What is it about?
Deep learning is a feasible assisting tool for pathologists in leveraging the analysis process of kidney cancer tissues. Unlike conventional inspection method which is labor-intensive and prone to human error, the use of deep learning can greatly reduce biases and provide accurate predictions with minimal supervision, disregards the complex nature of pathological images.
Featured Image
Photo by National Cancer Institute on Unsplash
Why is it important?
Pretrained Convolutional Neural Network (CNN), particularly EfficientNet architecture achieved the state-of-the-art classification results on annotation-free kidney cancer dataset (CPTAC-CCRCC). The outstanding accuracy (97%), specificity (94%), F1-score (98%) and AUC (96%) achieved in this study suggests its reliability as an automatic diagnosis system for assisting the pathologists in analyzing the kidney tissues in a more efficient way.
Perspectives
Apart from just achieving better classification results, the most important part of this project is how can it impacts the world by having effective detection of renal cancer, such as improving patients survival rate and reducing economic burdens. With such clinical assisting tool, we believe that the clinical procedures will be much improved. Of course, kidney cancer will still haunt us in the future, but with deep learning, this time, we will be one step ahead!
Mr Jia Chun Koo
Universiti Tunku Abdul Rahman
Read the Original
This page is a summary of: Deep Machine Learning Histopathological Image Analysis for Renal Cancer Detection, March 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3532213.3532313.
You can read the full text:
Resources
Contributors
The following have contributed to this page
