We would like to thank the Vodafone Americas Foundation, the Bill and Melinda Gates Foundation, and the S&#227;o Paulo Research Foundation (FAPESP) for funding this research. We would also like to thank the many faculty members from the Department of Entomology at University of California, Riverside, for their advice on this project.&#160;

1. Insect Colony and Rearing
<ol>
	<li>Mosquito Colony and Rearing
	<ol>
		<li>Rear Culex tarsalis, Culex quinquefasciatus, Culex stigmatosoma, and Aedes aegypti adults from lab colonies, which originated from wild caught individuals.</li>
		<li>Rear mosquito larvae in enamel pans under standard laboratory conditions (27 &#176;C, 16:8 hr light:dark [LD] cycle with 1 hr dusk/dawn periods), and feed them ad libitum on a mixture of ground rodent chow and Brewer&#8217;s yeast (3:1, v:v).</li>
		<li>Collect mosquito pupae into 200&#160;ml&#160;cups, and place them into experimental chambers. Alternatively, aspirate the adult mosquitoes into experimental chambers within 1 week of emergence. Make sure each experimental chamber contains 20 to 40 individuals of the same species/sex.</li>
		<li>Feed adult mosquitoes ad libitum on a 10% sucrose and water mixture. Replace food weekly.</li>
		<li>Moisten cotton towels twice a week and place them on top of the experimental chambers to maintain humidity within the cage. In addition, place a 200&#160;ml cup of tap water in the chamber at all times to help maintain the overall humidity level.</li>
		<li>Maintain the experimental chambers on a 16:8 hr light:dark [LD] cycle, 20.5-22 &#176;C and 30-50% RH for the duration of the experiment.</li>
	</ol>
	</li>
	<li>House Fly and Fruit Fly Colony and Rearing
	<ol>
		<li>Rear Musca&#160;domestica from a lab colony, derived from wild caught individuals. Catch wild Drosophila&#160;simulans individuals and rear them in the experimental chambers.</li>
		<li>Rear Musca domestica larvae in plastic tubs under standard laboratory conditions (12:12 hr light:dark [LD] cycle, 26 &#176;C, 40% RH) in a mixture of water, bran meal, alfalfa, yeast, and powdered milk. Rear Drosophila&#160;simulans larvae in a rearing chamber and feed them ad libitum on a mixture of rotting fruit.</li>
		<li>Aspirate adult Musca&#160;domestica into experimental chambers within 1 week of emergence. Adult Drosophila&#160;simulans can be directly reared in the experimental chambers. Prior to data collection, make sure each experimental chamber contains no more than 10-15 individual Musca domestica or 20-30 individual Drosophila simulans.</li>
		<li>Feed adult Musca&#160;domestica ad libitum on a mixture of sugar and low-fat dried milk, with free access to water. Feed adult Drosophila&#160;simulans&#160;ad libitum on a mixture of rotting fruit. Replace food weekly.</li>
		<li>Maintain experimental chambers on a 16:8 hr light:dark [LD] cycle, 20.5-22 &#176;C and 30-50% RH for the duration of the experiment.</li>
	</ol>
	</li>
</ol>
2. Record Flying Sounds in Experimental Chambers
<ol>
	<li>Experimental Chamber Setup 
	Note: An &#8220;experimental chamber&#8221; denotes the cage designed in our lab, in which the data was recorded. The sensor is fairly inexpensive. When built in bulk, a set up could be manufactured for less than $10.
	<ol>
		<li>Construct an experimental chamber, either of the larger size: 67 cm L x 22 cm W x 24.75 cm H, or the smaller size: 30 cm L x 20 cm W x 20 cm H. The experimental chamber consists of a phototransistor array and a laser line pointing at the phototransistor array. 
		NOTE: Additionally, the chamber consists of Kritter Keepers that are modified to include the sensor apparatus as well as a sleeve attached to a piece of PVC piping to allow access to the insects.</li>
		<li>Connect the phototransistor array to an electronic board. The output of the electronic board feeds into a digital sound recorder and is recorded as audio data in the MP3 format. See the logic design of the sensor in Figure 1.II and a physical version of the chamber in Figure 1.I.</li>
		<li>Modify the lids of the experimental chambers with a piece of mesh cloth affixed to the inside in order to prevent escape of the insects. 
		NOTE: When an insect flies across the laser beam, its wings partially occlude the light, causing small light fluctuations. The light fluctuations are captured by the phototransistor array as changes in current, and the signal is filtered and amplified by the custom designed electronic board.</li>
	</ol>
	</li>
	<li>Set Up the System to Record Sounds Produced by Flying Insects
	<ol>
		<li>Connect the experimental chamber to a power supply. Turn on the power.</li>
		<li>On the experimental chamber, find the laser lights and photoarray. Align the laser lights to the photoarray. To achieve the proper alignment, adjust the photoarray using the magnets on the outside of the experimental chamber which correspond to the magnets attached to the photoarray on the interior of the chamber until the laser light is centered on all the individual photodiodes.</li>
		<li>Perform two sanity checks to make sure the system is properly set up. 
		Note: The first step is to make sure that the system is powered, all wires are properly connected and the laser is pointing at the photo array. The second step is to conduct further checks on the laser alignment to ensure it can capture the sound of the insects&#8217; wingbeats.
		<ol>
			<li>Plug headphones (rather than the recorder) into the audio jack. Plunge hand in and out of the experimental chamber, near the photoarray, to break the plane of the laser light. Make sure the laser light is on (it will be a red beam of light) and that you break the plane of light a few times with your hand. Listen for changes in noise level as your hand goes in and out of the light beam. If you detect an audible difference, the sensor is able to capture sounds produced by the movement of big objects. If successful, move on to the next sanity check, otherwise, check if the headphone is properly connected and whether the laser is pointing at the photoarray. Adjust the photoarray accordingly until the sound of the hand moving in and out of the experimental chamber can be heard.</li>
			<li>Attach a thin piece of electrical wiring to an automatic toothbrush. Turn on the toothbrush, and plunge the wiring in and out of the experimental chamber close to the phototarray. Make sure the laser light is hitting the piece of wiring as it moves. If you detect an audible change in frequency, when the piece of wiring breaks the plane of laser light, the system is then ready to capture the sounds produced by the movement of tiny objects, i.e., insect sounds. If you do not detect an audible difference, go back to step 2.2.2 to re-align the laser lights and the photoarray.</li>
		</ol>
		</li>
		<li>After the system is properly set up, close the lid and add the insects.</li>
	</ol>
	</li>
	<li>Data Collection: Record Sounds Produced by Flying Insects
	<ol>
		<li>Turn on the recorder and make a voice annotation that includes the following information: name of the species in the experimental chamber, age of the insects, date and time, current ambient RT, and relative humidity. Pause the recording.</li>
		<li>Connect the recorder to the system, via the audio cable, and resume the recording. Leave the recorder to record for 3 days, then stop the recording.</li>
		<li>Download the data from the recorder into a new folder on a PC. Empty the recorder by deleting the data.</li>
		<li>Repeat the above recording process, until the remaining insects have died off and there are no more than 5 insects left alive in the cage.</li>
	</ol>
	</li>
</ol>
3. Sensor Data Processing and Detection of Sounds Produced by Flying Insects
<ol>
	<li>Use Software to Detect Sounds Produced by Flying Insects. 
	Note: The software (detection algorithm) is much faster than real-time. It takes less than 3&#160;hr&#160;to process a recording session, i.e., three days of data, on a standard machine with Intel(R) Core&#8482; CPU at 2.00 GHz and 8 GB RAM.
