Posts

Using AI to improve the quality of endoscopy videos

Cancers detected at an earlier stage have a much higher chance of being treated successfully. The main method for diagnosing cancers of the gastrointestinal tract is endoscopy, when a long flexible tube with a camera at the end is inserted into the body, such as the oesophagus, stomach or colon, to observe any changes in the organ lining. Endoscopic methods such as radiofrequency ablation can also be used to prevent pre-cancerous regions from progressing to cancer if they are detected in time.

Unfortunately, during conventional endoscopy, the more easily treated pre-cancerous conditions and early stage cancers are harder to spot and often missed, especially by less experienced endoscopists. Cancer detection is made even more challenging by artefacts in the endoscopy video such as bubbles, debris, overexposure, light reflection and blurring, which can obscure key features and hinder efforts to automatically analyse endoscopy videos.

In an effort to improve the quality of video endoscopy, a team of researchers from the Institute for Biomedical Engineering (Sharib Ali and Jens Rittscher), the Translational Gastroenterology Unit (Barbara Braden, Adam Bailey and James East) and the Ludwig Institute for Cancer Research (Felix Zhou and Xin Lu) have developed a deep-learning framework for quality assessment of endoscopy videos in near real-time. This framework, published in the journal Medical Image Analysis, is able to reliably identify six different types of artefacts in the video, generate a quality score for each frame and restore mildly corrupted frames. Frame restoration can help in building visually coherent 2D or 3D maps for further analysis. In addition, providing quality scores can help trainees to assess and improve their endoscopy screening performance.

Future work aims to employ real-time computer algorithm-aided analysis of endoscopic images and videos, which will enable earlier identification of potentially cancerous changes automatically during endoscopy.

This work was supported by the NIHR Oxford Biomedical Research Centre, the EPSRC, the Ludwig Institute for Cancer Research and Health Data Research UK.

(1)Real-time detection of artefacts of different types including specularity, saturation, artefact, blur, contrast, bubbles, each indicated with different coloured boxes on the image. Artefact statistics and quality score are generated. Frames suitable for restoration of blur, artefact and saturation are identified. (2) Fast and realistic frames restoration. Discriminator-generator networks are used. (3) Restoration of the entire video. Before restoration, many more frames were corrupted and fewer frames were of good quality compared to after restoration when over 50% of frames had been restored.

Graphical abstract summarising the main messages of the publication. © The Authors CC-BY-NC-ND 4.0

The search for pancreatic cancer biomarkers

Pancreatic cancer has the lowest survival rate of any cancer in the UK, due in part to the limited ability to diagnose it at an early stage. Earlier detection of pancreatic cancer is a major priority of cancer researchers, in order to identify tumours at an earlier stage when they are more easily treatable.

Identifiable biomarkers (naturally occurring molecules which can be related to the presence of a cancer) is one method that can be used to predict or diagnose pancreatic cancer. Currently, the previously-identified biomarkers available have a limited ability to accurately diagnose pancreatic cancer. There is a need to identify new biomarkers that more accurately predict the presence of pancreatic cancer for improved earlier diagnosis.

Dr Christiana Kartsonaki, a senior scientist at the MRC Population Health Research Unit in the Nuffield Department of Population Health, is leading investigations on the potential of protein biomarkers in blood, using data from the China Kadoorie Biobank. Blood samples from over 500,000 Chinese adults have been collected as part of this data set, allowing researchers to identify circulating proteins in the blood and see which individuals went on to develop pancreatic cancer.

During 9 years of follow-up, 700 individuals from the ~500,000 went on to develop pancreatic cancer. From their blood samples, Dr Christiana Kartsonaki and her colleagues will be able to identify a number of protein biomarkers that are associated with a future risk of pancreatic cancer. This study builds on their previous work on the associations of metabolic and lifestyle factors with risk of pancreatic cancer.

Identification of biomarkers may prove very useful in the establishment of strategies to utilise these proteins in predicting the development of pancreatic cancer and help with its diagnosis.

Results from this research will likely be published next year. Once biomarkers are identified, this work may help researchers understand the role that individual proteins play in the development and progression of pancreatic cancer, and whether they may have therapeutic potential as drug targets in its treatment.

