Seite - 318 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

fer fromasthmaworldwide [1]. In order to achieve a good symptoms control and limit theriskof futureexacerbations,aproper long-termasthmamanagement is indispensable [2]. Inhaled pharmaceutical therapy is themost commonmedical treatment for asthma. In order to fully benefit from the drugs, the prescribed dose has to reach the airways. Therefore,apoor inhalationtechniquecandecreasetheefficacyof theinhaler.Foranim- provedasthmacontrol, patientsmust respect the timeaswell as theprocedureof taking themedication. Thus, temporal adherence and inhalation technique adherence are key successelements forasthmamanagement [3,4]. It isestimated that70 to80%ofpatients do not use their inhaler correctly [4,5,6]. In fact, several factors are responsible for a non-adherence to the instructed inhalation technique, suchasan inappropriate inhalation flowrate, exhaling into the inhalerwhenusingadrypowder inhaler (DPI)andnothold- ing themedication longenough in theairways [6,7]. InastudyconductedbyLavoriniet al. [8], itwas shown that inappropriatedrug inhalationprofile iswidelycommonamong asthma patients. For example, sometimes patients inhale too fast (>90 L/min) using pMDIs [9] or too slow to reach the recommended flow rate (>30L/min) usingDPIs. This leads to anon-adherenceof themedicationuse [10].Thedamage causedbyanon adherence tomedication use varies frombeingminor to severe. It can affect symptoms suchascoughingandwheezing,but canalso lead tomore seriousconsequences suchas frequent asthma exacerbations and frequent hospitalizations, and thus highermorbidity andmortality rates [11,12,13,14]. Differentmethods have been developed tomonitor the adherence tomedication of users. Thesemethods canbe summarized in twomain categories: subjective andobjec- tivemethods [11,15]. The subjectivemonitoringmethods aremainly based on surveys, which consists of patients’ self-reports andobservationsmadebyphysicians.However, thefact that thesemethodsarebasedonreportedinformationmakesthempronetoahigh riskofbias[11,16,17].Ontheotherhand,objectivemonitoringtechniquesincludeavari- etyofbiochemicalmethodsandelectronicmonitoring systems.Biochemical techniques can accurately evaluate the adherence. However, they can be expensive, interventional anddemanding in termsof special technical expertise for their implementation [11,18]. Electronicmonitoring systems are based on built-in electric sensors and are developed for inhalersoften tocapture thefrequencyof themedicationuse.Toourknowledgethere are no non-invasive technologies that can objectivelymonitor the inhalation flow rate inDPIandpMDI inhalers [11,16,18,19,20,21].Most of these techniques are expensive, difficult touseandfocusonlyon temporaladherence.There is thereforeaneedforalter- nativemonitoring solutions thatmitigate theabove-mentionedshortcomings.Moreover, mobile health applications are currently becomingmore popular in our daily life, espe- cially forpatientswithchronicdiseaseswhoneedaregularcontrolof theirhealthcondi- tion.Dueto theirprovenusabilityandefficiency,developinghealthmonitoringsolutions usingsmartphonesappears tobeverypromising [22]. In this paper,wepropose anovel sound-based inhaler system that can remotely as- sessboth the inhalationadherenceand the inhalation technique.Byanalyzing the sound generatedduring the inhaleruse, theacousticmonitoringsystemcanprovideanaccurate assessmentof the inhalationprofile. In this context, a new solution has been proposed in [23] for the monitoring of asthmapatients. This solution comprises an inhaler add-on devicewhich is attached to the inhaler; it allows anaccurate assessment offlowrate during inhalationbasedon the soundgeneratedby the add-ondevice, and can therebyprovide an estimationof the in- Z.Jeddi etal. /Estimationof InhalationFlowParameters forAsthmaMonitoring318