that such methods of digital morphometry could have future applications in other areas of pathology and histology.