About the study

This study is funded by the Nuffield Department of Population Health, Pancreatic Cancer UK and the CRUK Oxford Centre. It was co-led by Associate Professor Michael Holmes, Professor Zhengming Chen, Dr Yuanjie Pang and Dr Christiana Kartsonaki.

Early stage ‘red flag’ symptoms for pancreatic cancer

Pancreatic cancer is the 11th most common cancer in the UK. However, the mortality rate remains the highest among all cancers, due to diagnosis at late stages. As a result, less than 20% of patients diagnosed with pancreatic cancer are suitable for surgery with curable intent, and only 16% of patients are likely to live longer than a year after diagnosis.

The survival rate is much higher when the cancer is found at an earlier stage. However, there is no national screening programme or reliable tests for pancreatic cancer. Most symptoms reported to be associated with pancreatic cancer are vague and non-specific, which increases the difficulty of general practitioners (GPs) recognising early signs of pancreatic cancer in the community.

Identifying red flag symptoms

To address this research gap, the ADEPTS study was set up, using linked data from GP records, hospital records, ONS mortality data, and cancer registry data from the QResearch database, with the aim to better understand the symptom profile of pancreatic ductal adenocarcinoma (PDAC) and pancreatic neuroendocrine neoplasms (PNEN, a rarer type of pancreatic cancer). The ADEPTS study is run by researchers from the Nuffield Department of Primary Care Health Sciences.

This is a case-control study. The team identified about 23600 patients diagnosed with PDAC and 600 patients with PNEN from the QResearch database in the last 20 years.

Up to 10 patients without cancer (controls) with the same age, sex, calendar year registered in the same general practice were identified and matched with each case (patient diagnosed with PDAC/PNEN). The team also identified a list of potential symptoms that may be associated with PDAC and/or PNEN through literature review, leading research charities like Cancer Research UK and Pancreatic Cancer UK, NICE guidelines, and patient representatives. The team explored the presentation of symptoms in different time windows (e.g. within 3 months, 6 months, 1 year, and 2 years before diagnosis) and the association with the diagnosis of PDAC and PNEN.

Through this analysis, a profile of symptoms that are associated with PNEN and PDAC can be determined, which can be used to update the QCancer (Pancreas) prediction model. The model can be used in primary care settings to help GP identify patients who are at high risk and investigate these patients in a timely manner.

So far, the team have already identified a number of red flag symptoms. The results will be published next year. They have also identified certain ethnic groups that are less likely to develop PDAC, along with certain co-morbidities (other health conditions beside pancreatic cancer) that could also be used to predict cancer risk.

Increasing public awareness and GP pathways

After publishing their study findings, the research team hope to engage with relevant stakeholders, to increase public awareness of symptoms that are associated with pancreatic cancer, such as weight loss, abdominal pain, jaundice, etc.

In conjunction with this, the ADEPTS study is working with GPs to improve better direct access to diagnostic investigation resources, such as ultrasound, CT scans and MRIs. This way, when a patient presents to their GP with symptoms, they can be quickly and accurately diagnosed in the hopes of identifying PDAC earlier.

Improved GP assessment tools are being developed as part of the study. By improving the identification and quantification of red flag symptoms associated with pancreatic cancer, the ADEPTS study will help GPs ensure that right patients are sent for the right investigatory methods, making efficient use of scarce or expensive resources such as MRI scans. By communicating its findings with GPs and patients, the ADEPTS study will increase public awareness of symptoms and prompt earlier diagnosis through investigation. Look out for the published findings next year.

About this study

 The ADEPTS study is funded by Pancreatic Cancer UK, and conducted by Weiqi Liao, Ashley Clift, Martina Patone, and Julia Hippisley-Cox from the Nuffield Department of Primary Care Health Sciences.

The QResearch database is founded and directed by Prof Julia Hippisley-Cox, who is the Principal Investigator of the ADEPTS project. External collaborators include Prof Carol Coupland (Medical Statistics) from the University of Nottingham, and Prof Stephen Pereira (Hepatology & Gastroenterology) from University College London.

Oxford Cancer alumni’s biotech success

Scenic Biotech was founded in March 2017 as a spin-out of the University of Oxford and the Netherlands Cancer Institute. The company is based on the Cell-seq technology developed by co-founders Sebastian Nijman and Thijn Brummelkamp in their academic labs.