	<ol>
		<li>For each folder containing data from a recording session, run the detection software to detect insect sounds. To run the software, open MATLAB, and type &#8220;circandian_wbf (dataDir)&#8221; in the command window, where dataDir is the directory of the recording data. Then press &#8220;Enter&#8221; to start. 
		NOTE: Download the detection software circadian_wbf from reference #16.</li>
		<li>Wait until the algorithm terminates, then check the detection results. The algorithm outputs all the detected insect sounds in a new folder named &#8220;dataDir_extf&#8221;, where dataDir is the same as in the previous step. Each sound file is a 1 sec long audio clip originally extracted from the raw recording, with a digital filter applied to remove noise. The occurrence time of each detected sound is saved in a file named &#8220;dataDir_time.mat&#8221;. Observe the example of a detected insect sound in Figure 2.</li>
	</ol>
	</li>
	<li>Detection Algorithm
	<ol>
		<li>Use a 0.1 sec&#160;long sliding window to slide through the recording. The sliding window starts from the beginning of the recording. For each window, follow the steps below.
		<ol>
			<li>Compute the fundamental frequency of the current window.</li>
			<li>If the fundamental frequency is within the range of 100 Hz to 1,200 Hz, then do the following:
			<ol>
				<li>Extract the 1-sec&#160;long audio clip centering at the current window from the recording; apply a digital filter to remove the noise in the clip and save the filtered audio into the folder &#8220;dataDir_extf&#8221;.</li>
				<li>Save the occurrence time of the current window into the file &#8220;dataDir_time&#8221;.</li>
				<li>Move the sliding window to the point that immediately follows the extracted audio clip.</li>
			</ol>
			</li>
			<li>If the fundamental frequency is NOT within the range of 100 Hz to 1,200 Hz, simply move the sliding window 0.01 sec&#160;forward.</li>
		</ol>
		</li>
		<li>Repeat the process until the sliding window reaches the end of the recording.</li>
	</ol>
	</li>
</ol>
4. Insect Classification
<ol>
	<li>Bayesian Classification Using Just the Flying Sound 
	Note: Bayesian classifier is a probabilistic classifier that classifies&#160;an object to its most probable class.
	<ol>
		<li>Sound Feature Computation
		<ol>
			<li>For each insect sound, compute the frequency spectrum of the sound using the Discrete Fourier Transform (DFT). Truncate the frequency spectrum to include only those data points corresponding to the frequency range: 100 Hz to 2,000 Hz. The truncated frequency spectrum is then used in the classification as the &#8220;representative&#8221; of the insect sound. 
			NOTE: The DFT is an algorithm that transforms signals in time domain to the frequency domain. It is a built-in function in most programming libraries, and can be called in the program with just one line of code.</li>
		</ol>
		</li>
		<li>Train a Bayesian classifier
		<ol>
			<li>Use the kNN density estimation approach14 to learn the posterior probability distribution using the sound feature. With the kNN approach, the training phase is to build a training dataset.
			<ol>
				<li>Randomly sample a number of insect sounds from the data collected for each species of insects.</li>
				<li>Follow the steps in Section 4.1.1 and compute the truncated frequency spectrum for each sampled sound. The truncated spectrums together with the samples&#8217; class labels (insect species name) composed the training dataset.</li>
			</ol>
			</li>
		</ol>
		</li>
		<li>Use the Bayesian classifier to classify an unknown insect
		<ol>
			<li>Compute the truncated frequency spectrum of the unknown insect sound.</li>
			<li>Compute the Euclidean distance between the truncated spectrum of the unknown object and all the truncated spectrums in the training dataset.</li>
			<li>Find the top k (k = 8 in this paper) nearest neighbors of the unknown object in the training dataset. Compute the posterior probability of the unknown insect sound belonging to a class as the fraction of the top k nearest neighbors which are labeled as class .</li>
			<li>Classify the unknown object to the class that has the highest posterior probability.</li>
		</ol>
		</li>
	</ol>
	</li>
	<li>Add a Feature to the Classifier: Insect &#160;Circadian Rhythm
	<ol>
		<li>Learn the class-conditioned distributions of the occurrence time of insect sound, that is, the &#160;circadian rhythm for each species of insects.
		<ol>
			<li>Obtain the occurrence time of each sound from the detection results (c.f. Section 3.2).</li>
			<li>For each species, build a histogram of the insect sound occurrence time.</li>
			<li>Normalize the histogram so that the area of the histogram is one. The normalized histogram is the &#160;circadian rhythm of the given species. It tells the probability of observing an insect, of that species, in flight within a certain time period.</li>
		</ol>
		</li>
		<li>Classify an unknown &#8220;insect sound&#8221; by combining the &#8220;insect sound&#8221; and the&#160;circadian rhythm
		<ol>
			<li>Given the specific point in time in which the unknown insect sound occurred, obtain the probability of observing an insect of class based on the&#160;circadian rhythm of class. 
			NOTE: The&#160;circadian rhythm is a probability distribution. It is an array specifying the probability of detecting a sound, produced by a specific species of insects, at a specific time of day. So once a time is given, one can simply check the array to get the probability.</li>
			<li>Follow the steps in section 4.1.2 to compute the posterior probability that the unknown sound belongs to class using the sound features. Multiply the posterior probability to the results from the previous step to get the new posterior probability.</li>
			<li>Classify the &#8220;unknown sound&#8221; to the class that has the highest new posterior probability.</li>
		</ol>
		</li>
	</ol>
	</li>
	<li>Add One More Feature to the Classifier: Insect Geographic Distribution
	<ol>
		<li>Learn the geographic distribution of the species of interest, either from data published in historical records, relevant literature, or simply gather the first-hand knowledge from field technicians/biologists. For demonstration purposes, use a simulation of the graphic distribution, as shown in Figure 7.</li>
		<li>Classify an &#8220;unknown insect sound&#8221; using &#8220;flying sound&#8221; and the two additional features.
		<ol>
			<li>Given the geographic location where the insect sound was intercepted, compute the probability of observing an insect from class at that specific location using the graphic distribution of species.
			<ol>
				<li>Follow steps in section 4.2.2 and compute the posterior probability that the &#8220;unknown sound&#8221; belongs to class using the sound features and the&#160;circadian rhythms. Multiply the outcome of this step to the results from the previous step, in order to get the new posterior probability.</li>
			</ol>
			</li>
			<li>Classify the &#8220;unknown sound&#8221; to the class that has the highest new posterior probability.</li>
		</ol>
		</li>
	</ol>
	</li>
	<li>A General Framework for Adding Features
	<ol>
		<li>Consider the Bayesian classifier that uses just sound features as the primary classifier. Follow the steps below to add new features to the classifier.
		<ol>
			<li>In the training phase, learn the class-conditioned density functions of the new feature.</li>
			<li>In the classification phase, given the new feature of the &#8220;unknown sound&#8221;, compute the probability of observing the feature in class using the density functions learned in the previous step. Multiply the new probability to the previous posterior probability corresponding to the &#8220;unknown sound&#8221; belonging to class which were computed based on just the odd features, to obtain the new posterior probability. Classify the unknown object to the class that has highest new posterior probability.</li>
		</ol>
		</li>
	</ol>
	</li>
</ol>

