The Study Of Odor Response Analysis And Odor Discrimination By Bio-electronic Nose With Implantable Electrode

Odor detection is widely applied to disease diagnosis, environment monitoring, security control, and food detection et al. With the development of bio-engineering and electronic technique, the concept of bio-electronic nose has been come up with. Chemical gas sensors in traditional e-nose are replaced with sensitive bio-materials in bio-electronic nose. Bio-materials such as cells or tissue are usually utilized in vitro, which damages their biological activity. Aiming to solving this problem, this study establishes the bio-electronic nose with implantable electrode. To conserve the intact olfactory system as working part that is sensitive to odors, the micro-electrode array is implanted into the animals’olfactory bulb for odor information acquisition. Acquired information is analyzed and processed by computer to decode odors that the animals smell. The testing results demonstrate that the bio-electronic nose with implantable electrode responds specifically and repeatedly to octanol, pentanal, and butyric acid. Besides, The response intensity is linear to the logarithmic concentration of testing odor. The life time of sensing system in this study is about 3 months. With appropriate decoding methods,3 specific odors are discriminated in about 60% accuracy rate. In order to assess the ability of sensing system detecting natural odors, the system is tested by odor mixtures. The testing results reveal that neurons respond to mixtures in two distinct ways, which are dominant and additive types.The main work and conclusions are listed as following:1. The construction scheme of bio-electronic nose with implantable electrode is adjustedAccording to odor map in olfactory bulb, collating electrode implantation position by locating electrode in vitro, we implant electrode in the center of anterior part of olfactory bulb precisely in every rat, so that the bio-electronic nose respond to same odors identically. The adjustment assures that the sensing systems are sensitive specifically to octanol, pentanal, and butyric acid. Besides,3 odors incite strongest response in the selected area of olfactory bulb.2. The algorithm for decoding odor information is designed based on physiological knowledge of animals’olfactory senseThree parameters of odor response are compared. They are neural firing rate, nonlinear energy, and distribution of spike phase. The analysis results demonstrate that neural response intensity is reflected by firing rate correctly. Firing rate is selected as response parameter because of its simple calculation procedures as well. On the other hand, taking random fluctuation in testing results into consideration, testing results are modeled by Gauss distribution. After obtaining distribution functions from multiple neurons, the testing odor is predicted by maximum likelihood estimation. The prediction method is easy to achieve and accuracy rate is satisfactory, too.3. Performances of bio-electronic nose with implantable electrode are testedThe bio-electronic nose with implantable electrode is tested regarding to its specificity, repeatable rate, stability, concentration-dependent capacity, sensitivity and discrimination accuracy. Results show that the sensing system detect octanal, pentanal, butyric acid specifically, repeatedly, and sensitively in accuracy rate of 63.37%. Besides, response sensitivity is 5.84 Hz/lg[mol/L] and resolution is 0.12 1g[mol/L], and leneaity is 0.8.4. The response patterns of sensing system towards odor mixture are exploredMost natural odors consist of multiple components, so the exploration of mixed response promote the practical application of the sensing system. Arbitrary mixtures of three odors mentioned above are tested on the sensing system. The results reveal that one neuron respond to different mixtures in different ways. There are two typical response patterns,dominant and additive type. The dominant pattern refers to that the mixed response pattern is dominated by one of the component in the odor mixture. The additive pattern refers to that the mixed response pattern is the addition of two components in the odor mixture.