The European Congress of Radiology 2018 saw a record number of delegates and companies gather in Vienna from 28th February to 4th March for one of the biggest and most innovative scientific meetings in Europe hosted by the largest radiological society in the world, the European Society of Radiology (ESR). A total of 28,474 participants attended the Congress where some overarching themes debated during the event included how artificial intelligence (AI) is shaping radiology and patient care, the importance of big data in the field and the challenges of radiomics.
Arguably one of the most important radiological revolutions during recent years is the introduction of endovascular thrombectomy (ET) for the management of acute ischaemic stroke. Solid evidence in the form of randomised control trials (such as the 2015 MRCLEAN, ESCAPE, SWIFT PRIME, EXTEND-IA and REVASCAT trials) robustly support the clinical benefits of the procedure. Excitingly, there has also been promising recent research (the 2018 DAWN trial) showing successful endovascular treatment in a wider pool of patients, in particular concerning the therapeutic window. In addition to the obvious necessity for accurate diagnostic neuroradiology to select appropriate ET candidates, the technicalities of the intervention itself are constantly evolving. For example, Dr Pedro Viela (Head of Neuroradiology at Beatriz Angelo Hospital, Lisbon) discussed how dealing with distal occlusions and ET treatment anaesthetic protocol are hot topics at present in this field.
What are your thoughts on this lady’s case: do you think she is a candidate for mechanical thrombectomy?
Another interesting theme at ECR 2018 was that of immunotherapy in cancer care, presented by Prof. Vicky Goh. Engaging the immune system by promoting sustained tumour specific T-cell killing via increasing production of cytokines can neutralise suppressor mechanisms when antibodies block inhibitory signals. This acts to avoid immune evasion that is a hallmark of cancer and a key mechanism in its progression cycle. Trials have shown that this type of treatment can prolong survival with synergistic effect1, however it does have significant side effects such as for the skin and gastrointestinal system. Radiologically, there is also the complex challenge of how to actually assess the efficacy of the treatment and also how to identify the patients most likely to benefit from it. This is crucial in order to justify the likelihood of adverse side effects resulting from the treatment itself.
Do you think use of pembrolizumab, an immunotherapy, is appropriate in this case? What other treatment could you recommend?
At a time when the world has its eye on ‘big data’, it is important to consider where this fits into the spectrum of medical care and ultimately how we can harness advancement in this area to help improve and save patients’ lives. The concept of ‘value-based radiology’ grows ever-more crucial in three key areas as outlined by Prof. Marc Dewey at ECR 2018. Firstly, radiologists must be vigilant about what specific imaging is actually necessary and in which patients: a process which can be aided by ‘decision support modules’ such as ESR’s iGUIDE or eGUIDE. Next, human error reduction by integration of artificial intelligence in pattern recognition within radiology will be invaluable to advance the specialty and increase accuracy under the watch of ‘the bionic radiologist’. This role will be supervision of machine-generated results produced by dynamic algorithms and can be compared to a pilot overseeing autoflight mode and seamlessly taking over when human action is required. Prof. Dewey’s department also found that patients prefer this approach with 84% of patients feeling more comfortable with analysis by radiologist plus computer instead of either alone. Lastly, structuring of radiology reports to allow choice from pre-determined descriptions of particular abnormalities in different imaging types as opposed to free text ought to improve objectivity and reduce variability of results reporting. This would hopefully lead to better marriage between the results themselves and optimal treatment planning.
Would a machine be able to diagnose this patient based on the imaging provided? What pitfalls do you see in this type of radiology?
ECR 2019 will again be in Vienna and will be held on 27th February to 3rd March next year.
Developed by doctors, MedShr is the easiest and safest way for medical professionals to discover, discuss and share clinical cases and medical images. Now over 450,000 members worldwide!
- Ref: Hodi FS et al. N Engl J Med 2010; 363, 711-723