Cell-seq is a large-scale genetic screening platform that allows the identification of genetic modifiers – or disease suppressors – that act to decrease the severity of a disease. These disease-specific genetic modifiers are difficult to identify by more traditional population genetics approaches, especially in the case of rare genetic diseases. By mapping all the genetic modifiers that can influence the severity of a particular disease, Cell-seq unveils a new class of potential drug targets that can be taken forward for drug development.

In a deal worth $375m, Scenic Biotech has recently entered into a strategic collaboration with Genentech, a member of the Roche Group. This will enable discovery, development and commercialisation of novel therapeutics that target genetic modifiers.

Detecting pancreatic cancer through blood tests

Pancreatic ductal adenocarcinoma (PDAC) makes up 95% of all pancreatic cancer cases and has the lowest survival rate, and early diagnostic methods have yet to be developed. As a result, diagnosis often comes at a later stage when treatment options are limited and prognosis is poor.

Diagnosis at this stage often comes from imaging techniques followed by tissue biopsies, which are not appropriate options to use as standardised, early screening methods. New ways to diagnose PDAC at an earlier stage are needed, without the use of invasive procedures.

Liquid biopsies are becoming a more popular option to fill this demand. Taking a blood sample is minimally invasive, quick, and can tell us a lot of information about a person from their cfDNA (cell free DNA). cfDNA is released from cells and circulates in the blood, containing information about the cell they come from.

Methylation on cfDNA often appears in cancer patients, making it an effective biomarker that can be used to diagnose the presence of cancer with high accuracy and specificity about the cancer (such as location). The concept has many applications, including in the earlier diagnosis of PDAC.

The identification of these biomarkers in blood is often limited to the technology used, with DNA being damaged by the harsh chemicals that are used in the processing. The recent development of TAPS technology at the University of Oxford has helped to overcome this, using a bisulphate-free method, and making it a perfect method for PDAC biomarker identification.

DPhil students Paulina Siejka-Zielinska and Felix Jackson and Postdoctoral Researcher Jingfei Chang from Dr Chunxiao Song’s lab in collaboration with Dr Shivan Sivakumar (consultant medical oncologist) have been investigating TAPS as a method to identify PDAC biomarkers. Using blood samples from PDAC patients and healthy individuals, they are applying TAPS technology to prove that it can be used to accurately detect pancreatic cancer biomarkers in cfDNA.

Preliminary results from this study suggest that cfDNA methylation can be used for the identification of PDAC, as well as being able to accurately distinguish between pancreatic cancer and other pancreatic disorders that effect the DNA, such as pancreatitis.

If this is the case, then the results from this study will make for solid grounds for the application of TAPS in the earlier screening for pancreatic cancer.

About the Song Lab

The Song Lab combine various chemical biology and genome technologies to develop novel tools to analyse the epigenome. The lab apply these tools to two main research areas: the use of epigenetic modifications in circulating cell-free DNA from the blood for non-invasive disease diagnostics including early detection of cancer, and understanding the contribution of epigenetic heterogeneity in cancer development.

Most recently, the TAPS technology developed at the Song Lab has led to the creation of the start up Base Genomics, which has been launched to set a new gold standard in DNA methylation detection using this TAPS technology. Base Genomics will initially focus on developing a blood test for early-stage cancer and minimal residual disease. You can read more about it here.

Funding boost for OxPLoreD early detection study

OxPLoreD is an observational cohort study sponsored by Johnson and Johnson that will recruit 1650 patients from across the UK with pre-cancerous lymphoproliferative disorders. These conditions include monoclonal B-cell lymphocytosis and monoclonal gammopathy of unknown significance that put individuals at higher risk of developing the blood cancers chronic lymphocytic leukaemia and multiple myeloma respectively.

The aim of the study is to look for new ways to find and treat blood cancer sooner by identifying clinical, genomic and immunological predictive markers of progression from these pre-cancerous conditions to malignant disease. The study will also explore the possibility of a future early intervention trial for the subgroup of patients at highest risk of progression.

OxPLoreD is one of the seven clinical trials that have received an £8m funding boost from UK Research and Innovation (UKRI) and will work in partnership with Genomics England. The funding will speed up the adoption of whole genome sequencing in the study of cancer. Genetic analysis is a critical tool that can allow clinicians to select the most appropriate treatments for each patient. In the OxPLoreD study, genetic analysis might be able to identify individuals at highest risk of disease progression that would benefit from earlier treatment. In the longer term this may also enable the identification of those people who would benefit from certain types of treatment.

