An updated overview of radiomics-based artificial intelligence (AI) methods in breast cancer screening and diagnosis

Overview

• Current breast cancer (BC) imaging methods have limited sensitivity, specificity, and positive predictive power
• Artificial intelligence (AI) and novel computational techniques like radiomics show promise for improving BC diagnosis and subtype classification
• This review article discusses the radiomics workflow and recent studies on using radiomics for BC diagnosis across different imaging modalities

Plain English Explanation

Doctors currently use various imaging tests like mammograms, MRI, and ultrasound to diagnose and understand breast cancer. However, these methods aren't always completely accurate - they can sometimes miss cancers or incorrectly identify non-cancerous growths as cancerous.

Recent advancements in artificial intelligence (AI) and medical imaging have created new opportunities to improve breast cancer diagnosis. One promising technique is called "radiomics," which uses complex computer algorithms to analyze medical images and extract detailed quantitative measurements. These radiomics features can then be used to detect signs of cancer and differentiate between breast cancer subtypes more accurately than traditional imaging alone.

The review article discusses how radiomics has been applied to different imaging tests for breast cancer, including mammograms, MRI, 3D breast imaging (digital breast tomosynthesis), and specialized contrast-enhanced mammograms. The authors also explore the current challenges in making radiomics techniques more reliable and specific for clinical use, with the goal of achieving better breast cancer classification and diagnosis in the future. As the field of AI and medical imaging continues to advance, it holds great promise for providing more accurate and personalized care for breast cancer patients.

Technical Explanation

This review article provides an overview of the current state of radiomics, a computational technique that extracts quantitative features from medical images to aid in disease diagnosis and classification. The authors focus on the application of radiomics to the diagnosis and subtyping of breast cancer (BC) across various imaging modalities, including MRI, mammography, contrast-enhanced spectral mammography (CESM), ultrasound, and digital breast tomosynthesis (DBT).

The radiomics workflow typically involves the following steps: image acquisition, segmentation, feature extraction, feature selection, and machine learning model building. The authors discuss how radiomics features, such as tumor shape, texture, and intensity, can provide additional information beyond standard visual assessment of medical images. Several recent studies have demonstrated the potential of radiomics to improve the sensitivity and specificity of BC detection and classification compared to traditional imaging approaches.

For example, radiomics analyses of MRI data have shown promise in differentiating between BC subtypes and identifying features associated with patient outcomes. Similarly, radiomics features extracted from mammograms and CESM images have been used to enhance breast cancer diagnosis and improve classification accuracy.

The authors also highlight current challenges in radiomics, such as the need for larger, more diverse datasets, standardized image acquisition and analysis protocols, and the development of explainable AI models to improve the interpretability of radiomics findings. Strategies to address these challenges, such as the use of multimodal approaches and the integration of radiomics with other clinical and genomic data, are discussed.

Critical Analysis

The review article provides a comprehensive overview of the current state of radiomics in the context of breast cancer diagnosis and subtyping. The authors effectively highlight the potential of radiomics to improve the sensitivity and specificity of imaging-based BC detection and classification, which could lead to earlier and more accurate diagnoses.

However, the article also acknowledges the significant challenges that need to be addressed, such as the need for larger, more diverse datasets, standardized protocols, and the development of explainable AI models to improve the interpretability of radiomics findings. These are critical issues that must be tackled to ensure the reliable and widespread clinical deployment of radiomics techniques.

Additionally, while the article discusses the use of radiomics across various imaging modalities, it does not provide a detailed comparison of the performance and trade-offs of these different approaches. A more in-depth analysis of the strengths and weaknesses of radiomics in the context of each imaging modality could help guide clinicians and researchers in selecting the most appropriate methods for their specific use cases.

Furthermore, the article does not address potential ethical and privacy concerns associated with the increasing use of AI and computational techniques in medical imaging. As these technologies become more prevalent, it will be important to consider the implications for patient data privacy, algorithm bias, and the transparency of decision-making processes.

Conclusion

This review article highlights the promising potential of radiomics, a computational technique that extracts quantitative features from medical images, to improve the diagnosis and subtyping of breast cancer. The authors discuss how radiomics has been applied across various imaging modalities, including MRI, mammography, and 3D breast imaging, and has demonstrated the ability to enhance the sensitivity and specificity of BC detection and classification compared to traditional imaging methods.

However, the article also acknowledges the significant challenges that must be addressed, such as the need for larger, more diverse datasets, standardized protocols, and the development of explainable AI models. As the field of AI and medical imaging continues to advance, addressing these challenges will be crucial to realizing the full potential of radiomics and other AI-driven techniques for providing more accurate and personalized care for breast cancer patients.