A team of mathematicians from RUDN University has successfully developed a cutting-edge artificial intelligence (AI) model capable of identifying breast cancer in histological samples with an unprecedented accuracy of nearly 100%. The breakthrough was achieved by incorporating an attention mechanism module, which significantly improved the model’s ability to pinpoint key features and areas of interest in the images. The results of this groundbreaking research were published in the esteemed scientific journal, Life.
Early detection of breast cancer is crucial for improving patient prognosis and overall treatment outcomes. However, histological examination, which is currently the gold standard for diagnosis, is prone to subjective interpretation and the variability in sample quality, resulting in potential inaccuracies and misdiagnosis. To combat these challenges, the interdisciplinary team of mathematicians collaborated with experts from China and Saudi Arabia to design an AI model that could enhance the accuracy of cancer recognition in histological images.
“The implementation of computer-based classification and analysis of histological images is of paramount importance in improving diagnostic accuracy and alleviating the burden on medical professionals,” said Dr. Ammar Muthanna, the Director of the Scientific Center for Modeling Wireless 5G Networks at RUDN University.
The mathematicians rigorously tested various convolutional neural networks, augmenting them with attention modules specifically designed to detect objects in images. The final model, which combined the DenseNet211 convolutional network with attention modules, achieved an impressive accuracy rate of 99.6%. Importantly, the researchers observed that the scale of the images had a significant impact on the recognition of cancerous formations, highlighting the need for further consideration of appropriate image approximations in real-world applications.
Dr. Muthanna emphasized the importance of attention mechanisms in medical image analysis, as they enhance feature extraction and overall model performance. The attention modules enabled the AI model to focus on relevant areas of the images and extract crucial information, thus revolutionizing the analysis of medical images.
This breakthrough in AI technology has the potential to revolutionize breast cancer diagnosis and significantly improve patient outcomes. By reducing the reliance on subjective interpretations and enhancing accuracy, this AI model holds promise for transforming the field of medical imaging and advancing the early detection and treatment of breast cancer.
The source of the article is from the blog krama.net