Microscope automation : machine vision to support in-cell signal detection
Farah, Hassan (2022)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2022062919170
https://urn.fi/URN:NBN:fi:amk-2022062919170
Tiivistelmä
Automation has become increasingly important in our life. With automation, we can achieve a reduction in cost, time, mistakes, and a multitude of other aspects that hinder the workflow in production.
The objective of this thesis was to use computer vision on fluorescence automated microscopy to help cell imaging in detecting signaling events from cells based on calcium spikes.
Automated microscopy has been used in drug discovery for quite some time, but there has only recently been a rapid advance in automated detection and analysis of cells, which would enable an amplitude of possibilities with the data obtained from cells that are analyzed after they have been detected.
This thesis introduces the processes necessary to achieve functional cell detection and analysis. The cell detection algorithm, which utilizes image segmentation, and the cell analysis algorithm, which employs ROI, are presented in detail with code examples in this thesis. The steps taken to improve the cell detection capability utilized processes in image segmentation. The processes that were used include blurring, greying, binary thresholding, and contour.
In conclusion, the data received from the detected cells offers numerous l possibilities such as decision making, visualizing the data, and predicting outcomes with the use of machine learning to name a few. The collected data is not only limited to the intensity of the cell, but also other attributes including size, color, and shape can also be collected for analysis purposes.
The objective of this thesis was to use computer vision on fluorescence automated microscopy to help cell imaging in detecting signaling events from cells based on calcium spikes.
Automated microscopy has been used in drug discovery for quite some time, but there has only recently been a rapid advance in automated detection and analysis of cells, which would enable an amplitude of possibilities with the data obtained from cells that are analyzed after they have been detected.
This thesis introduces the processes necessary to achieve functional cell detection and analysis. The cell detection algorithm, which utilizes image segmentation, and the cell analysis algorithm, which employs ROI, are presented in detail with code examples in this thesis. The steps taken to improve the cell detection capability utilized processes in image segmentation. The processes that were used include blurring, greying, binary thresholding, and contour.
In conclusion, the data received from the detected cells offers numerous l possibilities such as decision making, visualizing the data, and predicting outcomes with the use of machine learning to name a few. The collected data is not only limited to the intensity of the cell, but also other attributes including size, color, and shape can also be collected for analysis purposes.