Next‐generation pollen monitoring and dissemination

摘要

Although pollen allergy is the most frequent allergic disease in countries with a western lifestyle,1 knowing “the enemy,” that is, the current pollen exposure, is still a neglected scientific topic. This is remarkable, as pollen levels predict allergy symptoms in allergic individuals.2 In addition, higher pollen exposure results in more allergic sensitizations.3 Thus, knowing the local pollen count is of relevance for allergic individuals. Instrumental to the measurement of ambient pollen was the development of the manual Hirst trap in 1952, a simple and cheap but efficient instrument able to assess airborne pollen. And indeed, pollen is now monitored in almost every country of the world with this instrument,4 with some time series dating back to 1952. The EAACI conference 2018 in Munich showed that the field of pollen monitoring is now rapidly changing. In Bavaria, Germany (the county of the conference), a network of 8 automatic pollen monitors (robots) was installed this year. Novel is that the network called ePIN (electronic Pollen Information Network) is state-owned (www.lgl.bayern.de), circumventing the problem of declining finances many private pollen monitoring networks face. Second, data are delivered to any person online and for free, and third, the speed of delivery of data is within hours instead of days as in the conventional setting. Pollen data are now available online as soon as the pollen robots report their findings. The next novelty is that the dissemination of the data is revolutionized too. At the EAACI conference 2018, the worldwide first and only “Pollen Indicator” was unveiled (see Figure 1). The purpose of this “Pollen Indicator” was to show in one glance which pollen, and how many are in the air right now. The “Pollen Indicator” does not measure pollen, but instead is linked to a robot that does the job.5 The “Pollen Indicator” makes a light show of the current pollen levels, delivered by the robot. Horizontally at the top, a running screen with bright LEDs shows which pollen is currently displayed. Although thresholds for allergic symptoms due to pollen are still debated and vary with patient and location,6 vertically, a color-coded LED-column shows how much symptoms an allergic individual might expect. When the whole “Pollen Indicator” turns red, >100 pollen/m3 of the displayed pollen is in the air, and symptoms are expected to be severe. When pollen is <25 pollen/m3, a smaller light column glows up green, telling individuals that no symptoms are expected, even for sensitized individuals. A QR-code on the “Pollen Indicator” links to a website (www.zaum-online.de/pollen) that explains in more detail the ins and outs of what is shown. The “Pollen Indicator” makes pollen visible. The robot uses image recognition, a field that is changing rapidly, and although currently only three automatic pollen monitors are on the market, probably more will follow. The oldest automatic pollen monitors are installed in Japan. In Japan, one pollen type is dominant and of concern: those of the Japanese cedar (Cryptomeria japonica) to which most Japanese are allergic. Thus, reporting Japanese cedar pollen is enough. In other countries, this is not sufficient and the information which pollen type is flying is critical. Then, only two instruments with completely different working mechanisms are available. The BAA (Bio-Aerosol-Analyzer) is a robotic microscope that, like humans, recognizes pollen by image recognition.5 The second instrument is an “air flow cytometer,” measuring optical parameters of a pollen when it passes through two red and then an UV laser.7 Although the performance of both instruments is not perfect and does not beat the human eye, when running large numbers of samples in a network of pollen monitors, the “human eye method” is not without errors too. Simplified, both make similar errors but each of another kind. Importantly, the robots’ major error is “I don′t know this pollen” and seldom wrong classifi