The majority of patients diagnosed with pancreatic cancer have symptoms and an incurable disease at the time of diagnosis. Cancer in the pancreas is one of the deadliest forms of cancer and this project is a new way of trying to improve the prognosis for the patients. The purpose of this project is to use data from UK Biobank to investigate the possibility of identifying potentially curable pancreatic cancer, before patients get any or some of the high-risk symptoms we know today. We will use large amounts of data and new analytical tools, machine learning (ML) and artificial intelligence (AI). Machine learning is a mathematical method that uses large amounts of data to search for patterns that humans cannot see. With the help of ML and AI, we will look for previously unknown patterns for genetic profile, lifestyle, signs, symptoms, etc. that reveal pancreatic cancer at an earlier stage. With data from UK Biobank, we strive to combine phenotypic and genotypic information to identify a pre-diagnostic risk profile for pancreatic cancer.
Despite advances in diagnostic and surgical techniques and the identification of promising biomarkers, impact on survival for pancreatic cancer has been limited. Previous studies have shown improvement in survival when pancreatic cancer is diagnosed in earlier stages. Novel diagnostic strategies and more profound understanding of the disease are therefore urgently needed.
Timeline of Research Project:
Request for data December 2020
Application for ethics approval December 2020
Processing and analysis of data 2021
Publication of results in international peer-reviewed scientific journal 2022