If you’ve played golf or tennis for any extended period of time, you have felt the extensive strain that these activities place on your elbow. We have begun utilizing Vibromyography (VMG) as a noninvasive means for measuring the muscle effort produced in the upper extremities. This technology has been tested extensively under conditions of dynamic isometric contraction and has been shown to accurately predict muscle force during maximum voluntary contraction levels. In order to provide the maximal benefit of this technology to the clinician, physical therapist, or trainer, we have begun isokinetic testing in the upper arm. The pilot study began addressing how well VMG assessment correlates to Biceps Brachii and Triceps force using isokinetic contraction over a 135 degree range of motion of the elbow.

VMG recordings were taken from the Bicep Brachii (BB) and the medial head of the Triceps Brachii (TB) of an adult male (age 24) during seated elbow extension/flexion in a Biodex System 3 muscle dynamometer (Biodex Medical Systems, Shirley, NY) (Figure 2).

Recordings were obtained using two TSD250 VMG Transducers interfaced to the MP150 Data Acquisition System (BIOPAC Systems, Goleta, CA).

The protocol consisted of 8 isokinetic flexions at an angular velocity of 30^o/sec over an ROM of 135 degrees. Data from the transducers, along with torque, position, and velocity were simultaneously collected using 5 channels of the MP150 and preprocessed using BIOPAC Systems, Inc.  AcqKnowledge 4.2 software. Preprocessing included application of the Vibromyography Filter to the BB and TB data, this is a wavelet packet filter that converts the VMG data to an assessment of absolute muscle effort (Figure 3)

Figure 3. Five channel data acquisition showing 8 elbow extension and flexion events, with Channel 4 & 5 showing the processed BB and TB VMG signals used to represent total muscle force. X-axis represents elapsed time in seconds.

The eight individual flexion events were identified and averaged to obtain a single representative event (Figure 4).

Figure 4.  Average of 8 elbow flexion events into one event using the zero crossing of velocity as a reference point.  X-axis represents sample number; sampling is at 62.5 Hz.

Estimated muscle force from the VMG was converted to torque using kinematic data (Ramsay 2008). A polynomial fit was generated for the moment arms for respective muscle groups (Figure 5).

Figure 5. Polynomial fit of the moment arms as a function of angle over the ROM of 5^o – 125^o, from the measurements of Ramsay, et al, 2008.

The recorded torque from the Biodex was corrected for gravitational loading (mass of the limb and mass of the dynamometer arm).  Since the BB and TB share an antagonistic relationship, torque generated by the TB was subtracted from the torque generated by the BB (Figure 6).

Net torque (BB-TB; Figure 7) was compared to the gravitationally corrected elbow torque measured by the dynamometer using linear regression (Figure 8).

Figure 7. Measured elbow torque (top) and estimated torque from VMG recordings (bottom) as a function of elbow flexion angle.

Figure 8.  VMG Torque as a function of BB and TB muscle effort (shown as torque) demonstrates a linear  correlation for the 20-100 degree ROM of a seated elbow flexion

A linear correlation (R² = .54; p<0.0001) was observed between VMG and muscle force over the range of motion of 20-100^o. While antagonistic muscle activity is generally thought to be minimal during open-chain contraction, we have observed that antagonistic muscle activity is, in fact, substantial during dynamometer assessment and without inclusion of the triceps activity; correlation observed between VMG predicted torque and measured was poor.  The correlations observed here lead to the suggestion that VMG can provide a good estimate of muscle force during isokinetic contraction if all appropriate muscle is included in the assessment. As flexion of the forearm involves over eighteen different flexors, the inclusion of additional flexor activity in the analysis would be expected to improve the observed correlations.

References:

Ransay et al. (2008) Muscle moment arm and normalized moment contributions as reference data for musculoskeletal elbow and wrist joint models

One of the principal owners at Transform, Paul Watson, a quite impressive, experienced trainer originally from the Toronto area, took time out of his busy day to explain to some friends of ours about the value of MyoWave.

A client who had been working with Paul for roughly a decade, named Don, was also interviewed after undergoing a variety of exercises with three MyoWave sensors attached via elastic straps. He underscored what Paul told our friends, that seeing the muscle activity in real time and under various conditions, really drove home for him how his body was functioning and where he should focus more of his training.  MyoWave provided a real incentive to the client to work a little harder and to focus a little more on what the trainer was trying to accomplish.

This reminded me of an incident that occurred with the patient of a prominent physical therapist in Binghamton about a year or so ago. We had been asked by the PT to evaluate a patient in his seventies who appeared to have little to no function in his right foot due to a neurological problem.

He had been in therapy for some time and was losing hope that the work was making any difference; he still walked with a prominent limp, assisted by a cane, and could not discernibly move his ankle or his toes. We placed the gentleman on a shuttle and located one of the MyoWave sensors on the tibialis anterior of his “good” leg.

