Vol 32 ISSUE 03-04/23 October 2023 THE EUROPEAN FORUM FOR THOSE IN THE BUSINESS OF MAKING HEALTHCARE WORK Not always fair AI in healthcare: diverse and balanced training datasets of high quality needed Machine learning and artificial intelligence (AI) are playing an increasingly important role in medicine and healthcare, and not just since ChatGPT. This is es- pecially true in data-intensive specialties such as radiology, pa- thology or intensive care. The quality of diagnostics and deci- sion-making via AI, however, does not only depend on a soph- isticated algorithm but – cru- cially – on the quality of the training data. tem to be able to make a decision, many thousands or millions of x-rays of benign and malignant tis- sue changes are needed as training data. The algorithm classifies each of these training images and com- pares the result with the human- made diagnosis. If the algorithm’s diagnosis was the weights assigned to individual con- nections between the virtual neur- ons are reweighted to improve ac- curacy next time. Once all training data has been processed, a new, incorrect, dated is rare but the algorithm does require a huge amount of training data. When a high volume of digital data can be obtained, it is often from sources of inconsistent quality. This quantity and quality problem is further complicated by privacy issues in healthcare. Garbage in, garbage out Most IT users know the phrase “garbage in, garbage out”: if the input data is garbage, the output will be garbage as well. This also vious: In the US, an AI algorithm was used to estimate which in- patients would need additional care. The training data used the costs patients had previously in- curred as a marker for disease se- verity? This led to the fact that for African-American ad- ditional care was less likely to be recommended because these pa- tients had in the past incurred lower costs. This was not due to their lower disease severity but to their lower access to health care, patients m o c . e b o d a . k c o t s - s k r o w r o m a t e m © smaller data set is used to check the accuracy of the fully trained al- gorithm. This step is called vali- dation. Quality of the training data No doubt: The accuracy of the AI algorithm can only be as good as the quality of the training data. If, for example, the training data con- tains many x-rays, in which a ma- lignant tissue change was mis- takenly considered benign by the human expert or vice versa, the AI will learn from false examples – which will affect its accuracy later on. Slowly digitalized healthcare system training Obtaining high-quality data for algorithms, however, often turns out to be difficult since our healthcare system is only slowly being digitalized. Data that has been carefully and manually vali- holds true in AT where the ac- curacy of a classification algorithm always depends on the quality of the training data. But not all types of low-quality output are immedi- ately recognizable as such. There can be subtle biases in the classifi- cations of an algorithm caused by an unbalanced composition of the training data. The accuracy of diag- nostic algorithms, for example, is worse in populations whose data was underrepresented in the orig- inal training data. from predominantly Well-known examples are al- gorithms to classify malignant skin tumors which were trained with data fair- skinned (Caucasian) individuals. These algorithms show lower diag- nostic accuracy when they are sup- posed to make a correct diagnosis in a dark-skinned person? The source of the bias is not always ob- i.e. a pre-existing systemic dis- advantage. The unbalanced composition of training data is often described with the acronym WEIRD: “white, educated, industrialized, rich and democratic countries” are over- represented. Women and the elderly are also disadvantaged And it’s not just the ethnic and economic background that creates bias. Women are not fairly repre- sented in AI either. For example, researchers suspect that women are overrepresented in the diag- nosis of depression because, among other things, diagnostic al- gorithms query behaviors that are more common in women – regard- less of clinical depression?. In contrast, the Institute of Health In- formatics at University College is intelligence Artificial a buzzword. Underneath the buzz, AI consists of algorithms based on certain machine learning methods. One of these methods, which has received a lot of attention in recent years, is the artificial neural net- work. The layers of nerve cells that are involved in learning processes in the human brain are algorith- mically reproduced (albeit idea- lized). Highly complex learning tasks require many layers of ar- tificial neurons – this is deep learn- ing, another AI term that has be- come popular. Learning from training data A neural network and other forms of AI learn how to perform their task – such as making a specific di- agnosis – based on training data. Imagine, the task is to distinguish malignant from benign findings on chest x-rays. In order for the sys- www.healthcare-in-europe.com London showed that an AI al- gorithm for diagnosing liver dis- ease in women had a significantly lower hit rate: it was wrong in 44% of women, but only in 23% of men. Age as a factor for the development of bias And another factor that can play a role in the development of bias: age. Facial recognition algorithms, for example, are less accurate in an elderly population?. This is particu- larly disconcerting in view of the fact that robotics is increasingly being used in geriatric care, for example to inform and entertain elderly people and dementia pa- tients. Particularly in this field there is substantial research being conducted to improve machine rec- ognition of emotions based on fa- cial expressions. In order to address these biases and ensure equal treatment in a digitalized healthcare system di- verse, balanced and high-quality training data sets are imperative. This also requires legal certainty with regard to the use of patient data for research and development. This issue will be addressed in Germany by the Federal Ministry of Health’s? proposed laws in the context of the digitalization strat- egy and in the EU with the Euro- pean Health Data Space? ■ Report: Dr. Christina Czeschik CONTENTS AI RADIOLOGY WOMEN‘S HEALTH GENDER MEDICINE LABORATORY/ PATHOLOGY 1-3 4-7 8-9 10-11 12-16 www.healthcare-in-europe.com

