EH 6_2015

SPECI A L:DIGITALPATH O LOGY www.healthcare-in-europe.com 9DIGITAL PATHOLOGY Peter Hufnagl, Head of Digital Pathology, Charité Berlin A pathologist’s work is undergoing radical transformation Experts are discussing data volumes not yet seen in medicine I saw the future of pathology – and it’s digital Automation in pathology Healthcare is going digital. No doubt about it, Prof. Hufnagl predicts. Information and communication technologies have gone beyond moving data from one place to the other; they are triggering stellar improvements in healthcare: diagnoses are becoming ever more precise, therapies ever more personalised. The extent to which the individual clinical disciplines have progressed in their technological development varies greatly. In pathology, for example, change has only just set in. However, it is already obvious, that digitisation will change the discipline forever. Karoline Laarmann reports Surveying medical developments over the last two or three decades confirms the enormity of progress made during this period. Going into details would go far beyond the scope of this article, writes Walter Depner. ‘However, one topic should be highlighted – pathology – because it has bucked the general trend in a certain way Scepticism towards digital systems is still widespread and fundamental questions regarding the sense and purpose of the new technology in the pathology department continue to be raised. These concerns are primarily caused by the significant necessary financial investments in hardware and software. Basic equipment encompasses slide scanners, data storage devic- es and an image management sys- tem. There are many opinions as to whether and when the purchase of these digital pathology devices makes economic sense. However, there are – throughout Europe – very encouraging examples that dem- onstrate the successful step-by-step transformation from microscope- based histopathological diagnostics to digital diagnostics. ‘Unfortunately, these positive examples are not sufficiently publi- cised in the pathology community’, says Professor Peter Hufnagl, Head of Digital Pathology at the Institute of Pathology at Charité, Berlin, and President of the13th European Congress on Digital Pathology in Berlin in 2016. ‘The two types of workflow, digital and analogue, can co-exist very well and, for some institutions, such a two-pronged approach might be the best solu- tion in the long run. That means you don’t have to wait until you can afford fully-fledged digitisation. Start small!’ Workflows in the pathology lab will benefit with only one third of the diagnostic task being digitised. Communication with clinicians will become faster, simpler and more effective. For large institutes involved in research and teaching complete digitisation makes sense, Professor Hufnagl underlines: ‘The bio-banks being established in the academic environment particularly profit from a digital slide archive, since it allows larger sample groups to be selected and to check whether certain slides are suitable for a research project.’ While many pathology labs are still hesitant to jump on the digital bandwagon although the industry has recognised the market potential long ago and is actively driving its rapid growth through technological innovation. The fact that the use of a virtual microscope saves a lot of money on expensive samples is only one of the arguments manufacturers use to lure new customers. More importantly, however, digital pathol- ogy optimises quality assurance and thus contributes to a lab’s competi- tiveness. Digitisation provides a bet- ter overview of comparable cases; it allows the application of quantita- tive methods to check diagnoses and accelerates obtaining second opinions. Hufnagl is convinced that it is only a matter of time before digital pathology will prevail. In radiol- ogy, he points out, there were very similar concerns regarding on-screen readings – today this is standard operating procedure. The fact that, due to image volumes in pathology – 150.000 x 300.000 = 45000Mpixels – digitisation is taking root 20 years later than in radiology, is an advan- tage. We can learn from radiologists’ experiences and from many of their solutions, because they can be trans- lated into the questions that need answering in pathology.’ Having said that, there are indeed also fun- damental differences between the two image-based disciplines. The different CT slices, for example, are automatically adjusted – a step that is not so easy with histopathological samples that consist of three-dimen- sional tissue, such as a tumour. ‘The registration of slides with differ- ent stains from the identical block is sometimes difficult. Differences between slides, such as pressure points, tiny fissures etcetera are originated by mechanical alterations in the lab or changing the lab during processing. Software programmers have not yet found a solution to deal with all different artefacts in a suf- ficient manner.’ In addition to diagnostics with the virtual microscope, digital pathol- ogy is being pushed ahead by other technological innovations, such as molecular pathology, particularly Next Generation Sequencing (NGS). This procedure can detect gene mutations much faster and simpler than before. This allows inter alia precise characterisation of tumours, which in turn translates into custom- ised cancer therapies. Today NGS is considered the path-breaking tech- nology towards personalised health- care, because it is expected to have a crucial impact on diagnosis and prognoses of cancer and other dis- eases. ‘Thus digitisation makes pathol- ogy much more visible in the land- scape of clinical disciplines’, Peter Hufnagl confirms, adding: ‘This development is being supported on the organisational level, for exam- ple by the establishment of tumour boards that include pathologists and it will be accelerated by the new technologies.’ Consequently not only the work environment and the ‘job descrip- tion’ of a pathologist are undergoing radical transformation; the image of the pathologist is also changing. For years, pathology has been strug- gling with recruiting problems. Digitisation is giving the field a fresh, modern look and more jun- ior physicians are becoming interested in this field. For years, even decades, proven and successful methods and procedures have been used in pathology, involv- ing intensive manual work, with a relatively little automation. This dif- ference between other medical dis- ciplines is now seeing rapid change. Automation in pathology is now frequently discussed in many publi- cations, congresses and workshops etc. – as well as in our journal European Hospital, with topics from this field more frequently featured. In one such example (issue 4/2014, page 7) Professor Alessandro Lugli, at the Institute of Pathology in Bern, Switzerland, details the introduction of the Lean Management System in pathology. The system was adapted from the automotive industry, in which it is utilised to improve quality as well as save costs and thus increase efficiency. After three years of using the Lean system in Bern, Lugli could confirm those benefits for the Institute of Pathology as well. In this issue you can also read the opinions of Dr Bela Molnar (Budapest), one of Europe’s pio- neers in pathology automation. We are now entering a further stage in the development of modern pathology, or respectively of auto- mation in pathology – digital pathol- ogy. What does that term mean? Simply, it stands for the increased use of IT systems and, in other words, the transformation of histo- pathological results revealed with an analogue microscope to diagnos- tics, or the computer respectively. This requires high-resolution scanning of samples of tissue sec- tions so that they can be digitised in the computer. In theory, a tissue sample under the microscope offers several thousand fields of vision, but only one or two percent are actually relevant. With a digitised object however, after further assess- ments and documentation, we can immediately return to the important parts. This procedure also makes exchange with other national and international experts far easier. With the help of teleconferencing, or telepathology, experts can exchange knowledge and experience or seek second opinions on digitised histo- logical samples in real-time, which obviously also enhances patient care. However, it still remains very uncertain as to what extent and how quickly digital pathology will become established in clinical rou- tine. The systems are currently mainly used in academic and scien- tific settings, where digital sample scanners and the respective archiv- ing systems are predominantly used for teaching purposes. On the other hand, demand for diagnosticians is growing and pathologists also need to deal with an increasing number of non-medical tasks and issues. Says Dr Gian Kayser, Senior Consultant at the Institute for Pathology at the University Hospital Freiburg: ‘The technology will cer- tainly become more affordable and of higher quality in the foreseeable future. Within the next five to ten years an increasingly growing number of pathologists will be working completely digitally.’ Radiology is a particularly appro- priate example. We should bear in mind that digitisation in this field literally began more than 40 years (!) ago with the introduction of CT scanners, at a time when no one really believed in a real breakthrough and introduction of this technology on a broad scale. Transferred to the current situa- tion in pathology we can just about imagine the massive changes this medical area will face in the near future. However, unlike in radiology, one of the main problems here is the enormous amount of data. Experts are in fact talking about the kinds of data volumes not yet seen in medicine. Of course, even in the world of radiology it was initially hardly conceivable how the PACS (Picture Archiving and Communication Systems) would cope with information overload. Nowadays, a radiology practice is basically inconceivable without ‘dig- ital diagnostics’. And maybe the same will apply to pathology in 20 or 30 years’ time? As in radiology, the introduction of such a complex system will generate enormous costs and administration and data archiv- ing (with data retention require- ments of many years) will play an important part. Let’s hope that Kayser’s prognosis is correct.

Pages Overview