Tibialis Anterior..... Source: http://docpods.com/

We asked the patient to flex his ankle so that his toes pointed toward his head. Although we had never tested output of that muscle before, MyoWave reported a strong response, which we noted. Then we repeated the maneuver on his compromised leg. To the naked eye it was impossible to see that any movement was occurring but clearly his foot was not “dead” as he had feared.  While reporting a significantly lower muscle effort, it was immediately obvious to all of us watching the muscle activity reported on the screen that he was getting a response from the tibialis anterior due to the flexion of his presumably non-functional ankle and foot.

The impact to his psyche was immediate and powerful.  He was overjoyed that the work he had been doing did not appear to be in vain and the positive reinforcement he received was obvious in his effusive comments to the PT that he certainly wanted to continue therapy. As he walked out, arm in arm with his wife, it was clear that MyoWave did a lot more than simply report muscle activity.

We at Sonostics have found that the phenomenon of Social Media and the potential benefits it has as a marketing tool are too just too great to pass up, but like so many others, we had to start at the beginning. On the surface, social media is easy. The basic idea is that you generate some content (like this blog for example), get a bunch of people to subscribe to it, and then sit back and watch as your network expands into the hundreds of thousands. People can’t wait to hear what you have to say, and after a few product references, you end up with money pouring in like rain from the sky. … well maybe it’s not quite that easy.

The reality is that unless your company already has an established following, finding people to subscribe to your blogs, tweets, etc, is more of a challenge than some might think. There is something of a Catch-22. If you don’t have a lot of subscribers, writing content seems silly and wasteful since you’re limited to a very narrow viewing audience. On the other hand, if you don’t have any content, finding subscribers is that much more difficult. The solution appears to be in something of a compromise. The reality is that no matter what, generating meaningful content on a regular basis is absolutely critical. Whether they be two-line Facebook posts, or three-page long journal articles surrounding the detailed merits of a technology like MyoWave, content is king. Your limited audience will grow, albeit slowly, and then faster over time. But don’t fret, as long as you don’t make a habit of it, great content can be recycled later on so there really is no such thing as a wasted blog or Tweet.

But how does getting followers reading blog entries lead to business growth and increased sales of your products? While there’s no simple formula for success in this transition, one possible solution lies in lacing each of your blogs/posts/tweets with mentions of your company’s products and their associated benefits. For example, in a blog about social media, one might mention that those who suffer from knee pain stand to benefit greatly from MyoWave, a cutting-edge technology by Sonostics. You might go on to explain the benefits of the technology and even mention how easy it is to use and how it exceeds the capabilities of every other comparable technology in terms of price and utility. Those who suffer from Knee pain will be all too curious about the mention and will click through to the carefully placed product links.

At the end of the day, however, there really is no secret solution – no pie in the sky method for succeeding in social media. Each person’s successes or failures are unique and there is no right or wrong answer on how to succeed. The good news is that social media, and all of the tools available are almost 100% free, so what have you got to lose in trying it out?

If you’ve had successes or failures in social media, leave a comment on this blog regarding what you did, what you didn’t do, and how it turned out. We hope to hear back from you!

If you’ve ever experienced knee pain that doesn’t seem to go away, no matter what you do, you’ve come to the right place. Many people suffer from knee pain, and when they visit their physician to find out what’s going on, they get nothing but a prescription for a pain killer and are told to stop whatever it is that is causing their pain.

Knee pain can be a result of a myriad of things, but one thing that is rarely considered is muscle imbalance. In order to function properly, the muscles surrounding our joints need to be working in an efficient manner. For example, generally speaking, our quadriceps should be 3x stronger than our hamstrings! When this ratio is altered, muscle function is thrown off balance and we begin to experience unexplained pains. Sometimes this can be associated with a significant injury from high school or college, but sometimes there is no explanation at all.

This is where MyoWave comes in. MyoWave is designed to identify muscles imbalances surrounding the knee, and Sonostics has several locations where Knee Pain Screenings are being offered. When a client arrives for their Knee Pain Screening, they’ll be offered a pair of shorts to change into.

Once measurements of the thigh have been obtained, the MyoWave Screener will apply 2 elastic straps which hold 4 sensors to the thigh. These sensors listen to the muscle fibers as they vibrate; they are NOT electrodes. the screener will then attach the MyoWave “backpack” (a small box about the size of a SmartPhone which collects the vibration data and sends it on to the computer) to the waist band of the client’s shorts.

Next, the client performs 10 step-ups on each leg. Meanwhile, MyoWave is collecting and analyzing data obtained from the 4 sensors on the thigh. After completing the step-ups, the client and Screener sit down together to analyze the data and identify any muscle imbalances.

