Learning MRI and histology image mappings for cancer diagnosis and prognosis
Lead Research Organisation:
University College London
Department Name: Computer Science
Abstract
This project aims to exploit recent advances in machine learning to address acute problems in cancer management - most directly prostate cancer. The current standard approach of making treatment decisions via biopsy and histology has two key limitations; it is invasive and subjective/inconsistent. We will develop the computational tools supporting new solutions that resolve both issues. Specifically, we aim to enable non-invasive MRI to become the primary diagnostic tool. This would avoid a large number of unnecessary biopsies, which carry significant risk of life-changing side-effects, reserving the procedure for only marginal cases. We also plan to relate MR signals to quantitative tissue features enabling consistent assessment and thus more reliable treatment decisions.
The use of MRI in prostate cancer has become routine only in the last few years. Thus, data relating MRI to patient outcome (e.g. 5-10 year survival) is not available. However, we are uniquely positioned to obtain i) associated MRI and histology images, and ii) associated histology and patient outcome. In combination, these support a two-step learning and estimation process: from MRI to histological features; and from histological features to patient prognosis. Such mappings can provide invaluable new information for clinical decision making, as well as guide the design of maximally informative future MRI protocols. Such protocols will enable long-term data collection initiatives that support direct mappings from MRI to outcome.
The project involves engineering challenges that demand innovations at the cutting edge of image-based machine learning technology: i) accommodating uncertainty in the alignment of training images; ii) quantification and visualization of uncertainty in the output of learned models; iii) salient feature selection in high-dimensional input data; iv) development of experiment design optimization algorithms driven by implicit computational models (such as neural networks). We build on the latest ideas in deep learning to address these challenges. We tailor solutions relevant to the immediate problems at hand in prostate cancer, but that extend to related tasks in cancer imaging and medical imaging in general.
The use of MRI in prostate cancer has become routine only in the last few years. Thus, data relating MRI to patient outcome (e.g. 5-10 year survival) is not available. However, we are uniquely positioned to obtain i) associated MRI and histology images, and ii) associated histology and patient outcome. In combination, these support a two-step learning and estimation process: from MRI to histological features; and from histological features to patient prognosis. Such mappings can provide invaluable new information for clinical decision making, as well as guide the design of maximally informative future MRI protocols. Such protocols will enable long-term data collection initiatives that support direct mappings from MRI to outcome.
The project involves engineering challenges that demand innovations at the cutting edge of image-based machine learning technology: i) accommodating uncertainty in the alignment of training images; ii) quantification and visualization of uncertainty in the output of learned models; iii) salient feature selection in high-dimensional input data; iv) development of experiment design optimization algorithms driven by implicit computational models (such as neural networks). We build on the latest ideas in deep learning to address these challenges. We tailor solutions relevant to the immediate problems at hand in prostate cancer, but that extend to related tasks in cancer imaging and medical imaging in general.
Planned Impact
Prostate cancer is the most common type of cancer in males in many countries worldwide, including the UK. Currently more than 8 in 10 men in England and Wales survive prostate cancer for ten years or more since the time of diagnosis. Early and accurate diagnosis is key for improved prognosis. It is important to identify early both high-risk cases of invasive cancer that require immediate treatment and additionally low-risk cases that grow slowly and might never develop into a life-threatening disease. The latter becomes increasingly important as a large number of patients are subjected to unnecessary biopsies and treatments (over-treatment) and their associated side-effects.
Management and treatment of cancer is one of the biggest challenges in medicine worldwide, costing the UK economy more than £15bn a year. Prostate cancer is the most common type of cancer in males with around 47,300 cases diagnosed annually. This project advances towards the use of MRI for early and accurate patient stratification of prostate cancer patients with significant socio-economic impact.
From an economic point of view, early diagnosis minimises the negative impact of cancer on the active workforce, as it is associated with increased survival rates. In addition, addressing the problem of over-treatment, by identifying the patients with low-risk disease that will not progress to be life-threatening, has the advantage of avoiding costly treatments that are unnecessary.
From a societal point of view, the benefits of early patient stratification are evident both for the patients and their families. This project makes an important advance in early and accurate diagnosis based on patient-specific in-vivo imaging, enabling subsequent personalised treatment. This involves immediate cancer therapy options for high-risk cases, while potentially sparing severe discomfort and side-effects from unnecessary biopsy and loss of quality of life from unnecessary chemotherapy treatments and surgical interventions for low-risk prostate cancer cases. In both cases the patient will benefit from improved quality of life.
The project uses prostate cancer as its demonstrator, but the framework has wider applications with similar impact both in a range of other solid cancers (for example breast, brain, pancreas, liver etc) as well as other medical domains that currently rely on histology, such as neuroscience.
Management and treatment of cancer is one of the biggest challenges in medicine worldwide, costing the UK economy more than £15bn a year. Prostate cancer is the most common type of cancer in males with around 47,300 cases diagnosed annually. This project advances towards the use of MRI for early and accurate patient stratification of prostate cancer patients with significant socio-economic impact.
