BACHELOR Assignment
Assessment of hearth tissue quality via optical flow
Type: Bachelor CS
Period: 01/04/2021 - 30/06/2021
Student: Tafuro, M.
If you are interested please contact :
Introduction:
Tissue culture is the process of growing tissues or cells under controlled conditions, in an artificial medium separated from the organisms, in the lab. Different mediums (i.e. the chemical compounds used for growing the cells) determine different growing processes and quality of the final grown tissue.
AST and BIOS labs are working on a platform allowing for engineering heart tissues (EHT). This platform combines biology, micro-fabrication techniques and image processing to analyze the contraction of the EHT. Various combinations of compounds are tested in the lab, and the quality of each tissue is inspected visually by means of a microscope. A large number of videos is recorded for all the tissues, and automatic and robust visual analysis to select the best result tissues is required.
Assignment:
Optical flow is the pattern of motion in consecutive frames of a video caused by the movement of the objects contained in the scene or of the camera [1]. It is represented as a vector of the displacement of the points in consecutive frames, or as a vector of the apparent velocities [2].
The assignment is to think of an algorithm that analyzes the motion of the engineered heart tissues in the videos via optical flow and categorizes them into groups on the basis of the characteristics of the motion. Unsupervised learning techniques, e.g. clustering, will be explored to categorize the tissue videos.
References:
[1] Baker, Simon; Scharstein, Daniel; Lewis, J. P.; Roth, Stefan; Black, Michael J.; Szeliski, Richard (March 2011). "A Database and Evaluation Methodology for Optical Flow". International Journal of Computer Vision. 92 (1): 1–31
[2] Eddy Ilg, Nikolaus Mayer, Tonmoy Saikia, Margret Keuper, Alexey Dosovitskiy, Thomas Brox, FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks, CVPR 2017