Once the specific muscle imbalances have been identified, the Screener provides the client with exercises that are designed to specifically target certain muscle groups (eg, quadriceps, hamstrings, hip abductors, etc.). Before leaving the Knee Pain Screening, the client schedules a 2-3 week follow-up for re-assessment and is given a complete report of their muscle imbalances, pictures and descriptions of the exercises to be performed, as well as a diagram of the leg anatomy.

Click on the video below to see a typical screening:

After a 6-week program of performing specific exercises, we expect that the client will have begun to re-balance their knee, thereby decreasing their knee pain. It is expected that the client will continue to exercise and strengthen the knee, and will have the opportunity to purchase future Knee Screenings to get objective information on how they are progressing.

Here’s an example of how the program can help:

Click here to get more information or to schedule a Knee Pain Screening visit.

Joint pain is the second most common complaint that people have when they visit a physician. Over 50 million individuals in the U.S. are afflicted, and approximately one-third of the population over age 45.  Joint pain results in over 40 million physician visits each year.  In addition, joint pain is the leading cause of disability in the U.S., affecting approximately 18% (8 million) of the adult population.  Direct and indirect costs of joint pain were estimated at over $65 billion in 1992 (Yelin, et al, 1992) and have grown considerably since that time.

Of the various joint pains, knee pain dominants, accounting for more than 14 million physician visits each year in the U.S.  In fact, knee pain is the third most common complaint of individuals seeking the advice of a physician (Table 1) (CDC 2009).

Table 1. Most common complaints of ambulatory care patients who are not seeing a physician for a regular physical, pregnancy check-up, or post-surgical follow-up.

When a patient with joint pain visits a physician they will usually be told they have a form of arthritis. Arthritis is a general term for a variety of joint diseases. There are three types of arthritis:  the most common by far is osteoarthritis, which is often referred to as ‘wear and tear’ arthritis, and is commonly thought to be a natural outcome of aging.  In addition, there are the more rare conditions of rheumatoid arthritis (an autoimmune disease), and ankylosing spondylitis (general associated with spinal pain, and is thought to be genetic in origin).

Most ambulatory care visits for arthritis involve persons over age 45 and over (75%) and predominantly women (66%).  Individuals with joint pain most commonly visit a primary care provider (52%), followed by orthopaedic surgeons (20%), and rheumatologists (17%).  The standard of care for patients who visit a physician is pharmacologic treatment (anti-inflammatories and pain medications).  In a recent study of 36.5 million physician visits for joint pain, over 70% of patients were given a total of 59 million drug prescriptions.  The most commonly administered drug was non-steroidal anti-inflammatory (NSAIDS), followed by analgesics (both topical, non-narcotic and narcotic), and corticosteroids (Hootman, et al 1997).

This standard approach to the treatment of arthritis is currently undergoing close scrutiny as it does not at all address the underlying cause of the pain. Rather, the pharmacologic approach simply allows the individual to mask the pain until joint degradation reaches a point where the knee has to be replaced (total knee arthroplasty – TKA).  Previously, it was believed that nothing else could be done, but studies such as the Arthritis Foundation’s “Arthritis Self-Management Study” have shown that simple preventative approaches, such as exercise, can be as effective as pharmacologic treatment (Lorig, et al, 1993).  Unfortunately, such approaches are severely underutilized.

The difficulty in pursuing preventative approaches to arthritis, and more specifically, exercise interventions, is that, historically, there has been no means for identifying the specific exercises which need to be undertaken by the individual to relieve the pain of arthritis.  As a result, a broad range of exercises have been provided to the individual, and after a short time, most individuals abandon their exercise routine, both because it is so time consuming, and since there is no feedback on whether progress was being made.

Yet, the basis for exercise intervention is well established. Since the beginning of the 20th century it has been recognized that specific muscle imbalances (either among complementary or agonist/antagonist muscle pairs) results in inappropriate distribution of forces in a joint, resulting in inflammation and eventually degradation (Truslow, 1909) – note that trauma and congenital defects can also lead to inappropriate joint loading, though these situations are far more rare.  Correspondingly, restoration of muscle balance corrects the joint loading, allowing the joint tissue to heal and the pain will disappear.

The challenge that has faced the healthcare community since the early 20th century has been one of identifying the specific muscle imbalances.  With the advent of electronic technology in the 1920 and 1930s, the technique of measuring the electrical potentials in contracting muscles (electromyography – EMG) began to be developed and it was widely hoped that this technology would permit accurate muscle assessment. EMG started to be utilized in the late 1960s, but EMG techniques could not be found which would permit comparisons between muscles or over time, and given the difficulty of application, was never widely accepted in the primary care clinic.  Integrated circuit technology developed in the 1980s did produce developments in EMG technology which allowed it effective use during neurosurgery.