<ol>
	<li>Kahn, M. C., Celestin, W., Offenhauser, W. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Recording+of+sounds+produced+by+certain+disease-carrying+mosquitoes.">Recording of sounds produced by certain disease-carrying mosquitoes.</a> Science. 101 (2622), 335-336 (1945).</li><li>Reed, S. C., Williams, C. M., Chadwick, L. E. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Frequency+of+wing-beat+as+a+character+for+separating+species+races+and+geographic+varieties+of+Drosophila.">Frequency of wing-beat as a character for separating species races and geographic varieties of Drosophila.</a> Genetics. 27 (3), 349 (1942).</li><li>Belton, P., Costello, R. A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Flight+sounds+of+the+females+of+some+mosquitoes+of+Western+Canada.">Flight sounds of the females of some mosquitoes of Western Canada.</a> Entomologia experimentalis et applicata. 26 (1), 105-114 (1979).</li><li>Mankin, R. W., Machan, R., Field Jones, R. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=testing+of+a+prototype+acoustic+device+for+detection+of+Mediterranean+fruit+flies+flying+into+a+trap.">testing of a prototype acoustic device for detection of Mediterranean fruit flies flying into a trap.</a> Proc. 7th Int Symp Fruit Flies of Economic Importance. , 10-15 (2006).</li><li>Raman, D. R., Gerhardt, R. R., Wilkerson, J. B. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Detecting+insect+flight+sounds+in+the+field+Implications+for+acoustical+counting+of+mosquitoes.">Detecting insect flight sounds in the field Implications for acoustical counting of mosquitoes.</a> Transactions of the ASABE. 50 (4), 1481 .</li><li>Moore, A., Miller, J. R., Tabashnik, B. E., Gage, S. H. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Automated+identification+of+flying+insects+by+analysis+of+wingbeat+frequencies.">Automated identification of flying insects by analysis of wingbeat frequencies.</a> Journal of economic entomology. 79 (6), 1703-1706 (1986).</li><li>Moore, A., Miller, R. H. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Automated+identification+of+optically+sensed+aphid+(Homoptera:+Aphidae)+wingbeat+waveforms.">Automated identification of optically sensed aphid (Homoptera: Aphidae) wingbeat waveforms.</a> Annals of the Entomological Society of America. 95 (1), 1-8 (2002).</li><li>Li, Z., Zhou, Z., Shen, Z., Yao, Q. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Automated+identification+of+mosquito+(diptera:+Culicidae)+wingbeat+waveform+by+artificial+neural+network.">Automated identification of mosquito (diptera: Culicidae) wingbeat waveform by artificial neural network.</a> Artificial Intelligence Applications and Innovations. , 483-489 (2005).</li><li>Moore, A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Artificial+neural+network+trained+to+identify+mosquitoes+in+flight.">Artificial neural network trained to identify mosquitoes in flight.</a> Journal of insect behavior. 4 (3), 391-396 (1991).</li><li>Halevy, A., Norvig, P., Pereira, F. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=The+unreasonable+effectiveness+of+data.">The unreasonable effectiveness of data.</a> Intelligent Systems, IEEE. 24 (2), 8-12 (2009).</li><li>Banko, M., Brill, E. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Mitigating+the+paucity-of-data+problem+exploring+the+effect+of+training+corpus+size+on+classifier+performance+for+natural+language+processing.">Mitigating the paucity-of-data problem exploring the effect of training corpus size on classifier performance for natural language processing.</a> , 1-5 (2001).</li><li>Shotton, J., et al. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Real-time+human+pose+recognition+in+parts+from+single+depth+images.">Real-time human pose recognition in parts from single depth images.</a> Communications of the ACM. 56 (1), 116-124 (2013).</li><li>Ilyas, P. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Classifying+insects+on+the+fly.">Classifying insects on the fly.</a> Ecological Informatics. , (2013).</li><li>Mack, Y. P., Rosenblatt, M. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Multivariate+k-nearest+neighbor+density+estimates.">Multivariate k-nearest neighbor density estimates.</a> Journal of Multivariate Analysis. 9 (1), 1-15 (1979).</li><li>Benedict, M., Robinson, A. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=The+first+releases+of+transgenic+mosquitoes+an+argument+for+the+sterile+insect+technique.">The first releases of transgenic mosquitoes an argument for the sterile insect technique.</a> TRENDS in Parasitology. 19 (8), 349-355 (2003).</li><li>. Supporting Materials Available from: <a target="_blank" href="https://sites.google.com/site/insectclassification">https://sites.google.com/site/insectclassification</a></li><li>Chen, Y., Why, A., Batista, G., Mafra-Neto, A., Keogh, E. <a target="_blank" href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed&cmd=Search&doptcmdl=Citation&defaultField=Title+Word&term=Flying+Insect+Classification+with+Inexpensive+Sensors.">Flying Insect Classification with Inexpensive Sensors.</a> Journal of Insect. 27 (5), 657-677 (2014).</li></ol>

The authors declare that they have no competing financial interests.

