This project demonstrates the use of the K-means clustering algorithm for vessel segmentation in Fundus Autofluorescence (FAF) images. K-means is an unsupervised machine learning method used for clustering data when the labels are unknown. This approach is ideal for segmenting images based on patterns and can be applied to medical image analysis, such as detecting neovascularization or quantifying leakage.

In this case, K-means clustering is used for pre-processing FAF images before performing further analyses like neovascularization indices or leakage quantification. Requirements

The following libraries are required to run the code:

Python 3.6 or higher
OpenCV
NumPy

References

A. O’Neill, “Implementing the K-means algorithm in Python,” LinkedIn, link.