AbstractsMedical & Health Science

abstract__Abstract__ In recent years the interest for the application of an electronic nose (eNose) in medical Diagnostics is increasing. There is a need for this since eNose Diagnostics is non-invasive, easy to run, fast and cheap. The eNoses, now on the market, however, turn out to be unsuitable for large-scale application. This is mainly due to insufficient reproducibility of measurement results. In this work an eNose is used which is cheap and suited for mass applications. The usage of advanced temperature control guarantees the reproducibility between eNoses. In practice, this means that a once developed analysis model for a specific disease easily can be transferred to any number of other eNoses. Application of mass-produced components keeps the cost low. In the research it is shown that temperature variation is the main cause of the significant differences in measurement characteristic between the metal oxide sensors on which the eNose is based. To illustrate the practical applicability pilot studies are described for sepsis (bacterial infection of the blood), tuberculosis (TB) and metritis (infection of the uterus in cows). In the sepsis and metritis studies the measurements were conducted in the headspace of the blood cultures and uterus mucus respectively. In the tuberculosis study the exhaled breath of patients analyzed. For the sepsis diagnostics 30 eNoses are used to identify 11 to identify clinical relevant pathogens in blood. The eNose can significantly speed up the diagnostic process: on average 78% of the pathogens were correctly identified within 6-8 hours after inoculation in contrast to the 24 hours typically needed with the current methods. The TB-study was conducted in Dhaka (Bangladesh) with 3 eNoses. It turned out to be possible to distinguish between healthy people and those with active TB infection [sensitivity 93.5%, specificity 85.3%] but also to identify an active TB infection in a group of TB suspects [sensitivity 76.5%, specificity 87.2%]. These results are significantly better than the much-used screening test based on microscopy. Currently there is no objective diagnosis for metritis. A vet performs the diagnosis based on a number of characteristics such as temperature and appearance of the sample. The eNose proved to be more reliable and objective than a control panel of veterinarians [sensitivity 100%, specificity 91.6%].markdown