Alison Cave, UKRI challenge director says:

“Research tells us that one-in-two people in the UK population will get cancer. That stark statistic shows just how important it is for us to seek new treatments. The use of genetic analysis opens new possibilities in our drive to beat cancer. The projects for which we have announced funding today are exciting pointers to future diagnosis and precision treatments”

The funding has been delivered through UKRI’s Industrial Strategy Challenge Fund’s £210m data to early diagnosis and precision medicine (DEDPM) programme. The challenge aims to combine research data and evidence from the NHS to create new and improved ways of identifying disease and treatment pathways.

Prof. Sir Mark Caulfield, Chief Scientist at Genomics England says:

“The 100,000 Genomes Project, Genomics England has analysed the genomes of over 17,000 cancer participants and this suggests that up to half have revealed mutations of potential clinical significance. The DEDPM programme is a major opportunity to expand the application of whole genome sequencing into clinical trials involving cancer where support from the ISCF is likely to deliver significant clinical benefit”

For more information about the other trials funded by this scheme, see the UK Research and Innovation announcement.

Oxfordshire-based SCAN pathway wins BMJ award

Every year, the British Medical Journal (BMJ) runs a competition to find the cancer care team that has developed new approaches to improve cancer diagnosis and treatment. This year, six teams were shortlisted from across the UK and on the 7th October it was announced that the Oxfordshire-based SCAN pathway had won this year’s award.

The Suspected CANcer (SCAN) pathway is designed to accelerate cancer diagnosis in patients with non-specific cancer symptoms. The UK performs worse than many other developed nations in terms of cancer survival and this is in part due to the fact that 21% of cancers are diagnosed after emergency presentation, when they are often at a later stage and more difficult to treat successfully.

In an effort to improve these statistics, urgent referral pathways for suspected cancer have been developed for symptoms specific to one cancer site. However, one in five people diagnosed with cancer only ever report non-specific symptoms of cancer, such as unexplained weight loss, fatigue, nausea, or abdominal pain. These people often experience delays due to being referred sequentially to multiple different tumour site-specific clinics before receiving a diagnosis. The SCAN team identified this unmet need and designed and implemented a new diagnostic pathway that straddles primary and secondary care for patients with non-specific but concerning cancer symptoms.

Patients are referred by their GP to the pathway based in the Churchill Hospital, Oxford, where they are investigated with a whole body computed tomography (CT) scan and undergo blood and stool testing. The outcome of these tests directs the patient to the most appropriate clinical expertise to reach a diagnosis as quickly as possible.

Since its implementation across Oxfordshire in November 2017, the SCAN pathway has seen 2148 patients and diagnosed 201 incidences of cancer, most commonly lung, bowel, pancreas, lymphoma and breast. In addition to cancer diagnoses, the SCAN pathway has diagnosed a large number of serious non-cancer conditions, including tuberculosis, endocrine diseases and inflammatory bowel disease.

“One of the unique features of the SCAN Pathway is that for the remaining patients who do not receive a cancer diagnosis, we offer GPs the option for these patients to have a general medical review in a further attempt to reduce onward referrals.”

  • Julie-Ann Moreland, Macmillan Project Manager and SCAN Navigator, Oxford Radiology Research Unit

Since the SCAN pathway’s inception, the number of GP surgery visits and secondary care referrals prior to receiving a cancer diagnosis decreased by approximately 4-fold, saving a large number of NHS appointments, and the time to diagnosis has reduced. Patients have also responded positively about the service in patient satisfaction questionnaires.

“Prior to the SCAN pathway, patients with non-specific symptoms were having to go to the GP on average 7.8 times and be referred to numerous secondary care clinics before receiving a diagnosis. The SCAN pathway decreases the time to diagnosis and allows patients to start receiving important treatments earlier. This will not only improve patient outcomes but will also reduce the anxiety experienced by patients while waiting for a diagnosis”

 

 “I am delighted that the SCAN team have received this recognition from the BMJ. The judges made a special mention of the holistic care that the clinical team works so hard to provide. Given its success, we are introducing the pathway across the Thames Valley Cancer Alliance and other regions. We are gathering data as we go so we can learn how to improve the service for patients.”