From an economic point of view, early diagnosis minimises the negative impact of cancer on the active workforce, as it is associated with increased survival rates. In addition, addressing the problem of over-treatment, by identifying the patients with low-risk disease that will not progress to be life-threatening, has the advantage of avoiding costly treatments that are unnecessary.
From a societal point of view, the benefits of early patient stratification are evident both for the patients and their families. This project makes an important advance in early and accurate diagnosis based on patient-specific in-vivo imaging, enabling subsequent personalised treatment. This involves immediate cancer therapy options for high-risk cases, while potentially sparing severe discomfort and side-effects from unnecessary biopsy and loss of quality of life from unnecessary chemotherapy treatments and surgical interventions for low-risk prostate cancer cases. In both cases the patient will benefit from improved quality of life.
The project uses prostate cancer as its demonstrator, but the framework has wider applications with similar impact both in a range of other solid cancers (for example breast, brain, pancreas, liver etc) as well as other medical domains that currently rely on histology, such as neuroscience.
Organisations
Publications
Anazodo UC
(2023)
A framework for advancing sustainable magnetic resonance imaging access in Africa.
in NMR in biomedicine
Atzeni A
(2022)
Deep active learning for suggestive segmentation of biomedical image stacks via optimisation of Dice scores and traced boundary length.
in Medical image analysis
Bailey C
(2019)
VERDICT MRI validation in fresh and fixed prostate specimens using patient-specific moulds for histological and MR alignment.
in NMR in biomedicine
Battiston M
(2019)
Fast bound pool fraction mapping via steady-state magnetization transfer saturation using single-shot EPI.
in Magnetic resonance in medicine
Bocchetta M
(2019)
Segmentation of medial temporal subregions reveals early right-sided involvement in semantic variant PPA.
in Alzheimer's research & therapy
Bocchetta M
(2018)
Hippocampal Subfield Volumetry: Differential Pattern of Atrophy in Different Forms of Genetic Frontotemporal Dementia.
in Journal of Alzheimer's disease : JAD
Boonsuth R
(2023)
Feasibility of in vivo multi-parametric quantitative magnetic resonance imaging of the healthy sciatic nerve with a unified signal readout protocol.
in Scientific reports
Boonsuth R
(2021)
Assessing Lumbar Plexus and Sciatic Nerve Damage in Relapsing-Remitting Multiple Sclerosis Using Magnetisation Transfer Ratio.
in Frontiers in neurology
Cohen-Adad J
(2022)
Comparison of multicenter MRI protocols for visualizing the spinal cord gray matter.
in Magnetic resonance in medicine
Collorone S
(2020)
Reduced neurite density in the brain and cervical spinal cord in relapsing-remitting multiple sclerosis: A NODDI study.
in Multiple sclerosis (Houndmills, Basingstoke, England)
Devine W
(2019)
Simplified Luminal Water Imaging for the Detection of Prostate Cancer From Multiecho T2 MR Images.
in Journal of magnetic resonance imaging : JMRI
Grussu F
(2021)
Feasibility of Data-Driven, Model-Free Quantitative MRI Protocol Design: Application to Brain and Prostate Diffusion-Relaxation Imaging
in Frontiers in Physics
Grussu F
(2020)
Multi-parametric quantitative in vivo spinal cord MRI with unified signal readout and image denoising.
in NeuroImage
Iglesias JE
(2023)
SynthSR: A public AI tool to turn heterogeneous clinical brain scans into high-resolution T1-weighted images for 3D morphometry.
in Science advances
Jin C.
(2022)
LEARNING TO DOWNSAMPLE FOR SEGMENTATION OF ULTRA-HIGH RESOLUTION IMAGES
in ICLR 2022 - 10th International Conference on Learning Representations
Lin H
(2023)
Low-field magnetic resonance image enhancement via stochastic image quality transfer
in Medical Image Analysis
Ning L
(2020)
Cross-scanner and cross-protocol multi-shell diffusion MRI data harmonization: Algorithms and results.
in NeuroImage
Peter L
(2021)
Uncertainty-Aware Annotation Protocol to Evaluate Deformable Registration Algorithms.
in IEEE transactions on medical imaging
Pizzolato M
(2020)
Computational Diffusion MRI - MICCAI Workshop, Shenzhen, China, October 2019
Prados F
(2020)
Generalised boundary shift integral for longitudinal assessment of spinal cord atrophy.
in NeuroImage
Sen S
(2022)
Differentiating False Positive Lesions from Clinically Significant Cancer and Normal Prostate Tissue Using VERDICT MRI and Other Diffusion Models.
in Diagnostics (Basel, Switzerland)
Sen S
(2024)
ssVERDICT: Self-supervised VERDICT-MRI for enhanced prostate tumor characterization.
in Magnetic resonance in medicine
Singh S
(2022)
Avoiding Unnecessary Biopsy after Multiparametric Prostate MRI with VERDICT Analysis: The INNOVATE Study.
in Radiology
Slator PJ
(2019)
Combined diffusion-relaxometry MRI to identify dysfunction in the human placenta.
in Magnetic resonance in medicine
Thomas-Black G
(2023)
Multimodal Analysis of the Visual Pathways in Friedreich's Ataxia Reveals Novel Biomarkers.
in Movement disorders : official journal of the Movement Disorder Society
Tregidgo HFJ
(2023)
Accurate Bayesian segmentation of thalamic nuclei using diffusion MRI and an improved histological atlas.
in NeuroImage
Tur C
(2022)
Spatial patterns of brain lesions assessed through covariance estimations of lesional voxels in multiple Sclerosis: The SPACE-MS technique.
in NeuroImage. Clinical
Yang H
(2022)
Biomarkers and Disease Trajectories Influencing Women's Health: Results from the UK Biobank Cohort.
in Phenomics (Cham, Switzerland)
Zhang L
(2023)
Learning from multiple annotators for medical image segmentation.
in Pattern recognition
Zhang L.