The sensor/classification framework described here allows the inexpensive and scalable classification of flying insects. The accuracies achievable by the system are good enough to allow the development of commercial products and could be a useful tool in entomological research.
The ability to use inexpensive, noninvasive sensors to accurately and automatically classify flying insects would have significant implications for entomological research. For example, by deploying the system in the field to count and classify insect vectors, the system can provide real-time counts of the target species, producing real-time information that can be used to plan intervention/suppression programs to combat malaria. Moreover, the system can automatically separate insects by sex, and thus it can be used to free entomologists working on the Sterile Insect Technique15 from the tedious and time-consuming task of manually sexing the insects.
In using this system, the most critical step is to properly set up the sensor for data collection. If the laser and the photo array are not properly aligned, the data will be very noisy. After the insects are placed in the cage, the photo array should always be fine-tuned using the magnets on the outside of the cage. Note that flashing lights, camera flashes and vibrations near the cages will introduce noise to the data. Therefore, to obtain clean data, place the cage in a dark room, and wherever necessary, place dry towels under the cages in order to diminish the level of vibration.
The classifier presented in this work used just two additional features. However, there may be dozens of additional features that could help improve the classification performance. As the potential features are domain and application specific, the users could choose features based on their specific need or applications. The general framework of the classifier allows users to easily add features to the classifier to improve the classification performance.
To encourage the adoption and extension of our ideas, we are making all code, data, and sensor schematics freely available at the UCR Computational Entomology Page16-17. Moreover, within the limits of our budget, we will continue our practice of giving a complete system (as shown in&#160;Figure 1) to any research entomologist who requests one.

The idea of automatically classifying insects using the incidental sound of their flight dates back to the earliest days of computers and commercially available audio recording equipment1. However, little progress has been made on this problem in the intervening decades. The lack of progress in this pursuit can be attributed to several related factors.
First, the lack of effective sensors has made data collection difficult. Most efforts to collect data have used acoustic microphones2-5. Such devices are extremely sensitive to wind noise and ambient noise in the environment, resulting in very sparse and low-quality data.
Second, compounding these data quality issues is the fact that many researchers have attempted to learn very complicated classification models, especially neural networks6-8. Attempting to learn complicated classification models, with a mere tens of examples, is a recipe for over-fitting.
Third, the difficultly of obtaining data has meant that many researchers have attempted to build classification models with very limited data, as few as 300 instances9 or less. However, it is known that for building accurate classification models, more data is better10-13.
This work addresses all three issues. Optical (rather than acoustic) sensors can be used to record the &#8220;sound&#8221; of insect flight from meters away, with complete invariance to wind noise and ambient sounds. These sensors have allowed the recording of millions of labeled training instances, far more data than all previous efforts combined, and thus help avoid the over-fitting that has plagued previous research efforts. A principled method is shown below that allows the incorporation of additional information into the classification model. This additional information can be as quotidian and as easy-to-obtain as the time-of-day, yet still produce significant gains in accuracy of the model. Finally, it is demonstrated that the enormous amounts of data we collected allow us to take advantage of &#8220;The unreasonable effectiveness of data&#8221;10 to produce simple, accurate and robust classifiers.
In summary, flying insect classification has moved beyond the dubious claims created in the research lab and is now ready for real-world deployment. The sensors and software presented in this work will provide researchers worldwide robust tools to accelerate their research.

<table border="0" cellpadding="0" cellspacing="0" width="1293">
	<tbody>
		<tr height="21">
			<td height="21" style="height:21px;width:328px;">Name of Material/ Equipment</td>
			<td style="width:237px;">Company</td>
			<td style="width:257px;">Catalog Number</td>
			<td style="width:471px;">Comments/Description</td>
		</tr>
		<tr height="21">
			<td height="21" style="height:21px;width:328px;">Audio Recorder: ICD-PX312</td>
			<td style="width:237px;">Sony</td>
			<td style="width:257px;">4-267-065-11(2)</td>
			<td>With a 8 GB microSD extra memory</td>
		</tr>
		<tr height="21">
			<td height="21" style="height:21px;width:328px;">Insectary</td>
			<td style="width:237px;">Lee&#39;s Aquarium &#38; Pet Products.</td>
			<td style="width:257px;">20088 HerpHaven&#174;, Large Rectangle</td>
			<td style="width:471px;">14 1/2&#34; Long x 8 3/4&#34; Wide x 9 3/4&#34; high. Modified to house insects.&#160;</td>
		</tr>
		<tr height="42">
			<td height="42" style="height:42px;width:328px;">Laser Line Generator, 650nm (red)</td>
			<td style="width:237px;">Apinex (www.apinex.com)</td>
			<td style="width:257px;">LN60-650</td>
			<td style="width:471px;">5mW. This is a low powered laser, similar to a teachers lasers pointer</td>
		</tr>
		<tr height="63">
			<td height="63" style="height:63px;width:328px;">Photodiode Array</td>
			<td style="width:237px;">VISHAY SEMICONDUCTOR&#160; TEFD4300&#160; PIN PHOTODIODE, 650NM, 20DEG, T1</td>
			<td style="width:257px;">TEFD4300</td>
			<td style="width:471px;">We made a custom array of 15 of these Photodiodes wired in parallel</td>
		</tr>
		<tr height="63">
			<td height="63" style="height:63px;">Analogue to Digital Convertor Integrated Circuit&#160;</td>
			<td style="width:237px;">Custom made in our lab</td>
			<td style="width:257px;"></td>
			<td style="width:471px;">We made this item ourselves, but an easily available commercial product, Gino PCF8591 AD/DA Converter, provides the same functionality.&#160;</td>
		</tr>
	</tbody>
</table>

flying insect detection and classification with inexpensive sensors

An inexpensive, noninvasive system that could accurately classify flying insects would have important implications for entomological research, and allow for the development of many useful applications in vector and pest control for both medical and agricultural entomology. Given this, the last sixty years have seen many research efforts devoted to this task. To date, however, none of this research has had a lasting impact. In this work, we show that pseudo-acoustic optical sensors can produce superior data; that additional features, both intrinsic and extrinsic to the insect&#8217;s flight behavior, can be exploited to improve insect classification; that a Bayesian classification approach allows to efficiently learn classification models that are very robust to over-fitting, and a general classification framework allows to easily incorporate arbitrary number of features. We demonstrate the findings with large-scale experiments that dwarf all previous works combined, as measured by the number of insects and the number of species considered.

