Creating your question(s)
Company questions posed to AIMday will be communicated via this website to academics and researchers throughout the University of Edinburgh, allowing them to select the questions they feel they can add most value to according to their individual expertise.
When registering your question, please ensure that you can facilitate access to the necessary data or details to develop a solution to your challenge. When phrasing your question(s), please ensure that they are concrete and clear (with no abbreviations) and that they can be discussed openly. This style of question works best for the AIMday format.
It helps to add context and background to your question(s), so that it is clear what exactly your challenge is and how our academics could potentially help find the solution.
We also recommend that at least two company representatives are available to join the discussion of each your questions in order to have a better balance in the session. The company representatives should be well versed in the topic of the question(s) submitted and be prepared to provide further context and background.
We welcome up to three questions from each participating company or organisation.
Some example questions are listed below. You are welcome to contact us if you need help phrasing an appropriate question.
- How can synthetic or simulated imaging data be most effectively used to train AI models for rare conditions, while ensuring clinical validity and regulatory acceptance?
- What are the technical and workflow requirements for implementing remote or assisted MRI scanning in rural or resource-limited settings?
- How can multimodal data (such as ultrasound, ECG, and motion signals) be combined to generate earlier, more accurate cardiovascular risk scores without extending clinic time?
- What strategies could reduce the need for sedation in paediatric MRI by improving acquisition speed, noise reduction and AI-based motion correction?
- Can an AI-assisted clinical decision tool be developed to automatically detect and prioritise high-risk findings from routine imaging and sensor data, integrating results into electronic health records under robust governance?
- What design choices enable a miniaturised, biocompatible sensor for in situ measurement during procedures?
- How can on device AI guide ultrasound acquisition under tight compute and power limits while preserving accuracy?
- How can imaging plus passive sensing speed up detection of early cognitive decline and trigger timely review?
For additional inspiration please see questions submitted to AIMday Medical Imaging & Diagnostics Toronto 2024