Announcements - Philippe poject
During the night between Jan 28 and 29 Ivan, Tibi and Daniel obtained the first biological results with the Mosquito Scientific Instrument at Phil's lab. They used the Mosquito V3 made by Jonathan, equipped with a Joint-type transducer (the C6, see in this doc) fabricated by Francois. A new bath was used, specially designed by Ivan and Daniel for chemical activation of skinned muscle cells. The bath parts were received the same day and it was assembled by Ivan right before the experiment.
The video below shows Mosquito displacement/force data (on the right) in sync with video recording (on the left) during the experiment. The resolution of the original video is much better. Due to computer memory problems the original video is 30 frames/second, Phil's CCD camera can go up to 300 frames/sec. The acquisition rate was only 1K. Averaging was performed on the data after the acquisition.
The chemical solutions were exchanged manually by Ivan during the experiment, using a pipette to suck and inject them into the bath through a small orifice, specially designed for this purpose. In the video above we can clearly see when solutions are changed. Changing solutions generates a signal before the activation and relaxation, because the sample is perturbed by capillary forces as the solution exits and enters the small chamber. This is not so bad, compared to other systems that are used to study muscle contraction, like for example Aurora's system.
Before the experiment we had some problems with signal fluctuation, which was coming from a damaged delivery fiber. Damaged optical fibers become bend sensitive, and any mechanical vibration generates a signal. Tibi found a way to hold the fiber and to reduce the fluctuations.
See discussion about this experiment during our weekly meeting the next day
Fresh results obtained by Ivan and James at Phil's lab. You can see activation and relaxation of a small bundle of skinned muscle cells (their membrane have been digested). The activation is chemically induced. The activation and relaxation solutions were exchanged manually using a pipette. We need to work on a faster and a more quite solution exchange system. We are making a few more transducers and setting up a new Mosquito prototype. Next step is to measure these forces.
This week we made some progress at Phil's lab. I remind you that the end goal is to characterize the Mosquito and to obtain some meaningful results on animal muscle samples. The project was scheduled to end on December 02, but we accumulated some delay. I think SENSORICA made a lot of progress during the 6 months time of the project, so the delay does not apply to our overall situation, but other activities got in the way of Phil's project.
For the characterization experiments, we installed a piezo micromanipulator prototype made by Jonathan and used Tibi's LabView program to drive it. The video below shows the piezo in action during the first trials in Phil's lab.
Tibi integrated the piezo with the Mosquito program, one running on the NI DAQ and the other one running on the LabJack. This allows script-based automation of characterization experiments. The script language was created and tested.
Ivan mounted the optical fiber-based transducer on the blue micromanipulator and Tibi made some preliminary tests by pushing the transducer with the piezo under the microscope.
This piezo prototype is still very limited, its range is only approx. 40 microns. Moreover, as you can see in the video above, it takes a long time to relax when the voltage is reduced, which means that motion in both directions cannot be at the same speed. This limits the type of tests we can perform using this particular prototype, to those that are more static in nature, i.e. sensitivity, resolution... Tests like fatigue and time response require back and forth motion of the piezo, and cannot be done with this prototype.
In parallel, we advanced the biological part of the project. Ivan, Phil, James and Tibi participated in this last effort to understand why our samples could not be electrically activated/stimulated. We performed our experiments with fresh mouse diaphragm muscles. See sample preparation and solutions below.
The picture below shows our initial setup, using the bath designed by Ivan.
The red arrows in the picture below point to the graphite electrodes used to stimulate the sample. The white blob on top is silver conductive glue, attaching these electrodes to the metallic wire. See more pictures in this album. The sample sits in the middle of the bath, attached to the two hooks sitting inside of the darker circle in the center of the picture.
For the electrical stimulation we used Jame's electronic circuit, driven with a NI DAQ card, and powered by a power supply. Tibi's LabView program (still undocumented) was used to generate the electrical pulse pattern. A max of 17 Volt could be supplied to the sample. This arrangement produced results only once.
The picture below shows the front panel of the LabView program used for the Stimulation, in conjunction with James' electronic circuit.
We turned to a second electrical stimulator which could supply up to 40 Volt. This was still not enough to stimulate the sample in the same bath, but worked very well when the sample was placed in the different type of bath with very large surface graphite electrodes.
These results eliminated other hypothesis that we have made, related to the sample preparation and chemical solutions. It seams that the orientation and the strength of the electrical field generated between the electrodes is the main factor. We need to increase the surface of our electrodes and/or place them closer to the sample, and make sure that the sample sits well in between the electrodes. The voltage adjustment/optimization is another issue.
For the biological tests we will test different types of electrodes.
After our last group meeting on Tuesday 27 Nov, 2012 (meeting notes - members only) we accelerated the pace of Philippe's project. The deadline is fast-approaching and we need to allocate more resources to this project.
There are two important sets of experiments going on in parallel:
In the mean time, Tibi has installed the Mosquito Demo (based on the LabJack) in the lab to advance the biological tests. Tibi also finished a new LabView program to drive Jame's circuit for muscle cell stimulation. The entire stimulation system was successfully tested with the program and monitored on the oscilloscope. The screenshot below shows a stimulation signal on the front panel of the program. (Jame's circuit is driven by an NI_DAQ card, which is driven by the LabView program).
The goal of the day was to measure an actual contraction of a diaphragm muscle stripe collected at 10 am in the morning. Unfortunately, the cells could not survive until 6pm in the afternoon. The picture bellow shows the muscle stripe hold in place by two polymer open rings, which in turn rest on 2 hooks that are attached to the Mosquito transducer (on the left) and to a glass needle (on the right). See more pictures at the end this page.
