Multimodal Diagnosis of Anemia using AI Technology V1.0

  • Research type

    Research Study

  • Full title

    Multimodal Diagnosis of Anemia using AI Technology

  • IRAS ID

    339575

  • Contact name

    Miguel Rodrigues

  • Contact email

    m.rodrigues@ucl.ac.uk

  • Sponsor organisation

    University College London

  • Clinicaltrials.gov Identifier

    Z6364106/2024/02/125, UCL Data Protection Registration Number

  • Duration of Study in the UK

    1 years, 0 months, 30 days

  • Research summary

    Anemia is a widespread blood disorder affecting approximately 1.6 billion people globally. The worldwide prevalence of anemia among all age groups in 2019 was 22.8%. Traditional anemia diagnosis involves measuring hemoglobin concentration through venous blood samples, a process that is invasive and requires a clinical or outpatient setting. Moreover, this method can cause pain, localized infection, and generate medical waste.
    We are currently developing a AI model that is able to ingest multiple data modalities -- ranging from EHR data to PPG sensor data to conjunctival images -- to achieve personalized anemia prediction, personalized treatment planning and personalized monitoring. This model will be non-invasive, fast and accurate.
    We plan to recruit about 50 participants from UCLH. Each participant will be involved for data collection 3-4 times over two weeks, with each collection lasting about 10 minutes. All data collection is non-intrusive and does not pose any risk.

    At the end of the study, we will build a multimodal AI model to diagnose anemia based on the collected data. Successful development of this model will provide strong support for intelligent mobile device-based healthcare systems. It will significantly improve the efficiency of the healthcare system and patient wellbeing in England.

  • REC name

    London - Stanmore Research Ethics Committee

  • REC reference

    24/LO/0652

  • Date of REC Opinion

    9 Oct 2024

  • REC opinion

    Further Information Favourable Opinion