Related Experiment Video
Updated: Jul 17, 2026
06:05
Multiplex Immunofluorescence Combined with Spatial Image Analysis for the Clinical and Biological Assessment of the Tumor Microenvironment
Published on: June 2, 2023
Image Cytometry Data From Breast Lesions Analyzed using Hybrid Networks.
H A Mat Sakim1, N A Mat Isa, Raouf G Naguib
1CELIS, School of Electrical and Electronic Engineering, Universiti Sains Malaysia, Engineering Campus, Pulau Pinang, Malaysia, Tel: +604-5937788 ext. 6605, Fax: +604-5941023,
Summary
This study explored hybrid neural networks for predicting breast cancer axillary lymph node involvement. While accuracy was limited to 69%, high confidence was observed in correct predictions, aiding early breast cancer staging.
Area of Science:
- Oncology
- Biomedical Engineering
- Computer Science
Background:
- Accurate breast cancer staging is critical for effective treatment and therapy selection.
- Axillary lymph node involvement is a key indicator of tumor aggression and disease stage.
- Neural networks show promise in medical prognosis, including breast cancer.
Purpose of the Study:
- To investigate the performance of hybrid neural networks (Multilayer Perceptron and Radial Basis Function) for predicting axillary lymph node involvement in breast cancer.
- To introduce a confidence measurement for the neural network predictions.
Main Methods:
- Utilized hybrid neural networks combining Multilayer Perceptron and Radial Basis Function.
- Input features comprised four image cytometry features from fine needle aspiration samples of breast lesions.
- Evaluated network performance using accuracy and a proposed confidence measurement.
Main Results:
- The hybrid networks achieved a maximum accuracy of 69% in predicting axillary lymph node involvement.
- A significant portion of the correctly predicted cases were associated with a high confidence level from the networks.
Conclusions:
- Hybrid neural networks demonstrate potential for predicting axillary lymph node involvement, though current accuracy requires improvement.
- The proposed confidence measure is valuable for interpreting the reliability of predictions in breast cancer staging.