ARTIFICIAL INTELLIGENCE 3 Impact of technology AI in radiology: helper or bane of society and the environment? The climate crisis and AI – ar- guably two of the most hotly- debated and relevant topics of our time – share an intricate rela- tionship: While computation of complex AI routines commands an immense carbon footprint, it is these algorithms that might be the very key to mitigate the ef- fects of global warming. In a dedicated session at ECR 2023, radiologists explored the socie- tal and environmental impact of the technology in healthcare. Angel Alberich-Bayarri, CEO of Val- encia-based AI company Quibim, looked at the technology’s carbon footprint and the strategies to make AI sustainable in the opening talk. ‘AI is the new electricity, and it will continue to change the world,’ he told the audience. ‘The power for positive change that AI brings holds the possibility for negative impacts on society.’ This impact is felt, among others, in the massive amount of CO2 that is being generated when devel- oping a new algorithm. ‘Training a single deep learning, natural lan- guage processing algorithm can approximately take 600,000 pounds of carbon emissions,’ the expert explained. ‘It’s the same amount of CO2 produced by five cars over lifetimes.’ For example, training Google AlphaGo Zero, a tool developed to solve the Go game, generated 96 tons of CO2 over 40 days – the same amount as 1,000 hours of air travel. their Sustainability must come into the li- melight To date, sustainability has barely been explored in AI ethics, a field that studies ethical and societal is- sues facing developers, producers, consumers, citizens, policy makers and civil society organizations. Al- berich-Bayarri expressed his con- viction that times are calling for a change and sustainability must now come into the limelight. ‘Sus- tainable AI is in its infancy: it’s underdeveloped, under researched and underfunded, but we’re start- ing to explore it. It’s time to ana- lyse the environmental impact of AI. We can’t neglect the carbon footprint that we are creating when training a model.’ Efforts to direct AI uses towards good purposes are increasing, but there are no real business tools available to measure AI’s carbon footprint yet. ‘We’re starting to get concerned about the impact of AI on our environment,’ the Quibim CEO explained the in- herent contradiction. While AI is being proposed to power up so- ciety, its development and use make society unsustainable at the same time. Deep learning models are posing both financial and envi- ronmental costs – from hardware, electricity and cloud computing time to the carbon footprint gener- ated when fuelling modern tensor processing hardware. ‘The energy required to do so is considerable,’ he said. Tracking, shrinking, specializing: Stra- tegies to reduce AI’s carbon footprint There are already a few tools that AI developers can use to calculate their carbon emissions – for example the Machine Learning Emissions Calculator, to estimate the carbon footprint of GPU com- puting, and the Carbontracker, an experiment impact tool for real time assessment of energy con- sumption and carbon emissions. ‘This carbon tracking is being in- in grant ap- creasingly asked www.healthcare-in-europe.com m o c . e b o d a . k c o t s - o r p r a m r a t S © plications or proposals in hospitals to measure impact on the environ- ment. I encourage all of us to track our CO2 emissions.’ There are three main aspects to consider to foster sustainable AI. The first consists in designing smaller models when im- plementing AI. Several research ini- tiatives are exploring how to train models faster and more efficiently, relying on techniques like pruning, compression, distillation and quan- tization. ‘The goal is to shrink down the size of the models and use fewer computing cycles to de- crease financial and environmental costs of building and deploying AI,’ Alberich-Bayarri said. and The second aspect is to alternate deployment strategies, including the use of specialized hardware like application-specific integrated circuits (ASICs); prevention of idle consumption sub-optimal computational processing dis- tribution; and optimizing the use of existing hardware like general- purpose CPUs. Last but not least, the AI community must work to improve carbon awareness, the ex- pert appealed. ‘We have to partner with cloud providers who are aware about the issue. Some data processing centres are using im- mersion-based cooling for different devices, grid interactive UPS bat- teries or hydrogen batteries instead of diesel generators as a backup for main data processing centres.’ Not in a rush? Keep your data cool AI developers must also consider where they store the files depend- ing on how quickly they need to be retrieved, he urged. ‘Amazon, for instance, offers different port- folios: we can have a very warm storage with immediate access to data, but if we don’t need to use the data right away, we can store it in an iceberg instead and have the data in hours instead of seconds. That’s another way to reduce our environmental impact.’ Ana Jimenez-Pastor, Vice President of AI at Quibim, then focused on how AI can help mitigate the lack of radiologists in low- to middle-in- come countries. In these regions, she argued, the global shortage of radiologists is felt more acutely. For example, in Tanzania, roughly 60 radiologists have to tend to a 25 million population, or Kenya, where here are only three fellow- ship-trained breast radiologists to scan over four million women who need annual mammography. ‘AI could have a great impact in these countries with few to no radi- ologists available, by automating image interpretation and poten- tially enabling a few radiologists to manage of hundreds,’ she suggested. ■ the workload Report: Mélisande Rouger Ángel Alberich- Bayarri Ángel Alberich-Bayarri is a Telecom- munications Engineer with special- isation in electronics by the Technical University of Valencia and PhD in Bio- medical Engineering for his research on the application of advanced image pro- cessing techniques to magnetic reson- ance imaging. Founder and CEO of Quibim (Quanti- tative Imaging Biomarkers in Medi- cine), company dedicated to the advanced analysis of medical images using artificial intelligence. He is the author of more than 90 scientific ar- ticles in prestigious international journ- als and inventor of 5+ patents. He is also the author of more than 100 com- munications to international con- gresses, editor of international books and author of 20+ book chapters. He has participated in a high number of international research projects and clinical trials. He is an active member of several scientific societies, among which stands out his participation as a member of the Board of Directors of the European Society of Medical Im- aging Informatics (EUSOMII). Ana Jimenez- Pastor Ana Jimenez-Pastor is a Telecom- munications and Biomedical Engineer from the Polytechnic University of Val- encia (Spain). She currently works as VP of AI at Quibim, leading the coordi- nation and development of new and innovative AI solutions applied to the field of radiomics together with MLOPs pipelines to ensure data governance and AI models reproducibility. Ana has dedicated the last seven years to the research, development and integration in clinical practice of solutions that combine image analysis, quantification and AI. Ana has participated and coordinated different European H2020 projects fo- cused on the development of European infrastructures for the study of cancer using AI and image analysis. Author of more than ten publications and book chapters, she has presented her re- search in different national and inter- national congresses.