This new solution to attach the sample to the force transducer was tested by puling on the sample and measuring the force with the Mosquito. In technical terms, this is also called measuring the passive tension (when the muscle is NOT activated), The picture below shows the passive stretch Mosquito trace.
See more pictures of the system taken today.
We first installed and tested the bath. The bath was designed mostly by Ivan and Daniel.
We also tested the electrical stimulation on a mouse heart freshly dissected today by Ivan (sitting in the middle of the bath). The heart was beating!
After, we tested the Mosquito signal under the microscope. The video below shows the Mosquito signal on the right as the transducer is pushed by a glass needle (which sits on a micrometer).
Problems to solve
See more pictures in today's album.
We have performed the first characterization tests on the Joint-type transducer.
Two joint-type transducers
Ivan and James have performed stiffness measurements. The graph below shows a linear fit of Force [nN] vs Displacement [um]. The Force is applied at the tip of the lever in a direction perpendicular to the axis of the optical fiber transducer, by attaching different per-calibrated weights. Displacement is the displacement of the tip of the lever under these weights. In this case, the lever is 22mm long. Weights were attached very close to the tip. We obtain for this particular joint-type transducer 197.35 nN/um.
NOTE: we realized that there was an error of reading on the micrometer. Ask Tibi to correct the data!!
[Experiment performed and analyzed by Tibi] In this experiment the transducer was pushed with a micrometer in 1um steps, in a direction perpendicular to its axis (pure bending mode of operation) and the voltage was recorded. 500000 samples were acquired at 126kHz and averaged for each data point on the graphs below. For all graphs the displacement needs to be multiplied by 10, i.e. 1 is 10 microns and 12 is 120 microns.
Voltage [Volt] vs Displacement [10um]
For a 22mm long lever. The resolution is well above 1 micron. Almost linear on 600 microns
Voltage [Volt] vs Displacement [10um]
For a 10mm long lever. The resolution is very close to 1 micron. Linearity on 40 microns distance.CONCLUSION
This particular joint-type transducer exhibited a strong directionality and unwanted structures. We attribute that to the poor quality of the mirror (silver coating). Transducer 4 was used in this particular experiment.
20X picture of the mirror (on the tip of the lever) of transducer 4
The graph below is recorded with the same transducer but in a different orientation (a rotation around the axis of the transducer). In this case the lever is only 5mm long. On the ascending part we can see that the resolution is below 1 micron (every next point is above the previous one). The small shoulder in the middle of the ascending part is a real feature. See next graph.
Voltage [Volt] vs Displacement [10um]
For a 5mm long lever. The resolution is under 1 micron. Linearity on 50 microns distance after the first feature.
The graph below illustrated the complexity of the signal. The orientation of the transducer was modified again (rotated on its axis). The lever in this case is 5mm (distance between point of contact and gap). As the transducer is pushed in one direction the signal diminishes (contrary to the two cases above), reaches a minimum and increases again. the structure close to the minimum is NOT noise, suggesting that this is a real structure in the spatial intensity distribution of the light that reflects from the mirror.
Voltage [Volt] vs Displacement [10um]
Need to better control silver deposition and to control the cleaving process, in order to obtain a perpendicular cut.
[Experiment performed and analyzed by Tibi] The picture below shows free (dumped) oscillations (in air) of the joint-type transducer. The oscillation period for this 22mm long lever is around 6.79ms.
The system needed to characterize the Mosquito sensor is being installed in Phil's lab. Last Wednesday Ivan prepared the working space and, with some help from Tibi, mounted the first iteration of the system used to perform mechanical characterization of the joint-type transducer.
working space in the lab, the system can be seen in the left corner
The first test will be about stiffness. We'll use calibrated weights attached to the optical fiber to measure the bending of the fiber, using a micron precision ruler. The picture below shows the ruler (vertical) and the lever of the force transducer (horizontal on the right).
In the picture below we have the entire system, in its first iteration form. It consists of a multi-axis micromanipulator (blue) to hold the force transducer, which is placed in between a microscope objective and the micron-precision ruler. We use a 8mp digital camera behind the microscope objective. The micron-precision ruler is also placed on a multi-axis micromanipulator.
The same day, Tibi worked on the LabView program. We are using the platform already developed by Tibi. A new version of the Mosquito program was installed, which contains all the modules necessary to perform the characterization, including data conditioning and analysis tools. This MosquitoFull version is going to be completely operational next week.
At the same time, Ivan and Daniel are moving full steam with the 3D modeling of different parts of the system, including the physiological bath.
Next week we are planning to make the Mosquito fully operational, by connecting all the electronics and the NI DAQ card.
Jonathan is moving on the construction of the xyz micromanipulation system. All the parts have arrived.
We are moving forward with the setup for a more consistent fabrication of the joint-type transducer. Our second fabrication device is now operational. Tibi and Ivan are in charge. We already made one transducer.
At the same time, we are taking notes and designing a semi-automated fabrication device. Daniel is in charge with 3D modeling. Yassine might also help. Moreover, Jonathan contributed with two used fiber splicers, which can be used to fabricate the constriction transducer. We hope to have one ready soon to be tested by James, along with the joint-type transducer.
The optical fiber coating process has been revised by Tibi. A few fibers have been coated on July 27th and are now ready for being assembled into transducers. See picture album.
Jonathan is finishing the design of the piezo controller. The piezos have been received, we are still waiting for the voltage converters before we can assemble everything.
Installation of our first Mosquito system has begone at Montreal Heart Institute. This is a 6 months project funded by an ENGAGE grant. During this project we'll perform a thorough characterization of our Mosquito sensor. One or two scientific publications will be produced presenting the Mosquito as a new research tool in physiology and biotech. This is the last 100m sprint that will bring our first product to the finish line, to the market. See the Mosquito's path to market.