MRI Prostate
Generation of a foundation model for automated reporting of prostate MRI
-
Prostate cancer is a very common disease, and the second commonest cause of death from cancer for men in the UK. Diagnosing prostate cancer has recently changed to a pathway that requires a MRI scan to help find suspicious lesions within the prostate that can then be sampled using a needle (a biopsy). This has much improved the diagnosis in the UK and internationally. About 200,000 men in the UK will currently have an MRI scan of their prostate every year. However, there is a real shortage of trained experts to provide reports for these studies. This results in delays to diagnosis across the country and also sometimes incorrect reporting which can miss important cancers or cause patients to have unnecessary biopsies (with potential for infection and bleeding as two main risks of biopsy). Given the large numbers of men being investigated there is also a large healthcare system cost to diagnosing prostate cancer and avoiding unnecessary procedures would significantly reduce costs.
Based on the wealth of prostate MRI expertise at UCL and UCLH and the number of studies that we have done in men since 2004, we will develop the models to support an AI based automated reporting system that can address the shortage of trained radiologists in the UK and world wide. We also plan to develop along the way AI tools that radiologists can use to improve workflow and avoid incorrect interpretation of MRI studies.
Initially we aim to use images that we hold in our electronic health record (approximately 5000 prostate MRI studies). This will be used to explore and develop initial automated reporting and tools. These will then need to be further refined and tested before any of them can be used at UCLH and at other sites. This initial request for data sharing will seed the developmental process.
-
Veeru Kasi
Professor and Honorary Consultant in Urology
Shonit Punwani
Professor of Magnetic Resonance and Cancer Imaging
Yipeng Hu
Associate Professor, Dept of Med Phys & Biomedical Eng
yipeng.hu@ucl.ac.uk
View Our Synthetic Data:
Number of images exported:
13,271
The information you see here is synthetic data. It’s not real data, does not contain real patient details, and cannot be traced back to any real patients. It is only designed to look like real health records.
We created these data to show the kind of information used in a research project called Generation of a Foundation model for automated reporting of prostate MRI. This study aims to develop the tools to address the shortage of trained radiologists in the UK and worldwide, and to improve workflow and avoid incorrect interpretation of MRI studies.
Because this data is randomly generated using a tool called datafaker, some parts may not make sense — for example, a birth date might appear after a death date. That’s because the columns are made separately and don’t always link together in a realistic way.
Please note that this data is part of our preliminary version of synthetic datasets. We’re actively improving our process so that over time, more datasets will be available, and the data will look more and more like real-world data, without ever containing any real patient details.
This dataset is only for demonstration and learning purposes. Any similarity to real people is purely coincidental.
Number of free text reports
1,212,505
Number of patients’ structured data:
8,157
How to browse our synthetic data:
1) In the embedded table above, click the ‘view’ button next to the file you’d like to look at.
2) A new window will open up to Figshare, where the file is stored. You will see a collection of tiles containing the file folder on the top half of the page, and a project description on the bottom half of the page.
3) To view the data in your web browser, click the ‘eye’ icon on your desired file tile.
4) The tabular data will display in your browser. You can expand the screen as needed using the double headed arrow ‘full screen’ icon in the bottom right corner of the table.
5) To download the data, click the ‘download file’ icon on your desired file tile.
6) The files are in CSV format, which is like a simple version of an Excel spreadsheet.
Tip: Each row in the file is a ‘record’ (like a line in a spreadsheet), and each column is a type of information (like date, condition, or measurement).
