Final Presentations

Hello Readers,

My project is coming to an end. I am starting to work on my final presentation and my final product. I won't have much to do in the lab. I have some very exciting news! I changed my question to "How does ankle torque effect walking on various terrains in lower-limb amputees?" I will gather and analyze ankle torque data from various terrains and hopefully post an update with my results. The next few weeks I will be mainly be working on my final presentation and product and meeting with all my advisors to perfect everything.

Until next time,
Krishna Patel

UPDATE

Hello all!! Sorry I haven't posted in a while. Spring break has interrupted quite a bit of my project. My advisors were on spring break last week and so I wasn't able to be in the lab or do much. But do not worry, everything will be picking up this week. Stay tuned.

Until next time,
Krishna

How it all started!

Hello readers!

This week I met with Dr. John Tester, one of the research engineers on the BiOM project. Dr. Tester was hired by NAU in June 2000 as an engineer design professor. A few years after he was hired, there was a severe cutback in research funding at the campus level. NAU was fortunate enough to get money to support technology and help generate patents and intellectual property(IP). NAU received a stimulus package. Dr. Tester was the only candidate for product development available at NAU at that time. He was offered $500,000 to build a lab that could assist other labs design products for commercial use. He accepted the offer and took the money to buy key machinery. Dr. Nishikawa had been trying to get the BiOM project up and running so she contacted Dr. Tester. Dr. Tester didn't have time to take up a research project, so he declined her offer at first. But, Dr. Nishikawa was persistent and eventually Dr. Tester gave in. So Dr. Tester and Dr. Nishikawa met and hashed out all their ideas. Since Dr. nishikawa was a biologist, they quickly came to realize that biologists and engineers do not speak the same language nor do they have the same motives for conducting research. The proposal advocated for a new product developed in tandem with a commercial company, iWalk, inc. Dr. Tester was the lead on research and Dr. Nishikawa was the Co- Principal Investigator. iWalk, located in Boston, already had an existing device so one of Dr. Tester’s student, John Dyer, started preliminary work to control the mechanical device. He transferred his work to Jeremy perhaps some more background in him.. who only had to develop software because iWalk had the device ready. The university wanted to make a product for commercial industries. Currently, Dr. Nishikawa is looking to hire a biomedical engineer with a background in gait cycle analysis. Dr Tester is going on his sabatical this coming fall, which happens every six years for professors and he wants to make sure his graduate students can handle the project without him for a whole year.

Until next time,
Krishna Patel

Back to the Basics!

Hello readers!

This week I met with Uzma, who is a PhD candidate in the lab. Muscles are activated depending on what we are doing; that is, only muscles that are needed for a task as on. Muscles are not just on and off, they have various levels of activation, like pressing down on the accelerator in a car. It is very important that the correct activations are used in the prosthesis control program, so as to feel comfortable and safe for the user. This is what Uzma spends most of her time achieving. More activation is required for lifting heavy objects (since more work needs to be done to move the objects certain distance) and vise versa.  In the past, the lab originally just had the muscle 100% on or completely off. This was OK for a first approximation to test out the software capability, but now they are ready to fine tune the activation levels.

What makes this fine tuning procedure difficult is that most of what we understand about muscles properties come from maximum stimulation (100% activation) experiments. In fact, the BiOM has torque equations based off of these maximal stimulations but they don't always match up to how much activation is actually needed. We need to make sure that the equations match the various levels of activation required during various activities.

Uzma works with real mouse muscles at a lower activation to understand their properties. She removes muscles from mice, and puts them on a “Force Lever”, which can change the length of the muscle, and is a representation of the muscle still in the mouse. It measures how much force the muscle produces when activated to a certain level. When muscles are activated, they want to shorten. The lever uses electrodes to shock the muscle, which makes it shorten, producing a force that is recorded. It is also used to measure how much force the muscle is producing. One way scientists are making this more like it is in the mouse, is by shortening and lengthening the muscle periodically, as happens in our leg muscles as we walk. We use the data found from this experiment to derive equations when the muscle is cyclically activated.

Now, their muscle models work better than others because they incorporate one extra muscle protein, titin, into their calculations. Titin is the largest protein in our bodies and works as a giant string. For the purpose of this project, it stores energy in muscles and the muscle stretches, which can be used to shorten. The crazy thing about this spring protein is that it is much stiffer in active muscles than passive. Understanding how this affects body movement is under intense investigation in the lab. That's all I have for today. Next week, I will meet with Dr. Tester who is one of the research advisors for the project.

Until next time,

Krishna Patel

Understanding Matlab!

Hello all!

This week I met with Isaac Romero. Isaac primarily works with analysis program Matlab and the data gathered during simulations and testing of the bionic ankle. He graphs and compares all the different data and determines what can be modified to make the prostheses work more efficiently.

Jeremy Petak, another member of the team, wrote the Matlab code that Isaac works with. It pulls from lower-limb approximation. These groups work together to dorsiflex (contraction of the anterior muscle) and plantarflex (contraction of posterior muscle) to create torque to move the ankle. First, Isaac enters the subject's information into Matlab which then is automatically plugged into the equations of the muscle algorithm. To estimate human muscle strength, the lab uses mouse muscle information and scales those to roughly a human leg muscle. Now, there are many parameters that need to be found to run the prosthesis. if they are not correct, the user could hurt themselves. So, these are found via simulations in Matlab. The output data is compared to real human data. The programs optimizes the parameters to fit this data as closely as possible. The most important parameters are the muscle activation parameters (how much the muscle is on). Human EMG (Electromyography) data is used in simulations. Isaac uses this information and plugs it into pre-formed equations to parameters required to find torque.When this is done, the software is ready to be put into the BiOM for human testing.