  • Dr Brian Nicholson, Academic GP Lead, Nuffield Department of Primary Care Health Sciences

 

“The development and implementation of the SCAN Pathway has been the result of hard work and collaborative teamwork with passionate people who have strived to develop a service focusing on improving the experience for patients.

“To even be short listed for this award is an incredible achievement and so to win it has been a fantastic and unexpected surprise. We are all very proud of this new pathway and this is a brilliant way to receive recognition and celebrate that.”

  • Zoe Kaveney, Cancer Programme Manager at Oxfordshire Clinical Commissioning Group

 

The SCAN pathway was supported by the Accelerate, Coordinate, Evaluate (ACE) programme funded by NHS England, Cancer Research UK and Macmillan, and the Oxfordshire Clinical Commissioning Group.

Prof. Ellie Barnes comments on the 2020 Nobel Prize for Medicine

Prof Ellie Barnes comments on the recent Nobel Prize in Medicine, awarded to Harvey J. Alter, Michael Houghton and Charles M. Rice for their discovery of the Hepatitis C virus, a major global health problem and a cause of cancer

In 1989, Harvey J. Alter, Michael Houghton and Charles M. Rice used what at the time were state-of-the-art technologies available to identify the virus that causes Hepatitis C infection. This ground-breaking discovery allowed for the development of blood tests to diagnose the Hepatitis C Virus (HCV) and saved millions of lives over the last 40 years.

Testing for HCV has enabled the discovery of chronic infections that results from the Hepatitis C virus. Currently 71 million people are living with HCV, as there is no vaccine to prevent infection. HCV remains a silent disease that is often only diagnosed until symptoms of late-stage liver disease develop. In many cases, it goes undetected until severe complications occur, the most serious of which is hepatocellular carcinoma (HCC). By this point, existing treatments are often less effective at clearing the infection.

Hepatocellular carcinoma is the most common type of primary liver cancer, which is common in those who have had liver scarring due to Hepatitis B and C infections. 400,000 people globally die each year from HCV, with hepatocellular carcinoma continually on the rise. As a result, viral hepatitis is still one of the most serious global pandemics at large. Due to the lack of an effective HCV vaccine and early detection methods for the diagnosis of hepatocellular carcinoma, it is crucial to develop techniques that can aid its early detection and thereby increase the survival rate of cancer patients.

Prof Ellie Barnes at the Nuffield Department of Medicine, leads the DeLIVER study for the early detection of hepatocellular carcinoma that builds on the seminal work as recognised with this year’s Nobel Prize. On the topic of this year’s Nobel Prize winners, she says:

“Now, we need to repeat what those Nobel Prize winners did in 1989 for liver cancer. Like them, we can use today’s new advances in imaging and molecular technology to identify hepatocellular carcinoma at an earlier stage when it is still curable.

“The techniques to do this have advanced remarkably over the last 40 years and it should be possible, with carefully designed patient cohorts and inter-disciplinary effective co-working. By building on the work of Alter, Houghton and Rice, we can do it.”

The risk of liver cancer is increased by viral hepatitis infections, alcohol and obesity, causing the immune system to attack the liver leading to scarring and liver cirrhosis. Monitoring of people with these conditions can reduce mortality but current diagnostic tests for hepatocellular carcinoma fail to detect cancer in many cases.

The DeLIVER team is building on the work of Nobel Prize winners through the use of state-of-the-art multiparametric imaging, viral genetics, and liquid biopsy technologies (such as TAPS) to identify the early indicators of liver cancer by studying people at risk, such as those with Hepatitis C, over several years.

About DeLIVER

DeLIVER is a CRUK-funded programme led by Professor Ellie Barnes that aims to better understand the pre-cancerous changes in the liver and use this knowledge to inform new technologies for early HCC detection. The study will receive patient input from the British Liver Trust and the Hepatitis C Trust.

You can read more about it on the OxCODE website here.

Tackling blood cancers in Tanzania and Uganda

Scientists from Tanzania, Uganda and Oxford University have teamed up in a new child blood cancer program

The relationship between unexpected weight loss & cancer

New research will help GPs to identify the signs they should look for to swiftly diagnose cancer in people with unexpected weight loss