RADIOLOGY Prostate cancer treatment MR-guided radiotherapy: a game changer? 5 Advertisement Comfortable and secure: The swivelling handle system helps patients to position themselves on the examination table with as little assistance as possible – even when transferring from a wheelchair. Febromed provides support for everyday radiology work Safe and hygienic: ‘Get up®’ physical complaints, there are also costs for the em- ployer and the social system. ‘Get up®’ by Febromed helps reduce this strain in day to- day work to a minimum. tomography MRg-A-SBRT can significantly enhance the safety of radiotherapy for prostate cancer patients. © Netherlands Cancer Institute (NKI) radiotherapy is magnetic Prostate tech- niques have been transformed over the past two decades. One promisting technique in this con- text resonance- guided radiotherapy. The latest clinical results show a dramatic reduction in side effects, improv- ing patient outcomes and quality of life. Prostate cancer is the fourth most common cancer worldwide. Ac- cording to the International Agen- cy for Research on Cancer (IARC), there were over 1.4 million new cases worldwide in 2020, and by 2040 this is set to rise to 29.5 mil- lion new cases and 16.3 million deaths annually. Radiotherapy re- mains a fundamental component of effective treatment, with 50% of all cancer patients receiving it as part of their care. In external beam radiotherapy, radiation is delivered to the prostate five days a week from a machine outside the body. Prostate radiation therapy: 90% suc- cess but short-term side effects ‘This standard treatment option for prostate cancer can lead to short- and long-term side effects, how- ever, including urinary problems, bowel problems, fatigue, erectile dysfunction and damage to sur- rounding tissue,’ said US radiation oncologist Dr Jonathan E. Leeman. Oncologists and radiotherapist teams use several main strategies to minimize the risk of these side effects, including careful treatment planning and close patient moni- toring during and after radiation therapy. In a risk-adapted strategy, con- temporary radiotherapy treatment algorithms use pretreatment prog- nostic factors to stratify patients into low-, intermediate-, and high- risk groups. Then, precision radi- ation delivery techniques, such as intensity-modulated radiation ther- apy and image-guided radiation therapy, are used to help avoid ex- posure to normal tissues. Magnetic resonance-guided radio- therapy (MRgRT) uses offline MR www.healthcare-in-europe.com imaging to help plan radiation vol- umes in order to ensure accurate tumour targeting while sparing critical normal tissues. MRg- A-SBRT is the product of advances in MR-guided radiotherapy and stereotactic body radiation therapy (SBRT) and emerged at the begin- ning of the decade from a collabor- ation between US and European re- searchers and clinicians. These include the Department of Radiation Oncology at Stanford University, Memorial Sloan Ketter- ing Cancer Center (MSKCC), New- York the Belgian University Medi- cal Center Utrecht (UMC Utrecht) and the German Research Center for Oncology (DKFZ) in Heidel- berg. SBRT uses MRI ‘to deliver high doses of radiation to precise targets in the body’, said Leeman. Real-time monitoring enables treat- ment to be adapted on a daily basis, tailoring radiation delivery to prostate changes and thus reduc- ing side effects. Prospective study analyses data from 29 clinical trials This technology has already been used successfully to treat breast, prostate, pancreatic, liver, lung and limited metastatic cancers, in ad- dition to non-cancer indications such as cardiac ablation. MRg- A-SBRT enables clinicians to ac- curately target the prostate while sparing bladder, urethra, and rectal tissue. However, its impact on clini- cal outcomes and side effects com- pared to standard computed to- mography-guided SBRT (CT-SBRT) was unclear. A team of researchers from the Dana-Farber Cancer Institute has just published a large study of MR- guided daily adaptive SBRT in Cancer, which directly assesses its benefits compared to standard techniques for the first time. 60% reduction in short-term bowel side effects The study combined data from 29 clinical trials including a total of 2,547 patients to compare the side effects of MRg-A-SBRT to more conventional treatment methods with CT guidance but without daily adjustments. ‘We found that the risk of short-term urinary side ef- fects was reduced by 44% [a 1.79-fold reduction] and the risk of short-term bowel side effects was reduced by 60% [a 2.5-fold reduc- tion]’, said Leeman. While these results strongly sup- port the use of MRg-A-SBRT as an effective treatment option in pros- tate cancer, Leeman noted that longer follow-up is required to see whether the short-term benefits will lead to more impactful long- term benefits for benefits. ■ Report: Bernard Banga Jonathan E. Leeman Dr Jonathan E. Leeman is a radiation oncologist at Dana-Farber Cancer Insti- tute and Brigham and Women‘s Hospi- tal, Harvard Medical School, USA. He specializes in the treatment of prostate cancer and MRI-guided radiation treat- ment for multiple cancer types. Dr Lee- man‘s clinical expertise lies in the areas of MRI-guided radiotherapy, adaptive radiotherapy, and prostate cancer. He serves as the physician lead of the MRI-guided radiation program at Dana-Farber Cancer Institute. His re- search focuses on the treatment of prostate cancer and the development of MRI-guided radiation therapy. He is also an instructor and has specialties in cancer research, radiation oncology, prostate cancer, and head and neck cancer. Whether in magnetic reso- nance imaging (MRI), com- puted (CT), diagnostic radiology, or ra- diation therapy: in radiolo- gy, it’s all about the details. Highly specialised equip- ment in a professional en- vironment helps to make exact diagnoses and im- plement precise therapies. Everything else needs to be just as professional. We at Febromed can help with daily operations with the ‘Get up®’ handle system. With our handle system, pa- tients can perfectly position themselves on the examina- tion table at their own pace. This takes the pressure off of medical personnel, letting them concentrate on what’s essential: the examination. Reducing workload In radiology, a large section of patients have restricted mobility, something that can often present a considerable strain for medical staff. They have to use their full physical strength to move patients – and are in danger of becom- ing a medical emergency themselves. But besides the Safety isn’t just a handling issue; everything has also been carefully considered in terms of hygiene. The ‘Get up®’ handle system by Febromed is easy to disin- fect and meets the highest hygienic requirements of a medical environment. The material is extremely dura- ble, an investment in safety, but also in cost efficiency. The ‘Get up®’ handle system can not only be mounted on the ceiling, but also on the wall. Photos: Febromed www.febromed.de Febromed GmbH & Co.KG Am Landhagen 52 | D-59302 Oelde Telephone Telefax E-Mail 0049 2522 92019 20 0049 2522 92019 19 vertrieb@febromed.de

