Grant: Discovery Grants Program - Individual

Breast cancer is the most frequent type of cancer diagnosed in women. Early diagnosis and prognosis are crucial for breast cancer patients: with early diagnosis more therapeutics options can be considered while with early prognosis personalized therapy adjustments can be administrated effectively. Tumour formation and its degeneration in response to treatment alter tissue bio-acoustical and bio-mechanical characteristics. Therefore, novel AI-enabled imaging technologies that permit simultaneous quantification and comprehensive analysis of tissue bio-mechanical and bio-acoustical properties can potentially facilitate early diagnosis and prognosis for cancer patients. This research program aims at developing innovative techniques to be integrated within an intelligent ultrasound (US) platform for rapid characterization and monitoring of breast tissues non-invasively. Our short-term objectives are: (1) to adapt and optimize the methodologies for enhanced US elastography and quantitative US (QUS) multi-parametric imaging of the breast that quantify the tissue biomechanical and bio-acoustical properties, respectively; (2) to design, develop and investigate novel data-driven models for ultra-fast generation of QUS and elastography images; (3) to design, develop and optimize state-of-the-art AI architectures for comprehensive and high-throughput analysis of these parametric images concurrently. Utilizing US as the imaging modality to collect raw data, our proposed framework has the advantage of being safe, inexpensive, and portable, besides having rapid imaging speed with high spatial resolution. The techniques developed in this program will be evaluated through computer simulations, comprehensive studies on tissue-mimicking phantoms, and assessments on available clinical datasets. Over the long-term, our goal is to establish accessible AI-enabled US-based technologies for diagnosis and therapy response monitoring of various types of cancer to improve equity in cancer control. To achieve this, we aim to characterize ultrasonically various healthy and cancerous tissues at different stages of disease, and at different stages of response development to anti-cancer therapies. A simultaneous characterization of different tissue properties enables us to investigate tumour formation/degeneration processes from various points of view at cellular and tissue levels. We expect the complementary information provided through a comprehensive analysis of different tissue characteristics will increase the accuracy and robustness in cancer diagnosis/prognosis. The proposed platform, therefore, can eventually facilitate cancer detection at early stages and permit therapy response evaluation early after the treatment initiation. The benefits offered by the proposed platform can facilitate an equitable access of the eventual technology for remote and poorer communities in Canada, which tend to include more Indigenous residents, people of color, and marginalized groups.