One of the questions I’ve been asking through my research and clinical work is deceptively simple: could digital tools help identify when someone is going through perimenopause — even if they don’t realise it themselves?
Perimenopause often presents subtly in primary care: vague fatigue, anxiety, changing cycles, poor sleep. These symptoms are frequently documented but rarely connected unless explicitly asked about. With so much rich data in GP records, is it possible for artificial intelligence to spot patterns earlier than we do?
This isn’t science fiction. As part of my current NIHR-funded work, I’ve been exploring how we might use structured and unstructured data to support earlier, more equitable diagnoses of perimenopause in general practice. It’s exciting – and humbling – to consider what computers might notice that we, in our time-limited consultations, may miss.
But it’s not just about detection. These tools must be designed with care, clinical context, and ethics front of mind. Bias, privacy, and explainability all matter. We can’t outsource judgment — but we can augment it.
If we get this right, we could offer thousands of women earlier support, better information, and a sense of being truly heard. That, to me, is the promise of responsible digital innovation in healthcare.
This idea was also the subject of my TEDx talk in Bristol — watch it here.