Two experiments are presented here. For both experiments, the data used were randomly sampled from a dataset that contains over 100,000 objects.
The first experiment shows the ability of the proposed classifier to accurately classify different species/sexes of insects. As the classification accuracy depends on the insects to be classified, a single absolute value for classification accuracy will not give the reader a good intuition about the performance of the system. Instead, rather than reporting the classifier&#8217;s accuracy on a fixed set of insects, the classifier was applied to datasets with an incrementally increasing number of species, and therefore increasing classification difficulty.
The dataset began with just 2&#160;species of insects; then at each step, one more species (or a single sex of a sexually dimorphic species) was added and the classifier was used to classify the increased number of species (the new dataset). A total of ten classes of insects (different sexes from the same species counting as different classes) were considered, with 5,000 exemplars in each class.
The classifier used both insect-sound (frequency spectrum) and time-of-intercept for classification. Table 1 shows the classification accuracy measured at each step and the relevant class added at that step.
According to Table 1, the classifier achieves more than 96% accuracy when classifying no more than 5&#160;species of insects, significantly higher than the default rate of 20% accuracy. Even when the number of classes considered increases to 10, the classification accuracy is never lower than 79%, again significantly higher than the default rate of 10%. Note that the ten classes are not easy to separate, even by human inspection. Among the ten species, eight of them are mosquitoes, with six of them being from the same genus.
The second experiment is to show how accurately the system can sex flying insects, specifically, to distinguish male Ae. aegypti mosquitoes from females. For the first part of the experiment, assume that the &#8220;misclassification cost&#8221; of misclassifying males as females is the same as the cost of misclassifying females as males. With this assumption, the classification results are shown in Table 2.I. The classification accuracy to sex Ae. aegypti is about 99.4%.
For the second part of experiment, assume the cost is not asymmetric, i.e., misclassification of females as males is much more costly than the reverse. With this assumption, the decision threshold of the classifier was changed to reduce the number of high-cost misclassifications. With the threshold properly adjusted, the classification results in Table 2.II were achieved. Of 2,000 insects in the experiment, twenty-two males, and zero females were misclassified.
<img alt="Figure 1" fo:content-width="5in" src="/files/ftp_upload/52111/52111fig1highres.jpg" width="500" /> 
Figure 1. (I) One of the experimental cages used to gather the data. (II) A logical version of the sensor setup with the components annotated. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/52111/52111fig1highres.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 2" src="/files/ftp_upload/52111/52111fig2highres.jpg" /> 
Figure 2. (I) An example of a 1 sec audio clip containing a &#8220;flying insect sound&#8221; generated by the sensor. The sound was produced by a female Cx. stigmatosoma. The insect sound is highlighted in red/bold. (II) The &#8220;insect sound&#8221;, which has been cleaned and saved into a 1-sec&#160;long audio clip by centering the insect signal and padding with 0s elsewhere. (III) The frequency spectrum of the insect sound, obtained using Discrete Fourier Transform. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/52111/52111fig2highres.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 3" src="/files/ftp_upload/52111/52111fig3highres.jpg" /> 
Figure 3.&#160;A Bayesian network that uses a single feature for classification.
<img alt="Figure 4" src="/files/ftp_upload/52111/52111fig4highres.jpg" /> 
Figure 4.&#160;A Bayesian network that uses two independent features for classification
<img alt="Figure 5" src="/files/ftp_upload/52111/52111fig5highres.jpg" /> 
Figure 5.&#160;The&#160;circadian rhythms of Cx. stigmatosoma (female), Cx. tarsalis (male), and Ae. aegypti (female), learned based on observations generated by the sensor that were collected over 1&#160;month duration. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/52111/52111fig5highres.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 6" src="/files/ftp_upload/52111/52111fig6highres.jpg" /> 
Figure 6.&#160;A Bayesian network that uses three independent features for classification.
<img alt="Figure 7" src="/files/ftp_upload/52111/52111fig7highres.jpg" /> 
Figure 7.&#160;The assumptions of the geographic distributions of each insect species and sensor locations in the simulation to demonstrate the effectiveness of using location-of-intercept feature in classification. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/52111/52111fig7highres.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 8" src="/files/ftp_upload/52111/52111fig8highres.jpg" /> 
Figure 8.&#160;The general Bayesian network that uses n features for classification, where n is a positive integer. <a href="https://www-jove-com.remotexs.ntu.edu.sg/files/ftp_upload/52111/52111fig8highres.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<table border="0" cellpadding="0" cellspacing="0">
	<tbody>
		<tr>
			<td>Step</td>
			<td>Species Added&#160;</td>
			<td>Classification Accuracy</td>
			<td></td>
			<td>Step</td>
			<td>Species Added&#160;</td>
			<td>Classification Accuracy</td>
		</tr>
		<tr>
			<td>1</td>
			<td>Ae. aegypti &#9794;</td>
			<td>N/A</td>
			<td></td>
			<td>6</td>
			<td>Cx. quinquefasciatus &#9794;</td>
			<td>92.69%</td>
		</tr>
		<tr>
			<td>2</td>
			<td>Musca domestica</td>
			<td>98.99%</td>
			<td></td>
			<td>7</td>
			<td>Cx. stigmatosoma &#9792;</td>
			<td>89.66%</td>
		</tr>
		<tr>
			<td>3</td>
			<td>Ae. aegypti &#9792;</td>
			<td>98.27%</td>
			<td></td>
			<td>8</td>
			<td>Cx. tarsalis &#9794;</td>
			<td>83.54%</td>
		</tr>
		<tr>
			<td>4</td>
			<td>Cx. stigmatosoma &#9794;</td>
			<td>97.31%</td>
			<td></td>
			<td>9</td>
			<td>Cx. quinquefasciatus&#160;&#9792;</td>
			<td>81.04%</td>
		</tr>
		<tr>
			<td>5</td>
			<td>Cx. tarsalis &#9792;</td>
			<td>96.10%</td>
			<td></td>
			<td>10</td>
			<td>Drosophila simulans</td>
			<td>79.44%</td>
		</tr>
	</tbody>
</table>
Table 1.&#160;Classification accuracy with increasing number of classes.
<table border="0" cellpadding="0" cellspacing="0">
	<tbody>
		<tr>
			<td></td>
			<td></td>
			<td colspan="2">Predicted class</td>
			<td></td>
			<td colspan="2"></td>
			<td colspan="2">Predicted class</td>
		</tr>
		<tr>
			<td colspan="2">I (Symmetric cost)</td>
			<td>female</td>
			<td>male</td>
			<td></td>
			<td colspan="2">II (Asymmetric cost)</td>
			<td>female</td>
			<td>male</td>
		</tr>
		<tr>
			<td rowspan="2">Actual class</td>
			<td>female</td>
			<td>993</td>
			<td>7</td>
			<td></td>
			<td rowspan="2">Actual class</td>
			<td>female</td>
			<td>1,000</td>
			<td>0</td>
		</tr>
		<tr>
			<td>male</td>
			<td>5</td>
			<td>995</td>
			<td></td>
			<td>male</td>
			<td>22</td>
			<td>978</td>
		</tr>
	</tbody>
</table>
Table 2. (I) The confusion matrix for sex discrimination of Ae. aegypti mosquitoes with the decision threshold for females is set at 0.5 (i.e., same cost assumption). (II) The confusion matrix of sex the same mosquitoes, with the decision threshold for females set at 0.1.