RADIOLOGY Plenary discussion Sustainable radiology: why it takes more than “greener” imaging systems 7 Despite the strong Irish pres- ence at this year’s ECR, it’s clear that radiology is lacking in the “green” department: healthcare still causes a large share of glo- bal greenhouse gas (GHG) emissions, not least due to diag- nostic imaging. But while it might be easy to pin the blame on power-hungry CT and MRI systems, things might not be quite as clear-cut after all: Dr Sarah Sheard, Consultant radi- ologist at Imperial College Healthcare, UK, invited her ECR audience to take a closer look at radiology’s climate footprint. While carbon dioxide (CO2) has become a synonym for gases that accelerate global warming, other compounds used in healthcare have a far greater impact – such as the anaesthetic gas desflurane (C3H2F6O), or sulphur hexafluoride (SF6), which is used as a contrast agent in ultrasound imaging and has a global warming potential (GWP) 22,900 times greater than CO2. ‘For this reason, the NHS has announced that it will completely phase out the use of desflurane by the beginning of next year and re- place it with lower-GWP anaes- thetic gases,’ Dr Sheard reported. The huge impact of HVAC Radiology may be largely built around the use of huge, power- hungry scanners, but the actual share of these systems on the ener- gy consumption of a typical radiol- ogy ward is comparably small, the expert continued. Recent investi- gations have shown that the jour- ney of a patient through a interven- Carbon emissions from electricity and consumables (in kg CO2e) of imaging modalities, as estimated by attributional (ALCA) and consequential (CLCA) life cycle analyses. © McAlister et al., Lancet Regional Health – Western Pacific 2022 (CC BY-NC-ND 4.0) tional radiology unit generates substantial GHG volumes, a major- ity of which come not from the im- aging itself, but from energy used to maintain climate control, fol- lowed by emissions related to single-use surgical supplies. Ac- cording to this paper, HVAC – heat- ing, ventilation and air con- ditioning – contributes to almost half of the total emissions (49.2%), while the energy expenditure (plug load) for imaging only makes up 2.78%. ‘A huge challenge is seeing the impact from heating and cool- ing our departments, especially in old, inefficient buildings,’ Dr Sheard said, pointing out the need for more energy-efficient con- struction and renovation. In imaging, however, the devil is often in the details. For one, a re- cent study found significant vari- ations in the carbon footprint of the most common modalities: the average carbon emissions per scan imaging were highest for MRI (17.5 kg CO2 equivalents), followed by CT (9.2 CO2e). By contrast, chest X-ray and ultrasound generated much smaller footprints per scan, using only 0.76 and 0.53 CO2e, re- spectively. Strikingly, the amount of energy used for a single scan on a previously inactive system was far greater than that for any ad- ditional scans performed on a sys- tem that was already running. This, the radiologist said, makes a strong case for operating imaging systems at full capacity whenever possible. and tests Sustainability can be further im- proved by reducing overdiagnosis, unnecessary inter- ventions, the expert suggested: ‘The greenest scan is the one you don’t need to do.’ This mindset re- duces emissions from the energy required by the scanner itself but may also avoid additional patient travel – another major source of healthcare-related GHG emission. for sharps and Waste management Another area with huge potential for improvement is waste manage- ment, Dr Sheard said, pointing out how incorrect disposal of consum- ables can greatly enlarge any pro- cedure’s carbon footprint. For example, contrast bottles, tubing sets and residual contrast are often indiscriminately thrown into the bin infectious waste, she reported. This type of waste is incinerated – a highly car- bon-intensive procedure that gen- erates 1.85 tonnes of CO2 for every tonne of waste. ‘Often, only a very small part of the equipment – for example, the spike used for injec- tion – needs to go into the sharps bin. All of the plastic tubing can be thrown into the non-infectious/ offensive waste stream.’ Leftover contrast agent can be col- lected in a special tub and sent to the supplier for proper disposal, while the rinsed contrast bottles are perfectly fine to go into regular municipal recycling. The expert urged that staff must be educated on this topic: ‘Often, people just lack confidence to sep- arate things, and are worried about risking contaminating waste streams. Therefore, they take the safest street, which is to put every- thing into the highest clinical haz- ard category. Much more clear in- struction is needed here.’ In a radiology ward with two CT scanners, where this approach was put into practice, Dr Sheard re- ported that the amount of waste disposed in the sharps/infectious bin was reduced by 98%, saving Sarah Sheard Dr Sarah Sheard is a consultant car- diothoracic radiologist and Head of Thoracic Imaging at Imperial College Healthcare NHS Trust (ICHNT) in Lon- don, UK. At ICNHT, Dr Sheard chairs a cross-disciplinary Imaging Environ- mental Sustainability Group and sits on the hospital trust Green Advisory Com- mittee. € 11,300 and over 20,000 kg of CO2 equivalents per year. In conclusion, Dr Sheard pointed out that there are many more ways to reduce the carbon footprint of radiology – for example, by using renewable energy sources, or buy- ing imaging systems according to their environmental impact. How- ever, the most important factor as well as arguably the biggest chal- lenge is behavioural change, she appealed to her colleagues: ‘Any- thing you can do to normalise sus- tainable healthcare, to start the dis- cussion in your department and empower the people around you, is going to be very valuable. Fin- ally, get a team together, network, and use your position as clinician to bother the chief executives, to put this issue on their agenda.’ ■ Report: Wolfgang Behrends The ESR‘s Green Meeting Concept – a mission statement Acknowledging the impact of climate change, the European Society of Radiology (ESR) has included sustainability into its guid- ing principles. As an umbrella organisation for European radiologists with over 130,000 members, it has a considerable voice in promoting eco-friendly practices for radiology. This is reflected in the society‘s “Green Meeting Concept”: ‘The European Society of Radiology (ESR) is a progressive organisation, committed to supporting sustainable development and the right to live in a healthy, ecologically diverse environment. We encourage our employees, partners and customers to adopt practices that promote awareness of environmental conservation and sustainable use, and we endeavour to adhere to these principles throughout our entire enterprise. In our continuous quest for innovation, we will always pay close attention to good economic practice, ecological concerns and social responsibility, in order to achieve several aims: innovative events, conser- vation of natural resources, and human well-being. Our organisation observes these principles of sustainability according to the following guidelines: The environmental impact and sustainability of any new activity or project will be assessed in advance. All necessary measures will be taken to prevent contamination of the environment and to promote social sustainability. Our employees, partners and customers will be informed of the environmental and social aspects of our activities in an appro- priate manner. We will strive to ensure our partners meet the appropriate environmental standards. We will strive to continually increase environmentally and socially responsible behaviour within the organisation in every way possible. This philosophy consists of: An understanding of the importance of environmental protection and the provision of comprehensive information to our partners and employees. Active encouragement of the responsible use of resources such as electricity, water and paper. A commitment to buying, as far as possible, only local, eco-friendly and Fairtrade products. The promotion of environmental awareness as a major objective of the company. The involvement of all partners in this important task. www.healthcare-in-europe.com ACCUTRON® MR. THE MOBILE MRI SPECIALIST. Quick confi guration, versatile use and compatible with up to 7 teslas. Accutron® MR allows a high patient throughput. Next, please. Hauptstrasse 255 · 66128 Saarbruecken www.medtron.com

WOMEN‘S HEALTH 9 Fluorescence-guided surgery Identifying breast cancer in margins Breast surgery is a traumatic ex- perience for a woman, no matter whether breast-conservation sur- gery (BCS) or a mastectomy. Trauma levels are greatly en- hanced, if pathological evalu- ation findings of an excised breast tumour following a lum- pectomy suggest that additional cancer may still be in the mar- gins, and a second surgical pro- cedure is required. A new system with the ability to accurately identify residual tumour in the operating room could reduce re- excision rates, and potentially allow surgeons to excise the tu- mour with narrower margins. imaging fluorescence An intracavity visualisation system using to identify residual cancer in a tu- mour bed during a surgical pro- cedure has been successfully tested for performance and safety in a prospective randomized clinical trial involving more than 400 pa- tients undergoing BCS at 14 US cancer treatment centres. Results from the Investigation of Novel Surgical Imaging for Tumor Ex- cision (INSITE) trial revealed that the investigational Lumicell Direct Visualization System identified cancer remaining after a lumpec- tomy procedure in 7.6% of the par- by from ticipants it imaged. The system, de- veloped biotechnology company Lumicell, made 22 of 27 findings deemed negative by standard histopathol- ogy assessment. In addition, 14.5% of patients with positive margins had all margins cleared and were spared a second surgery. sites The study also evaluated the safety of Lumicell’s investigational optical imaging agent Lumisight (peguli- cianine), which becomes fluor- escent at residual tumour sites. Pe- gulicianine had an excellent safety profile, consistent with other im- aging agents used in breast conser- vation surgery, reported INSITE’s lead investigator Barbara Smith, MD, PhD. Complete tumour ex- cision during BCS remains a huge challenge because preoperative im- aging and current intraoperative examination methods do not ac- curately identify the microscopic extent of tumours. An estimated 20% to 40% of lumpectomies have positive margins, requiring a sec- ond surgery several days later fol- lowing histopathology analysis. A system with the ability to ac- curately identify residual tumour in surgical margins in the operating room not only could reduce re- excision rates, but also potentially allow surgeons to excise the tu- mour with narrower margins. The pegulicianine fluorescence- guided system (pFGS) includes the injectable activatable fluorescent imaging agent, a handheld probe, and a patient-specific tumour de- tection algorithm. Images are viewed on a computer screen in the operating room. All patients were administered 1 mg/kg of the pegulicianine dye two to six hours before surgery. After receiving standard-of-care lumpectomy, 35 patients were randomised to the control arm, and the remaining were assigned to the treatment arm. For the latter group, imaging with the system following tumour excision added approximately 10 minutes to surgical suite time. Real-time images from inside the breast cavity, based on a patient-calibrated can- cer detection software, guide the surgeon in the removal of residual cancerous tis- sue. © Lumicell