Computer-assisted medical interventions (CAMI) provide assistance to a physician (surgeon, interventional radiologist, etc.) to make the best diagnostic or therapeutic decisions for their patient and to execute the interventions safely and clinically efficiently. This entails constructing and continuously updating a digital twin using patient-specific multimodal information (such as images and signals) and prior knowledge (including atlases, models, and statistical data) and leveraging these for intervention planning and simulation. The execution phase is facilitated by various medical devices, whether robotic or not. Clinical scenarios featuring deformable and dynamic organs necessitate real-time monitoring and adjustment of the intervention plan. Achieving minimal invasiveness mandates the integration of devices and sensors to enhance perception and dexterity. Two illustrative examples of image-based prostate cancer treatments (laparoscopic prostate surgery and ultrasound-based brachytherapy) are used to provide insight into the current state of the field and to address existing challenges and bottlenecks of CAMI.