(2020)
Disentangling human error from the ground truth in segmentation of medical images
in Advances in Neural Information Processing Systems
Description | This award has helped the ongoing translation of microstructural MRI techniques, in particular VERDICT-MRI but also luminal water imaging, for prostate cancer imaging. VERDICT has been through a series of clinical trials. The most recent (Singh Radiology 2022) confirms that the technique can distinguish clinically significant cancer from benign non-invasively with sufficient accuracy to change clinical practice in particular by substantially reducing the number of patients that require invasive prostate biopsy. |
Exploitation Route | VERDICT MRI is now moving into clinical practice. |
Sectors | Healthcare |
URL | https://news.sky.com/story/prostate-scan-could-save-thousands-of-unnecessary-biopsies-charity-says-12664827 |
Description | See key findings: used to help propel VERDICT into clinical usage and potentially in national screening programmes for prostate cancer, which are currently being designed. |
First Year Of Impact | 2019 |
Sector | Healthcare |
Impact Types | Societal |
Description | Assessing Placental Structure and Function by Unified Fluid Mechanical Modelling and in-vivo MRI |
Amount | £1,124,021 (GBP) |
Funding ID | EP/V034537/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start | 01/2022 |
End | 07/2024 |
Description | EPSRC Centre for Doctoral Training in Intelligent, Integrated Imaging In Healthcare (i4health) |
Amount | £6,034,274 (GBP) |
Funding ID | EP/S021930/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start | 08/2019 |
End | 03/2028 |
Description | JPND: Stratification of presymptomatic amyotrophic lateral sclerosis: the development of novel imaging biomarkers |
Amount | € 1,600,000 (EUR) |
Funding ID | MR/T046473/1 |
Organisation | JPND Research |
Sector | Academic/University |
Country | Global |
Start | 06/2020 |
End | 07/2023 |
Title | 'Select and Retrieve via Direct Upsampling' Network (SARDU-Net) |
Description | "Select and retrieve via direct up-sampling" network (SARDU-Net) is a new method for model-free, data-driven quantitative MRI (qMRI) experiment design. SARDU-Net identifies informative measurements within lengthy acquisitions and reconstructs fully-sampled MRI signals from a sub-protocol, without prior information on the MRI contrast. It combines two deep networks in an encoder-decoder fashion. Firstly, a selector selects a signal sub-sample, which is passed to a predictor, which retrieves fully sampled input signals. SARDU-Net can be run with standard computational resources and can increase the clinical appeal of qMRI. |
Type Of Material | Computer model/algorithm |
Year Produced | 2019 |
Provided To Others? | No |
Impact | SARDU-Net has made important impacts: - it was used to submit an entry to the 2019 MICCAI Challenge on Computational Diffusion MRI known as 'MUDI' (MUlti-dimensional Diffusion MRI), which won the 1st prize in October 2019 - it was used for a submission for the 2020 annual meeting of the International Society for Magnetic Resonance in Medicine (ISMRM), which was accepted as an oral power pitch presentation and awarded the 2nd prize for Mansfield Innovation Award by the British & Irish chapter of the ISMRM; - it is being currently used as part of the same Researchfish Award to design prospective MRI acquisitions. |
Description | Bloomsbury festival |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | Regional |
Primary Audience | Schools |
Results and Impact | My PhD student and myself along with other colleagues and Phd students from CMIC and WEISS presented "all things imaging and what it is for" to school kids and the general public at the Bloomsbury Festival this year. Our stall at the Discovery Hub (Senate House, Malet Street) illustrated our various imaging technics (and what we use them for), as well as addressing those fundamental questions around how do we know where the lungs sits, what an MRI image actually looks like, is AI faster than school kids. Also, of course, an opportunity to touch on more serious questions around the applications in medical context such as cancer imaging. Bloomsbury Festival theme this year was "Breathe" and our exhibit - "From Top to Toe - the breathing body" . Thursday and Friday were dedicated to schools, who had booked tours and brought children aged 8-12 years. We had approximately 120 students visiting us. We had excellent feedback from the enthusiastic students. During the weekend we had general audience and approximately 200 people visited during the hours of the exhibition. All children were happy to engage with the activities while we had more in depth discussions about our methods and how they aid diagnosis with parents and grandparents, but also some independent patients. |
Year(s) Of Engagement Activity | 2022 |
URL | https://bloomsburyfestival.org.uk/ |