each positive When intraoperative images dem- onstrated a red signal indicating pFGS uptake, the surgeons excised additional shave margins in the in- terior and inferior lateral orien- tations, until the signal was elimin- ated from the image displayed on a monitor. The researchers assessed the diagnostic accuracy of pFGS by comparing or negative image to the histopath- ological margin assessment at that site. They reported that pFGS de- tected tumour tissue that had been left behind by standard lumpec- tomy in 27 patients. In the 17% who had positive margins after standard surgery, nine of 62 pa- tients had all the margins cleared by pFGS, potentially avoiding a second surgical procedure. This technology is still in an inves- tigational phase, and additional clinical trials are needed. In ad- dition to breast cancer, its use is being explored for excision of prostate cancer, brain cancers, gas- trointestinal cancers, sarcoma, and peritoneal metastases. To date, over 800 patients have been clinically evaluated in seven different cancer types at 18 US medical centres. ■ Report: Cynthia E. Keen MRI controversial Breast cancer screening of high-risk women Breast MRI is increasingly being used as a primary screening breast cancer screening exam for young women. The technique is also being ordered as a supple- mental screening exam following mammography or breast ultra- sound for women of all ages at high risk. But use of breast MRI as a screening tool is con- troversial. to perform It takes longer for patient set-up and than mam- mography, may be a scarce im- aging resource for radiology de- partments, and is very expensive compared to the cost of mam- mography and breast ultrasound exams. However, several recent studies strengthen the argument for breast MRI use as an essential breast cancer screening exam for high-risk women. Women who have extremely dense breasts are four to six times more likely to develop breast cancer than women with mostly fatty breasts. BRCA1 and/or BRCA2 gene mutations are an additional risk factor, putting those affected at an estimated 60% risk of devel- oping breast cancer, compared with 12.5% of women worldwide who have normal BRCA1/2 genes. Malignant breast tumours of BRCA carriers double in volume very rapidly, increasing their size up to four to five times over twelve months. This rapid tumour size growth, combined with the fact www.healthcare-in-europe.com aggressive that BRCA1 carriers tend to devel- op triple-negative cancers, makes early detection cru- cial. like that look The sensitivity rate of MRI in de- tecting early breast cancer in BRCA carriers is as great as 96%, but cancers sub- centimetre lesions with benign ap- pearing features may be over- looked. Researchers at Sheba Medical Center in Ramat Gan, Is- rael, have developed and clinically tested an artificial intelligence (AI) tool designed to accurately classify enhancing foci seen on breast MR images to help prevent them from being misinterpreted as benign. In Cancer case of a successful classificati- on of the AI network. © Anaby et al., Cancers 2023 (CC BY 4.0) its first evaluation test, the AI tool accurately identified more than half of the foci which were initially considered benign but identified as malignant twelve months later. The researchers had identified 53 BRCA carriers who were diagnosed with breast cancer between 2012 and 2021, and who had had two con- secutive breast MRI exams within an 18-month period. They were matched with cancer-free BRCA carrier controls who had been seen at the same clinic with breast MRIs performed in the same time frame. Prior MRIs of both groups were used for analysis, with subsequent diagnostic MRIs serving to identify the tumour area and its delin- eation. Radiologists manually marked the central slice of the tu- mour on the sub-dynamic contrast enhanced (DCE) image and used anatomical landmarks to mark the corresponding region in the prior scan. They marked lesions or ab- normalities reported in the prior MRI of cancer-free women, or if there were none, marked a promi- nent enhancing focus. The cor- responding region of the sub- sequent MRI was also marked. Lead author Debbie Anaby, PhD, and colleagues report that the AI tool successfully classified 65.6% of cancer cases as cancerous and 88.7% of cancer-free cases as non- cancerous. Triple-negative tumours were more frequently successfully classified as cancerous compared with other histological tumour types. Interestingly, in the cancer- free cohort, almost 40% of en- hancements detected in the prior scans did not appear in the follow- up MRI. The researchers are plan- ning a large prospective test and broadening the AI network to an automatic detection of suspicious abnormalities. ‘The ability to detect triple-negative breast cancer at such an early stage, while ab- normalities are mostly foci with no rim-enhancement, may reduce the clinical burden of the disease,‘ they write. Who specifically benefits from breast MRI screening? The Breast Imaging Clinic at the University of California San Fran- cisco (UCSF) uses both mam- mography and breast MRI as screening exams for high-risk women. These patients have an an- nual mammogram, followed by a breast MRI six months later. ‘Does breast MRI cancer screening in high-risk women detect early-stage breast cancer?‘ Lauren Ton, a third- year medical student, asked rhe- torically during a scientific session of the Society of Breast Imaging’s (SBI) annual meeting. ‘Yes! Our breast MRI screening detects 99% of cancers before tumour stage 2.’ UCSF recommends that high-risk patients start having mammograms (or breast ultrasound if very young) in their late 20’s or early 30’s. If a family member has had breast cancer, they recommend screening start ten years before the age the family member was diag- nosed. Ton and colleagues con- ducted a retrospective study to re- the clinicopathologic and view imaging outcomes in dynamic- enhanced breast MRIs of patients diagnosed with breast cancer, in- vestigating who specifically bene- fits from breast MRI screening. The study included all high-risk pa- tients with diagnosed breast cancer who had breast MRI screening be- tween 2012 through 2022, a total of 79 women. They ranged in age from 27 to 71 years. Forty-six per- cent were pre-menopausal. Most (63%) had heterogenous fibrog- landular tissue; only 14% had ex- tremely dense breasts. BRCA car- riers represented 42%, 32% had a personal history of breast cancer, and 11% had a family history. The most commonly detected cancer on breast MRI was invasive ductal carcinoma (55%), with or without ductal carcinoma in situ. Invasive lobular carcinoma represented 6%. The median lesion size was 1.2 cm, with the majority being mass. Fifty- nine percent were detected at tu- mour stage 1, and 40% at tumour stage in situ. Only 4% of these early-stage breast cancers had axil- lary lymph node involvement. Ton: ‘These results support the growing role of screening breast MRI as it enables early breast cancer detection.‘ ■ Report: Cynthia E. Keen