I am still working on becoming IRB approved.

Until next time,

Krishna Patel

The BiOM!

Hello Readers!

My plan for the next few weeks is to meet with various lab members and understand their role is. Hopefully, I can get a greater grasp on the details of the research. I will try to convey what I learn in a way that's easy to understand.

I first met with Eric Lockwood, who is working on the engineer aspect of the research. Eric gave me a plethora of information, so let's dive into it.

First and foremost: data collection. The team compares the data that they get from testing against computer simulations through Matlab to see how well the device is running. The team simplifies the lower leg muscles by clumping them all together into the anterior (front of leg) and posterior muscle (back of leg) groups. The anterior muscle dorsiflexes (pulls the toe towards the shank) the foot at the ankle. The posterior muscle aids in stabilization of the lower leg and plantar flexes the foot at the ankle; this is the motor of the leg, pushing the foot into the group to help us walk forward. The goal of the lab's research is to reproduce human walking with a motorized lower limb prosthesis. Eric optimizes the Matlab code, and send it off to our collaborators in Boston, Bionx. They compile the code into a format that the BiOM prosthesis computer can read. The team then uploads the new code into the ankle.

Next on our list, actually getting the code into the prosthesis computer. This is not as easy as it seems, since the prosthesis is usually only meant to do this once, but we have to do it all the time! A Piecer board is connected to the BiOM, which allows to to talk to our laptops. From there, we upload the code to the ankle. There are three parts to this code: Motor Controller, Gait Controller and IMU (Inertial Measurement Unit). Motor Controller code tells the motor how much power should be delivered to the ankle. The gait Controller code figures out where the person is in the walking gait. It also communicates to the program to determine which activations are needed to power the motor. The IMU code does calculations and measurements. An encoder reads where the motor is at with respect to degrees. Once the program is uploaded to the prosthesis, real testing can begin! During testing, the lab uses a Pole Test, which is used to simulate a gait cycle and check if there are any major bugs. Then, they move to human  subject trials, but we will leave that for next time.

Update, I am currently working on getting IRB (Institutional Review Board) approval so I can finally work with the subjects! I will keep you updated as I go through the process as it is quite a long process to complete.

Until next time,

Krishna Patel

BionX 101

Hello. Sorry it's been such a long time since you've heard from me but I'm here now. This week Dr. Zhixiu Han, who is an engineer at BionX and is working with the lab testing a winding filament hypothesis based controller for a robotic prosthesis, flew in from Massachusetts and gave a talk to engineer students at NAU. I went to the talk because it was the perfect way to get an introduction to what the company has done in the past, what they are working on right now and what they hope to accomplish in the future. Also, I am in the process of getting onto the protocol so hopefully I will begin work in the lab soon.

The company was founded in 2006 by Hugh Herr, the head of Biomechatronics research group at the MIT Media Lab. Dr. Herr had to get both of his leg's amputated below the knees after a suffering severe frostbite caused by a rock climbing trip in New Hampshire. About 50 people work for BionX in Massachusetts. The commercial release date for the prosthesis, which is made of carbon fiber material, was early 2011 and 1,000 ankles have been sold and distributed since then. The device costs roughly $45,000. 

Now, let's talk about what role the body plays in all this. The human ankle acts as a spring by exerting and releasing energy and force into the foot. The goal of the prosthesis is to generate the same amount of ankle torque that is given off by a human ankle during the gait cycle. The gait cycle is the sequence of events that occur during normal walking. The BiOM system focuses on Biometric Propulsion which consists of two parts: stiffness modulation and net-positive power assist. Stiffness modulation is resistance to foot flat and net-positive power assist is power assist with toe or heel off. There is a picture explaining the different stages of the gait cycle to make this easier to understand. The BiOM replaces the human tendon and muscle unit. The BiOM prosthetic runs on a battery. In a day an average person can go through 1-3 batteries depending on how active the person is. This particular prosthetic reduces wear and tear of the hip and knee muscles and joints that would normally experiences a lot of force with a passive ankle. Passive ankles are ankle prostheses that aren't operated with a motor. Typically, a person with a lower-limb amputation and a passive ankle would have to get knee and/or hip replacement surgery 5-7 years after receiving the passive ankle. Using the BiOM, persons with lower-limb amputations would need surgery about 10-15 years after receiving the Bionic ankle prostheses. The BiOM uses a high-energy series elastic actuator which allows to the motor to give off more power by pre-charging the spring in the ankle to provide the proper energy required to make walking seem as natural as possible. It also uses 6 Degrees of Freedom Inertial Sensing to provide for real time terrain discrimination. 

The Gait Cycle

In the future the company hopes to develop a Mechanical User Interface which will discover different ways to attach the prostheses to the human body. They also hope to develop an Electronic User Interface which will address how the prostheses will distinguish how the device will know how to use nerve signals in order to control the muscle ankle. Another goal is to develop User Intent Control which will hopefully allow a person using the BiOM to perform activities other than walking and running like swimming. 

Dr. Hugh Herr's TED Talk on the BiOM:
https://www.ted.com/talks/hugh_herr_the_new_bionics_that_let_us_run_climb_and_dance

Until next time,
Krishna Patel