Watch this Scientific Journal Video about Flying Insect Detection and Classification with Inexpensive Sensors at JoVE.com

Flying Insect Detection and Classification with Inexpensive Sensors

We proposed a system that uses inexpensive, noninvasive pseudo-acoustic optical sensors to automatically and accurately detect, count, and classify flying insects based on their flying sound.

An inexpensive, noninvasive system that could accurately classify flying insects would have important implications for entomological research, and allow ...

flying-insect-detection-and-classification-with-inexpensive-sensors

Nanyang Technological University

Research

JoVE Journal

Bioengineering

25.0K Views.  University of California, Riverside. We proposed a system that uses inexpensive, noninvasive pseudo-acoustic optical sensors to automatically and accurately detect, count, and classify flying insects based on their flying sound.

Video: Flying Insect Detection and Classification with Inexpensive Sensors

Fluorescence detection methods for microfluidic droplet platforms

The development of microfluidic platforms for performing chemistry and biology has in large part been driven by a range of potential benefits that accompany system miniaturisation. Advantages include the ability to efficiently process nano- to femoto- liter volumes of sample, facile integration of functional components, an intrinsic predisposition towards large-scale multiplexing, enhanced analytical throughput, improved control and reduced instrumental footprints.1
In recent years much interest has focussed on the development of droplet-based (or segmented flow) microfluidic systems and their potential as platforms in high-throughput experimentation.2-4 Here water-in-oil emulsions are made to spontaneously form in microfluidic channels as a result of capillary instabilities between the two immiscible phases. Importantly, microdroplets of precisely defined volumes and compositions can be generated at frequencies of several kHz. Furthermore, by encapsulating reagents of interest within isolated compartments separated by a continuous immiscible phase, both sample cross-talk and dispersion (diffusion- and Taylor-based) can be eliminated, which leads to minimal cross-contamination and the ability to time analytical processes with great accuracy. Additionally, since there is no contact between the contents of the droplets and the channel walls (which are wetted by the continuous phase) absorption and loss of reagents on the channel walls is prevented.
Once droplets of this kind have been generated and processed, it is necessary to extract the required analytical information. In this respect the detection method of choice should be rapid, provide high-sensitivity and low limits of detection, be applicable to a range of molecular species, be non-destructive and be able to be integrated with microfluidic devices in a facile manner. To address this need we have developed a suite of experimental tools and protocols that enable the extraction of large amounts of photophysical information from small-volume environments, and are applicable to the analysis of a wide range of physical, chemical and biological parameters. Herein two examples of these methods are presented and applied to the detection of single cells and the mapping of mixing processes inside picoliter-volume droplets. We report the entire experimental process including microfluidic chip fabrication, the optical setup and the process of droplet generation and detection.

Droplet-based microfluidic platforms are promising candidates for high throughput experimentation since they are able to generate picoliter, self-compartmentalized vessels inexpensively at kHz rates. Through integration with fast, sensitive and high resolution fluorescence spectroscopic methods, the large amounts of information generated within these systems can be efficiently extracted, harnessed and utilized.

The development of microfluidic platforms for performing chemistry and biology has in large part been driven by a range of potential benefits that ...

Detection and Isolation of Circulating Melanoma Cells using Photoacoustic Flowmetry

Circulating tumor cells (CTCs) are those cells that have separated from a macroscopic tumor and spread through the blood and lymph systems to seed secondary tumors1,2,3. CTCs are indicators of metastatic disease and their detection in blood samples may be used to diagnose cancer and monitor a patient&prime;s response to therapy. Since CTCs are rare, comprising about one tumor cell among billions of normal blood cells in advanced cancer patients, their detection and enumeration is a difficult task. We exploit the presence of pigment in most melanoma cells to generate photoacoustic, or laser induced ultrasonic waves in a custom flow cytometer for detection of circulating melanoma cells (CMCs)4,5. This process entails separating a whole blood sample using centrifugation and obtaining the white blood cell layer. If present in whole blood, CMCs will separate with the white blood cells due to similar density. These cells are resuspended in phosphate buffered saline (PBS) and introduced into the flowmeter. Rather than a continuous flow of the blood cell suspension, we induced two phase flow in order to capture these cells for further study. In two phase flow, two immiscible liquids in a microfluidic system meet at a junction and form alternating slugs of liquid6,7. PBS suspended white blood cells and air form microliter slugs that are sequentially irradiated with laser light. The addition of a surfactant to the liquid phase allows uniform slug formation and the user can create different sized slugs by altering the flow rates of the two phases. Slugs of air and slugs of PBS with white blood cells contain no light absorbers and hence, do not produce photoacoustic waves. However, slugs of white blood cells that contain even single CMCs absorb laser light and produce high frequency acoustic waves. These slugs that generate photoacoustic waves are sequestered and collected for cytochemical staining for verification of CMCs.

We have developed a flow cytometer using laser induced ultrasound to detect circulating melanoma cells as an early indicator of metastatic disease.

Circulating tumor cells (CTCs) are those cells that have separated from a macroscopic tumor and spread through the blood and lymph systems to seed ...

Bacterial Detection &amp; Identification Using Electrochemical Sensors

Electrochemical sensors are widely used for rapid and accurate measurement of blood glucose and can be adapted for detection of a wide variety of analytes. Electrochemical sensors operate by transducing a biological recognition event into a useful electrical signal. Signal transduction occurs by coupling the activity of a redox enzyme to an amperometric electrode. Sensor specificity is either an inherent characteristic of the enzyme, glucose oxidase in the case of a glucose sensor, or a product of linkage between the enzyme and an antibody or probe.
Here, we describe an electrochemical sensor assay method to directly detect and identify bacteria. In every case, the probes described here are DNA oligonucleotides. This method is based on sandwich hybridization of capture and detector probes with target ribosomal RNA (rRNA). The capture probe is anchored to the sensor surface, while the detector probe is linked to horseradish peroxidase (HRP). When a substrate such as 3,3',5,5'-tetramethylbenzidine (TMB) is added to an electrode with capture-target-detector complexes bound to its surface, the substrate is oxidized by HRP and reduced by the working electrode. This redox cycle results in shuttling of electrons by the substrate from the electrode to HRP, producing current flow in the electrode.

