Related Experiment Video
Updated: Jul 7, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.8K
Convolutional Neural Network Model for Segmentation and Classification of Clear Cell Renal Cell Carcinoma Based on
Vlad-Octavian Bolocan1,2, Mihaela Secareanu2, Elena Sava2
1Department of Fundamental Sciences, Faculty of Midwifery and Nursing, University of Medicine and Pharmacy "Carol Davila", 050474 Bucharest, Romania.
Journal of Imaging
|December 22, 2023
Summary
This study demonstrates an AI algorithm
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Computed tomography (CT) imaging presents challenges in differentiating renal cell carcinoma (RCC) from benign tissues and determining subtype.
- Accurate diagnosis is crucial for effective treatment and improved patient outcomes.
Purpose of the Study:
- To evaluate an AI algorithm's capability to enhance the identification and classification of renal cell carcinoma (RCC) using CT imaging.
- To assess the algorithm's potential to improve patient treatment strategies and outcomes.
Main Methods:
- Utilized the European Deep-Health toolkit, employing Convolutional Neural Networks (CNNs) with U-net for image segmentation and resnet101 for classification.
- Assessed clinical efficiency through kidney and tumor segmentation (Dice score) and RCC categorization accuracy.
- Emphasized data preparation and processing for algorithm implementation.
Main Results:
- Achieved kidney segmentation accuracy of 0.84.
- Obtained a mean Dice score of 0.675 for tumor segmentation.
- Demonstrated a renal cell carcinoma classification accuracy of 0.885.
Conclusions:
- The developed AI technique shows significant potential in improving the diagnosis of kidney pathology, specifically renal cell carcinoma.
- The algorithm's performance in segmentation and classification suggests its utility in clinical settings for enhanced diagnostic accuracy.
