Predicting recurrence of PMP arising from LAMN using Machine Learning

  • Research type

    Research Study

  • Full title

    Prediction, modelling and understanding the risks of recurrence of Pseudomyxoma peritonei arising from appendiceal epithelial mucinous neoplasms after Cytoreductive surgery and Hyperthermic intraperitoneal chemotherapy using Machine learning.

  • IRAS ID

    342832

  • Contact name

    Thomas D Cecil

  • Contact email

    tom.cecil@hhft.nhs.uk

  • Sponsor organisation

    Hampshire Hospitals NHS foundation Trust

  • Duration of Study in the UK

    1 years, 10 months, 29 days

  • Research summary

    Pseudomyxoma Peritonei (PMP) is a rare cancer of the abdomen originating from the appendix. Despite advances in treatment, the recurrence rate is very high. It is not yet understood what the risk factors are for this recurrence rate. Recent evidence suggests that modern Machine Learning (ML) techniques will better predict risk factors than traditional methods.
    This study therefore looks to develop a trustworthy and explainable model to predict recurrence and survival rates of PMP in patients who have undergone complete cytoreduction (the most common form of treatment). We will do this through three stages:
    Large datasets from Hampshire Hospitals, using over 30 variables, will be analysed by the University of Southampton to create a model framework
    A focus group of PMP patients will be consulted on attitudes and acceptability of using AI for guidance on treatment
    Clinician interviews will be conducting to ascertain attitudes and acceptability of using AI for guidance on treatment amongst staff
    Results from points 2 and 3 will be fed into the model to facilitate trust in this AI solution.

  • REC name

    Yorkshire & The Humber - Bradford Leeds Research Ethics Committee

  • REC reference

    25/YH/0049

  • Date of REC Opinion

    19 Mar 2025

  • REC opinion

    Further Information Favourable Opinion