Bacterial Detection & Identification Using Electrochemical Sensors

We describe an electrochemical sensor assay method for rapid bacterial detection and identification. The assay involves a sensor array functionalized with DNA oligonucleotide capture probes for ribosomal RNA (rRNA) species-specific sequences. Sandwich hybridization of target rRNA with the capture probe and a horseradish peroxidase-linked DNA oligonucleotide detector probe produces a measurable amperometric current.

Electrochemical  sensors are widely used for rapid and accurate measurement of blood glucose and  can be adapted for detection of a wide variety of ...

Biosensor for Detection of Antibiotic Resistant Staphylococcus Bacteria

A structurally transformed lytic bacteriophage having a broad host range of Staphylococcus aureus strains and a penicillin-binding protein (PBP 2a) antibody conjugated latex beads have been utilized to create a biosensor designed for discrimination of methicillin resistant (MRSA) and sensitive (MSSA) S. aureus species 1,2. The lytic phages have been converted into phage spheroids by contact with water-chloroform interface. Phage spheroid monolayers have been moved onto a biosensor surface by Langmuir-Blodgett (LB) technique 3. The created biosensors have been examined by a quartz crystal microbalance with dissipation tracking (QCM-D) to evaluate bacteria-phage interactions. Bacteria-spheroid interactions led to reduced resonance frequency and a rise in dissipation energy for both MRSA and MSSA strains. After the bacterial binding, these sensors have been further exposed to the penicillin-binding protein antibody latex beads. Sensors analyzed with MRSA responded to PBP 2a antibody beads; although sensors inspected with MSSA gave no response. This experimental distinction determines an unambiguous discrimination between methicillin resistant and sensitive S. aureus strains. Equally bound and unbound bacteriophages suppress bacterial growth on surfaces and in water suspensions. Once lytic phages are changed into spheroids, they retain their strong lytic activity and show high bacterial capture capability. The phage and phage spheroids can be utilized for testing and sterilization of antibiotic resistant microorganisms. Other applications may include use in bacteriophage therapy and antimicrobial surfaces.

Lytic phage biosensors and antibody beads are able to discriminate between methicillin resistant (MRSA) and sensitive staphylococcus bacteria. The phages were immobilized by a Langmuir-Blodgett method onto a surface of a quartz crystal microbalance sensor and worked as broad range staphylococcus probes. Antibody beads recognize MRSA.

A structurally transformed lytic bacteriophage having a broad host range of Staphylococcus aureus strains and a penicillin-binding protein (PBP 2a) ...

Glutamine Flux Imaging Using Genetically Encoded Sensors

Genetically encoded sensors allow real-time monitoring of biological molecules at a subcellular resolution. A tremendous variety of such sensors for biological molecules became available in the past 15 years, some of which became indispensable tools that are used routinely in many laboratories.
One of the exciting applications of genetically encoded sensors is the use of these sensors in investigating cellular transport processes. Properties of transporters such as kinetics and substrate specificities can be investigated at a cellular level, providing possibilities for cell-type specific analyses of transport activities. In this article, we will demonstrate how transporter dynamics can be observed using genetically encoded glutamine sensor as an example. Experimental design, technical details of the experimental settings, and considerations for post-experimental analyses will be discussed.

This article will demonstrate how to monitor glutamine dynamics in live cells using FRET. Genetically encoded sensors allow real-time monitoring of biological molecules at a subcellular resolution. Experimental design, technical details of the experimental settings, and considerations for post-experimental analyses will be discussed for genetically encoded glutamine sensors.

Genetically encoded sensors allow real-time monitoring of biological molecules at a subcellular resolution. A tremendous variety of such sensors for ...

A Microfluidic-based Electrochemical Biochip for Label-free DNA Hybridization Analysis

Miniaturization of analytical benchtop procedures into the micro-scale provides significant advantages in regards to reaction time, cost, and integration of pre-processing steps. Utilizing these devices towards the analysis of DNA hybridization events is important because it offers a technology for real time assessment of biomarkers at the point-of-care for various diseases. However, when the device footprint decreases the dominance of various physical phenomena increases. These phenomena influence the fabrication precision and operation reliability of the device. Therefore, there is a great need to accurately fabricate and operate these devices in a reproducible manner in order to improve the overall performance. Here, we describe the protocols and the methods used for the fabrication and the operation of a microfluidic-based electrochemical biochip for accurate analysis of DNA hybridization events. The biochip is composed of two parts: a microfluidic chip with three parallel micro-channels made of polydimethylsiloxane (PDMS), and a 3 x 3 arrayed electrochemical micro-chip. The DNA hybridization events are detected using electrochemical impedance spectroscopy (EIS) analysis. The EIS analysis enables monitoring variations of the properties of the electrochemical system that are dominant at these length scales. With the ability to monitor changes of both charge transfer and diffusional resistance with the biosensor, we demonstrate the selectivity to complementary ssDNA targets, a calculated detection limit of 3.8 nM, and a 13% cross-reactivity with other non-complementary ssDNA following 20 min&#160;of incubation. This methodology can improve the performance of miniaturized devices by elucidating on the behavior of diffusion at the micro-scale regime and by enabling the study of DNA hybridization events.

We present a microfluidic-based electrochemical biochip for DNA hybridization detection. Following ssDNA probe functionalization, the specificity, sensitivity, and detection limit are studied with complementary and non-complementary ssDNA targets. Results illustrate the influence of the DNA hybridization events on the electrochemical system, with a detection limit of 3.8 nM.

Miniaturization of analytical benchtop procedures into the micro-scale provides significant advantages in regards to reaction time, cost, and integration ...