GENDER MEDICINE 11 Knowledge gaps Covid-19: higher mortality of male patients The Covid-19 pandemic has af- fected scientific research in nu- merous ways – for example by highlighting knowledge gaps in gender medicine. In many studies differences in morbidity and mortality between women and men surfaced incidentally. While the extent and causes of these differences remain largely unexplored, the preliminary in- sights confirm the need for further research. As early as spring 2020, studies of Covid-19 patients in China indi- cated that mortality among men was higher than among women. Data from Italy corroborated these findings and at the same time sug- gested that more women than men were infected. The latter result however was not confirmed con- sistently: other studies reported either rates among men and women or even higher rates among men. Quickly, a causal relationship was estab- lished between the expression of the ACE2 receptor, which works differently in men and women. ACE2 is an enzyme whose mem- brane-bound form is SARS-CoV-2’s entry point to cells. infection identical In many tissues of the female body, ACE2 expression is higher than in tissues in the male body, among others due to the fact that the ACE2 gene is located on the X chromo- some. The genes on the second X chromosome of women are largely Tackling gender inequality “silenced”, i.e. they are not ex- pressed in the quantity, as one would assume, but are inactivated. This silencing however is incom- plete: in the presence of two X chromosomes ACE2 is expressed 110% compared to the expression rate when only one X chromosome is present. This might explain the higher incidence of Covid-19 in women: women, the assumption goes, have more entry points for SARS-CoV-2 on their cell mem- branes. But how can the lower mortality of women be explained? Lower mor- tality in women remains statis- tically significant even when the data is adjusted for other factors such as age, life style, previous dis- eases, time of onset of therapy, etc. The same holds true for the fact that men are at a higher risk to de- velop a severe Covid-19 infection. The stats: a meta-analysis of 90 studies has shown that men com- pared to women have a so-called odds‘ ratio (OR) of 2.84 for ICU ad- mission due to Covid-19 and an OR of 1.39 for Covid-19-caused death. One possible explanation is the fact that the genes of certain pro- teins that the immune system needs are located on the X chro- mosome and thus are slightly over- expressed in women compared to men. An example would be TLR7, which plays an important role in fighting viral infections. The male immune system and the cy- tokine storm to an response immune system In a system that is as complex as the human it would be misleading in a given situation to talk about a “stronger” or “weaker” defence against a cer- tain virus and consider the first one to be tied to a more favourable prognosis. Indeed, certain func- tional impairments or organ da- mages are caused by the activities of the immune systems rather than by the virus itself. Following a Covid-19 – or other – infection, men are more likely to develop a so-called cytokine storm, i.e. an ex- cessive in- flammation which is caused among others by the cytokine IL-6, a pro- tein that is expressed in men to a higher degree than in women. The damage caused by the cytokine re- sponse, which often leads to severe acute syndrome (SARS), explains for example the efficacy of glucocorticoids, which in fact are an immunosuppressant, in severe Covid-19 cases. Moreover, there are indications that reduced testosterone levels in male patients might improve their Covid-19 prog- nosis. The beneficial effect of oes- trogen on Covid-19 mortality was shown in a Swedish study which included postmenopausal women. Patients who received an oestrogen replacement had lower Covid-19 mortality rates than patients with- out such hormone substitutes. Whether the opposite is also true, i.e. patients with an anti-oestrogen respiratory © Mourosi et al., Infection, Genetics and Evolution 2022 (CC BY 4.0) therapy after a breast cancer diag- nosis have a higher Covid-19 mor- tality due to the lower oestrogen levels, has not been established conclusively. Again, the interaction is complex: some researchers assume that the less severe Covid-19 course in women can be attributed among others to the lower (rather than higher) expression of ACE2 par- ticularly in the lung tissue of women. Gender medicine: relevant for all, no matter which sex This brief look into the research on Covid-19 and sex shows that the relationships and interactions be- tween sex hormones, immune sys- tem and specific viruses are com- plex. A better understanding of these factors could improve ther- apies for given diseases for both sexes as it might pave the way for new approaches or confirm effi- cacy of known medicines. Some re- searchers point out that the role of sex was insufficiently taken into consideration during the pan- demic. A review found that only about 18% of the clinical studies published during the first year of the pandemic contained gender- specific sub-group analyses. A re- cent survey conducted with the participation of the Institute for So- cial Medicine at the Charité and the German Medical Women‘s As- sociation (Deutscher Ärztinnen- bund) found that only nine of the 16 interviewed virologists were aware of the fact that there are sex differences in Covid-19 epidemiol- ogy; only three considered sex-spe- cific differences in their discipline, virology, to be important. New ap- proaches are required if infectiol- ogy and immunology are to prog- ress towards truly individualised medicine. ■ Report: Dr. Christina Czeschik Radiology – still a “man’s world”? A panel of experts assessed the challenges women have to face in radiology and explored strat- egies to help take gender in- equalities at ECR 2023 in Vienna, Austria. As populations become increas- ingly diverse, organizations need to adapt, Laura Oleaga, MD, PhD, told the audience. ‘Gender equity is a key factor in achieving excellence in academic medicine,’ the pro- fessor of radiology said. ‘It con- tributes to excellence in all pro- fessions, allowing for a global vision of all situations from dif- ferent perspectives and sensi- tivities.’ So far, however, this is only partly reflected in reality: While the pro- portion of women in academic medicine is similar to men, the pro- portion of female professors is sig- nificantly lower, the expert pointed out. ‘In 2019, the proportion of women heads of institutions in higher education was 23.6%, and they made up 31.1% of the members of scientific boards,’ she said. ‘Women represented 33.8% of all researchers in the EU in 2018, and as we go up in the career pyra- mid, the number of women de- creases.’ Forty-eight percent of all www.healthcare-in-europe.com doctoral students were women, while they represented 24% of the highest roles in academia. In Eu- rope, women represent 54% of physicians and 40% of radiologists. Female representation in radiology decreases at increasing levels of leadership, she noted. Women underrepresented in crucial future fields Radiology is a technological field, an area where women remain a mi- nority, Prof. Oleaga pointed out. ‘There are less women in digital in- formation technology, computing, physics, mathematics, engineering and AI. We have to pay attention to that, because these are the fields that are driving many of the future fields jobs.’ The percentage of women with AI skills is much lower than men, and women are also leaving technological in greater numbers than men. ‘This trend makes us think that progress to- wards gender balance could be compromised,’ she said. ‘We have to promote scholarships to attract women in technology and to incor- porate and retain them in these fields.’ To influence change, in- creasing the number of female role models and mentors is essential, the radiologist explained. ‘A lack of mentorship and role mentoring may be deterring women from pur- suing radiology,’ she said. ‘Mentor- ship and sponsorship of women programs directly impact female A rare picture: women are underrepresented in radiology. © amorn – stock.adobe.com advancement and career satis- faction. Leadership and manage- ment programs represent effective ways to cultivate leadership skills to promote the professional career,’ she concluded. The next speaker, Dr Carolyn Horst from Kings College London, UK, shared her perspective as a radiol- ogy trainee with the delegates. She showed a picture of her two-year- old daughter as part of her disclos- ures. ‘Having a child as a female physician can feel like something you have to disclose,’ she said. ‘I say this jokingly, but it can feel like something that hampers you pro- fessionally.’ To prove this point, she then talked about her non-linear training story, starting medical school when she was 28, and then working as a jun- ior doctor for two years, as she was always told her career would go. Until she decided to embark on a PhD and have a baby. ‘I took an eight-month break to have my child, and now I’m back in radiol- ogy training and still have a couple more years before I reach the golden field of consultants,’ she said. ‘The truth is it doesn’t always happen the way you’ve been told it was going to happen, especially when you have a family.’ In the UK, the expectation for trainees is that they are young, willing and able to move anywhere they might be needed, may have some indepen- dent financial means if they stay in London, and have no caring re- sponsibilities, she explained. ‘This approach diverts good trainees with any or all of the above press- ures to GP and other specialties, which have short and less popular training schemes,’ she said. In the UK, only 37% of consultant radi- ologists are female, even though there has been parity of men and women in medical schools for dec- ades. ‘We have an image problem,’ Dr Horst said. In addition, there is a structural problem with regards to the way radiology is being taught. ‘The apprenticeship model is often successful when the trainer is like the trainee. So, we need women training other women.’ Finally, women with caring respon- sibilities tend to be seen as ‘uncom- mitted’ because they leave after they have completed their work to pick up children – and don’t have time to stay and talk with col- leagues. ■ Report: Mélisande Rouger

SAFETY AND VALUE. It's in our blood. TAKE CARE, STAY SAFE WITH VACUETTE® SAFETY PRODUCTS We know how much your daily work takes from you. The least we can do is keep you and your patients safe. It's why every detail of our VACUETTE® Safety Products are designed to make a difference to nurse, patient, and lab safety. Let us take care of blood collection. Greiner Bio-One GmbH / Kremsmünster, Austria / E-MAIL office@gbo.com We are a global player / Find the contact details of your local partner on our website.

Designed around 20 gram packettes of ultrasound gel, Parker’s newest warmer adds comfort to our single use packettes' long list of benefits. Like our other Thermasonic Gel Warmers, advanced microprocessor technology allows for custom temperature control, while a powerful heating element brings 20 packettes of gel to temperature quickly and eciently. Engineered for performance, durability and control. Warming up Ultrasound 20 GRAMS AT A TIME The NEW The NEW Thermasonic® Gel Warmer Thermasonic® Gel Warmer For Single Use Packettes For Single Use Packettes ISO 13485:2016 Learn more at: parkerlabs.com/thermasonic © 2019 Parker Laboratories, Inc. Parker Laboratories, Inc. © 2019 P Thermasonic is a registered trademark of Parker Laboratories, Inc. Thermasonic is a registered trademark of Parker Laboratories, Inc. ® The sound choice in patient care is a trademark of Parker Laboratories, Inc. The sound choice in patient care is a trademark of Parker Laboratories Inc AD-83-2 REV 0 AD-83-2 REV 0