Cancer MIA

We focus on developing computer vision and machine and deep learning techniques to automatically interpret medical images for cancer applications, e.g.

• analysis of super-resolution single molecule localization microscopy data to understand cellular architecture of of prostate cancer cells (2018 Nature Scientific Reports paper)
• histopathology based automated subtype classification of ovarian and breast cancer
• quantitative imaging from PET/CT (relation to ct-DNA) for head and neck cancer
• brachytherapy dose accumulation related to cervical cancer
• skin dermoscopy image analysis for melanoma diagnosis
• image guided robotic surgery and augmented reality for kidney cancer (partial nephrectomy)
• enhancement and analysis of confocal laser endoscopy for gastrointestinal cancer diagnosis
• deep learning based image reconstruction of diffuse optical tomography for breast cancer detection
• brain tumour management from MRI (external link)