PREDICT ME

  • Research type

    Research Study

  • Full title

    Predicting Difficulty of Rectal Cancer Surgery through Preoperative MRI scans using Machine Learning Models

  • IRAS ID

    318773

  • Contact name

    Jim Khan

  • Contact email

    jim.khan@porthosp.nhs.uk

  • Sponsor organisation

    Portsmouth Hospitals University NHS Trust

  • Duration of Study in the UK

    1 years, 1 months, 28 days

  • Research summary

    This study is a retrospective, single-centre data analysis study analysing MRI scans and surgical outcomes in rectal cancer patients that underwent rectal cancer surgery. We aim to analyse MRI scans from 500 patients from our center on several anatomical and tumour characteristics and to train a machine learning model on identifying and processing these characteristics. The outcomes will then be pooled with other available data for data analysis to determine relationship and correlation with factors associated with difficult rectal cancer surgery. Outcomes from these analyses will subsequently be used to develop a predictive risk model for preoperatively assessing difficulty of rectal cancer surgery and tested for use in a clinical setting in follow-up trials.

  • REC name

    Yorkshire & The Humber - Leeds West Research Ethics Committee

  • REC reference

    22/YH/0284

  • Date of REC Opinion

    25 Nov 2022

  • REC opinion

    Favourable Opinion