Detection of Exosomal Biomarker by Electric Field-induced Release and Measurement (EFIRM)

Exosomes are microvesicular structures that play a mediating role in intercellular communication. It is of interest to study the internal cargo of exosomes to determine if they carry disease discriminatory biomarkers. For performing exosomal analysis, it is necessary to develop a method for extracting and analyzing exosomes from target biofluids without damaging the internal content.
Electric field-induced release and measurement (EFIRM) is a method for specifically extracting exosomes from biofluids, unloading their cargo, and testing their internal RNA/protein content. Using an anti-human CD63 specific antibody magnetic microparticle, exosomes are first precipitated from biofluids. Following extraction, low-voltage electric cyclic square waves (CSW) are applied to disrupt the vesicular membrane and cause cargo unloading. The content of the exosome is hybridized to DNA primers or antibodies immobilized on an electrode surface for quantification of molecular content.
The EFIRM method is advantageous for extraction of exosomes and unloading cargo for analysis without lysis buffer. This method is capable of performing specific detection of both RNA and protein biomarker targets in the exosome. EFIRM extracts exosomes specifically based on their surface markers as opposed to size-based techniques.
Transmission electron microscopy (TEM) and assay demonstrate the functionality of the method for exosome capture and analysis. The EFIRM method was applied to exosomal analysis of 9 mice injected with human lung cancer H640 cells (a cell line transfected to express the exosome marker human CD63-GFP) in order to test their exosome profile against 11 mice receiving saline controls. Elevated levels of exosomal biomarkers (reference gene GAPDH and protein surface marker human CD63-GFP) were found for the H640 injected mice in both serum and saliva samples. Furthermore, saliva and serum samples were demonstrated to have linearity (R = 0.79). These results are suggestive for the viability of salivary exosome biomarkers for detection of distal diseases.

Detection of Exosomal Biomarker by Electric Field-induced Release and Measurement EFIRM

Exosomes are microvesicular structures found within biofluids that potentially carry important disease discriminatory biomarkers. Here, a novel method is used to specifically extract exosomes and rapidly test the exosomal cargo for both RNA/protein targets following the disruption of exosomes using non-uniform electric cyclic square waves.

Exosomes are microvesicular structures that play a mediating role in intercellular communication. It is of interest to study the internal cargo of ...

Label-free Single Molecule Detection Using Microtoroid Optical Resonators

Detecting small concentrations of molecules down to the single molecule limit has impact on areas such as early detection of disease, and fundamental studies on the behavior of molecules. Single molecule detection techniques commonly utilize labels such as fluorescent tags or quantum dots, however, labels are not always available, increase cost and complexity, and can perturb the events being studied. Optical resonators have emerged as a promising means to detect single molecules without the use of labels. Currently the smallest particle detected by a non-plasmonically-enhanced bare optical resonator system in solution is a 25 nm polystyrene sphere1. We have developed a technique known as Frequency Locking Optical Whispering Evanescent Resonator (FLOWER) that can surpass this limit and achieve label-free single molecule detection in aqueous solution2. As signal strength scales with particle volume, our work represents a &#62; 100x improvement in the signal to noise ratio (SNR) over the current state of the art. Here the procedures behind FLOWER are presented in an effort to increase its usage in the field.

We have developed a label-free biosensing system based on optical resonator technology known as Frequency Locking Optical Whispering Evanescent Resonator (FLOWER) that is capable of detecting single molecules in solution. Here the procedures behind this work are described and presented.

Detecting small concentrations of molecules down to the single molecule limit has impact on areas such as early detection of disease, and fundamental ...

Sensitive Detection of Proteopathic Seeding Activity with FRET Flow Cytometry

Increasing evidence supports transcellular propagation of toxic protein aggregates, or proteopathic seeds, as a mechanism for the initiation and progression of pathology in several neurodegenerative diseases, including Alzheimer's disease and the related tauopathies. The potentially critical role of tau seeds in disease progression strongly supports the need for a sensitive assay that readily detects seeding activity in biological samples. 
By combining the specificity of fluorescence resonance energy transfer (FRET), the sensitivity of flow cytometry, and the stability of a monoclonal cell line, an ultra-sensitive seeding assay has been engineered and is compatible with seed detection from recombinant or biological samples, including human and mouse brain homogenates. The assay employs monoclonal HEK 293T cells that stably express the aggregation-prone repeat domain (RD) of tau harboring the disease-associated P301S mutation fused to either CFP or YFP, which produce a FRET signal upon protein aggregation. The uptake of proteopathic tau seeds (but not other proteins) into the biosensor cells stimulates aggregation of RD-CFP and RD-YFP, and flow cytometry sensitively and quantitatively monitors this aggregation-induced FRET. The assay detects femtomolar concentrations (monomer equivalent) of recombinant tau seeds, has a dynamic range spanning three orders of magnitude, and is compatible with brain homogenates from tauopathy transgenic mice and human tauopathy subjects. With slight modifications, the assay can also detect seeding activity of other proteopathic seeds, such as &#945;-synuclein, and is also compatible with primary neuronal cultures. The ease, sensitivity, and broad applicability of FRET flow cytometry makes it useful to study a wide range of protein aggregation disorders.

Cell-to-cell transfer of protein aggregates, or proteopathic seeds, may underlie the progression of pathology in neurodegenerative diseases. Here, a novel FRET flow cytometry assay is described that enables specific and sensitive detection of seeding activity from recombinant or biological samples.

Increasing evidence supports transcellular propagation of toxic protein aggregates, or proteopathic seeds, as a mechanism for the initiation and ...

Fast Enzymatic Processing of Proteins for MS Detection with a Flow-through Microreactor

The vast majority of mass spectrometry (MS)-based protein analysis methods involve an enzymatic digestion step prior to detection, typically with trypsin. This step is necessary for the generation of small molecular weight peptides, generally with MW &#60; 3,000-4,000 Da, that fall within the effective scan range of mass spectrometry instrumentation. Conventional protocols involve O/N enzymatic digestion at 37 &#186;C. Recent advances have led to the development of a variety of strategies, typically involving the use of a microreactor with immobilized enzymes or of a range of complementary physical processes that reduce the time necessary for proteolytic digestion to a few minutes (e.g., microwave or high-pressure). In this work, we describe a simple and cost-effective approach that can be implemented in any laboratory for achieving fast enzymatic digestion of a protein. The protein (or protein mixture) is adsorbed on C18-bonded reversed-phase high performance liquid chromatography (HPLC) silica particles preloaded in a capillary column, and trypsin in aqueous buffer is infused over the particles for a short period of time. To enable on-line MS detection, the tryptic peptides are eluted with a solvent system with increased organic content directly in the MS ion source. This approach avoids the use of high-priced immobilized enzyme particles and does not necessitate any aid for completing the process. Protein digestion and complete sample analysis can be accomplished in less than ~3 min and ~30 min, respectively.

A quick protocol for proteolytic digestion with an in-house built flow-through tryptic microreactor coupled to an electrospray ionization (ESI) mass spectrometer is presented. The fabrication of the microreactor, the experimental setup and the data acquisition process are described.

The vast majority of mass spectrometry (MS)-based protein analysis methods involve an enzymatic digestion step prior to detection, typically with trypsin. ...