The invention of microscope had a significant impact on biology since it has allowed the exploration of what was previously invisible to the naked eye. In the past century, we have seen rapid development of microscopy techniques. Microscopy has been a qualitative technique for decades. While qualitative observations are the critical and first step toward biological discoveries, modern biology is transforming from qualitative observations to quantitative understanding. The manual processing of microscopy data is subjective, time-consuming, unreliable and in many cases unfeasible. When analyzing cellular processes, we often face terabytes of image/video data. For example, when assessing if drugs/compounds can successfully prevent cancer metastasis (the spread of cancer from one organ to another non-adjacent organ), a typical screen to evaluate the effectiveness of 20 compounds results in the acquisition of at least one million cellular images containing about 200 million cells. To process the data and determine the effect of the compounds on cancer cell proliferation, spreading speed/patterns and the dynamics of interactions between the different proteins involved, quantitative analysis is essential.
Novel imaging analysis methods are making it feasible to extract quantitative data from vivid images. Two closely related components are required: Computational Bioimage Analysis and Quantitative Imaging Informatics. The former refers to the automatic analysis of acquired biological images (2D, 3D or 4D) using computational solutions and the extraction of different features; the latter relates to processing the features and extracting quantitative information, which will be valuable to understanding the underlying biological phenomena and building traceable classification/prediction models. New knowledge, skills, and approaches are urgently required to develop novel, efficient and reliable computational methods for emerging challenges in biological studies. The general question is how to provide timely and suitable solutions for the images/videos acquired in different scientific and industrial applications.
To fill the gaps in the field, the Computational Bioimage Analysis (CBA) Unit is a research and development team in IMCB located at Biopolis, Singapore. Our team focuses on the application of image processing, computer vision, machine learning, pattern recognition and mathematical modeling in biological and biomedical image analysis and informatics.
The CBA Unit at IMCB:
- develops computational, automatic and quantitative image analysis solutions and platforms in the field of biological and biomedical image informatics
- tailor-makes fast prototype bioimage analysis and informatics solutions including, but not limited to, High-throughput Screening (HTS) and High Content Screening (HCS)
- performs analysis of histology image/stacks in 2D/3D and using general bioimage quantification methods.