Related Experiment Video
Updated: Dec 19, 2025
09:37
Combining Reflectance Confocal Microscopy with Optical Coherence Tomography for Noninvasive Diagnosis of Skin Cancers via Image Acquisition
Published on: August 18, 2022
2.8K
Proposal of the CAD System for Melanoma Detection Using Reconfigurable Computing
Wysterlânya K P Barros1, Daniel S Morais1, Felipe F Lopes1
1Laboratory of Machine Learning and Intelligent Instrumentation, Federal University of Rio Grande do Norte, Natal, RN 59078-970, Brazil.
Sensors (Basel, Switzerland)
|June 7, 2020
Summary
This study introduces dedicated hardware for real-time cancer detection using Field-Programmable Gate Arrays (FPGA). The system combines Artificial Neural Networks (ANN) and Digital Image Processing (DIP) for accurate melanoma classification.
Area of Science:
- Biomedical Engineering
- Computer Engineering
- Artificial Intelligence
Background:
- Early cancer detection is crucial for effective treatment.
- Existing methods for skin lesion analysis can be computationally intensive.
- Field-Programmable Gate Arrays (FPGA) offer potential for high-speed, low-power embedded systems.
Purpose of the Study:
- To develop and evaluate dedicated hardware for real-time cancer detection.
- To integrate Artificial Neural Networks (ANN) with Digital Image Processing (DIP) on an FPGA.
- To compare the performance of the hardware implementation against a software-based approach.
Main Methods:
- Utilized Digital Image Processing (DIP) for feature extraction from skin lesion images.
- Implemented a Multilayer Perceptron (MLP) Artificial Neural Network (ANN) on an FPGA for classification.
- Validated classification results using an open-access dermatological database.
- Analyzed execution time, hardware resource utilization, and power consumption.
Main Results:
- The FPGA-based hardware achieved real-time performance for cancer detection.
- The integrated ANN and DIP system demonstrated effective melanoma classification.
- The hardware implementation showed advantages in execution speed and power efficiency compared to software.
Conclusions:
- Dedicated FPGA hardware provides an efficient platform for real-time cancer detection.
- The combination of ANN and DIP on hardware is a viable approach for medical image analysis.
- This approach offers a promising solution for improving diagnostic capabilities in resource-limited settings.
