Online ISSN: 2515-8260

Author : A, Chandan B K, Dr. Jayachandran


IDENTIFYING GLOMERULI IN HUMAN KIDNEY TISSUE IMAGES USING PATTERN RECOGNITION METHODS

Chandan B K, Dr. Jayachandran A

European Journal of Molecular & Clinical Medicine, 2022, Volume 9, Issue 3, Pages 11208-11217

Glomerular disease is the result of conditions that affect a particular part of your kidneys called glomeruli. Glomeruli is a small network of blood vessels that are the "cleansing units" of your kidneys. They filter the waste and remove excess fluid from your bloodstream. “an approximate of one million glomeruli in each human kidney”. Normal glomeruli typically range from 100-350μm in diameter with a roughly spherical shape when glomeruli are damaged and unable to function properly, it is called glomerular disease or glomerular malfunction. Glomerular malfunction can damage your kidneys which could also eventually lead to kidney failures. Therefore, this paper presents a convolutional neural network model for image segmentation which uses pattern recognition methods to identify the position of glomeruli in the human kidney tissue images. The paper aims at the comparison between neural networks namely faster region based convolutional neural network and mask region based convolutional neural network for effective identification of glomeruli in human kidney tissue images