Please activate JavaScript!
Please install Adobe Flash Player, click here for download

Digital Pathology

19 15 minutes during on-going surgery, there can be no digitisation. However, we (including Professor Peter Hufnagel at the Charité Ber- lin and Professor Gian Kayser at Freiburg University Medical Centre) believe that, more and more pathologists will mainly work digitally, and also beyond university settings, in the distant future, because the requirements from this field will continue to grow and the in- vestment costs for scanners and data storage are likely to fall due to increased competition. In our view the full potential of DP, which is as versatile as it is promising, has not yet been exhausted, by far. However, high quality standards in image generation, processing and analysis must be established independent of manufacturers. They should be the basis for the continuously growing and ambi- tious community of pathologists in diagnostics, research and tea- ching. DP can help achieve a new measure of quality, particularly in the growing field of cancer diagnostics (companion diagnostics) with immunohistochemical bio- markers such as Her2 or PD1/ PD-L1, which are decisive for treatment. The tool will become a robust, reproducible, secure, com- prehensively quantitative and observer-independent aid for diagnosis. With the help of DIA, important biomarkers, such as prognosis parameters, can be completely, quantitatively, digi- tally evaluated (e.g. Her2-FISH in Z-stack) even three-dimensi- onally. DIA is therefore superior to the previous methods, as it makes millions of tumour cells analysable, if necessary even in several layers (for FISH signals). Furthermore, observer-inde- pendent, digital evaluations will lead to a location-independent, comprehensive increase in dia- gnostic quality for certain pro- blems, be it in a large centre or a peripheral practice. Along with many well-known pathologists we believe that DP will have become an established part of a hybrid workflow con- sisting of DP and conventional microscopy in clinical routine within the next 10 years. The advan- tages speak for themselves and there is no end in sight for the rapid developments and resulting opportunities for application. I believe many pathologists are following the rapid developments in digital pathology with excitement and interest, as well as with scep- ticism. Most of them are not yet happy to swap their microscopes for computers. However, within the next ten years I believe that a hybrid diagnostics workflow, consisting of conventional microsco- py and digital image analysis, will be established. A comprehensive change to purely digital diagnostics is still a long time coming. WALTER DEPNER milar problems to those seen in radiology years ago, during the initial conversion to digital image processing. Although terabyte (TB) mass storage is now comparatively af- fordable, the amount of data involved in DP is extremely high. One case with five sec- tions can take up between 2.5 to 20 giga- bytes (GB), i.e. the equivalent of a vast sto- rage requirement of several 1,000 TB, which converts to several PB per year, assuming all cases from one large centre were to be digitally stored every year. This new dimen- sion of data exceeds the requirements in radiology by tenfold and would therefore be unprecedented in clinical medicine. By contrast, there is the new dimension of the interpretability of an immeasurable number of cells with their colourings for immunohistochemistry, CISH or FISH. New analysis software – digital image analysis (DIA) – from companies such as Definiens or Indicalab, can quantitatively evaluate all cells from one section. This can be a rou- tine section for one patient or samples from many patients via tissue micro arrays (TMAs), where more than 400 samples can be put on one slide (such as GrandMaster, 3-D-histech, Hungary). This will deliver com- pletely new and detailed views of tumours and their specific biomarkers. Although previously we could gain good and rea- listic estimates with the bare eye, we can now generate far more exact evaluations of hundreds of thousands of tumour cells in one section, for instance for each individu- al tumour cell. Until recently this was com- pletely impossible. The integration of individual parts of DP into clinical routine is already happening in inter- disciplinary tumour boards and conferences, in telepathology and through the transmis- sion of digital slides in the context of stu- dies or consultations. It has facilitated a much faster, worldwide exchange between experts and colleagues without the need to physically send sections. As there has been an abundance of new technologies and innovations over the last few years and a continuously expanding market for imaging equipment (scanners and the camera tech- nology used in them) for image processing (reconstruction and visualisation software) as well as for digital image analysis (DIA) the need for standards for diagnosis and re- search has become equally imperative. If, one day, DP was to become established in daily routine it will have to withstand comparison with the microscope in daily practice, in terms of practicability, sturdi- ness and, most importantly, speed. The microscope is unlikely to disappear. Some areas, such as quick section diagnosis where a diagnosis has to be achieved within Carol-Immanuel Geppert MD from the Institute for Pa- thology at Erlangen University Hospital

Pages Overview