In the 1970s and 1980s, feedback control muscle dynamometry was developed, and again, there was great hope that a means for assessing muscle activity would be available in the clinic. However, dynamometry requires that any exercise activity used in an assessment be open-chain so that the application to functional activity is unclear.  Moreover, muscle dynamometers report torque generated at a joint, rather than the activity of a specific muscles, so that direct application to a muscle training regimen was always indirect. The inability for muscle dynamometers to assess closed-chain activities resulted in insurance companies removing this assessment technique as a reimbursable evaluation, and their presence in clinics has been greatly reduced over the past decade.

Starting in the 1990s, a research group in Italy began to pursue the phenomenon of “muscle sound” as an alternative means for assessing the activity of specific muscles.  This research progressed from using microphones to record the sounds that muscles create when they contract, to small accelerometers to record the actual muscle body vibrations at the surface of the skin.  This new recording technique has been called mechano-myography (MMG) in order to distinguish the technique from EMG, but more accurately, the technique should be called vibromyography, or VMG, as it is the muscle vibrations on the surface of the skin which are being recorded.

Recent work out of the State University of New York at Binghamton, has shown that advanced signal processing techniques can be applied to a recorded VMG signal from a muscle undergoing contraction during a functional activity, resulting in a reflection of the absolute force being generated by the muscle.  This approach works for muscles in either isometric contraction or in isotonic contraction.  Moreover, the repeatability of the recording permits comparisons between different muscles, between different people, or comparisons in the change of a person’s muscles over time.  In other words, there now exists a convenient and reliable technique available for assessing muscle balance during closed-chain functional activity.

Arthritis has not attracted as much attention as other chronic health conditions such as diabetes, cancer and heart disease, perhaps because the associated rates of mortality and hospitalization are relatively low.  However, the burden of arthritis on individuals and society is extremely high.  Indeed, given the high prevalence of arthritis, the impact on the workplace, the healthcare system, and the quality of life for the affected population, the impact of joint pain is widely considered to be higher than that of other chronic diseases.   The goal of Sonostics Inc. is to provide communities throughout the U.S. (and perhaps eventually the world) with the resources necessary to indentify the specific muscle imbalances existing in individuals with joint pain as early as possible, and then help these individuals address their imbalances.  The company is starting with a focus on knee pain – evaluating both those with knee pain, and those who want to ensure that no muscle imbalance exists which could result in future knee pain and joint damage.  It is expected that there will then be progressive developments in the technology permitting additional joint assessments, such as elbow, hip, back and neck.  The development of VMG technology for the non-invasive assessment of muscle balance appears to be extensive, and the impact on the healthcare will be immense.

References
CDC. Summary health Statistics for U.S. Adults. National Health Inverview Survey 2009.

http://www.cdc.gov/nchs/fastats/docvisit.htm

Hootman JM, Helmick CG, Schappert SM. Magnitude and characteristics of arthritis and other rheumatic conditions on ambulatory medical care visits in the U.S.  Arthritis and Rheumatism 47:571-581, 2002.

Hurley MV, The role of muscle weakness in the pathogenesis of osteoarthritis. Rheum dis Clin North Am. 25:283-286, 1999.

Lorig KR,  Mazonson PD, Holman HR. Evidence suggesting that health education for self-management in patients with chronic arthritis has sustained health benefits while reducing health care costs.  Arthritis Rheum 36:439-446, 1993.

Turslow W. The principles of muscle balance as applied to orthopedic practice.  J. Bone and Joint Surgery, 1909.

Yelin E, Callahan LF for the National Arthritis Data Work Group. The economic cost and social and psyghological impact of musculoskeletal conditions. Arthritis Rheum 38:1351-1362, 1992.

The video shows how young athletes as well as older adults can discover if muscle imbalance is preventing them from functioning at top level and without pain.  This measurement device, which converts muscle vibrtion to absolute effort – totally non-invasively –  provides objective metrics of muscle performance heretofore impossible to capture.

You can see how far you have ridden and perhaps how many calories you have burned, or watts you have generated, or what your average revolutions per minute were – and from that information you may be able to derive what sort of progress you’re making while devoting those hours of blood, sweat and tears to improving your leg muscle efficiency and your cardiovascular condition.

What you don’t get, however, is insight as to how those muscles in your legs are interacting, how they contribute to the ride, the rate at which they fatigue, and perhaps how they respond to hydration or nutrition.

It was with that goal in mind that Tim and Sara decided to tackle the Keiser stationary bikes at the Island Health & Fitness Center in Ithaca, NY last week. They decided to ride at a fixed gear with reasonably high rate of resistance – 18 on the Keiser scale – maintaining a constant cadence (65-70 RPMs) for five minutes while they remained firmly in the saddle.  A later test will explore the difference in muscle contribution between saddle riding versus standing.

They strapped vibromyography sensors to their vastus lateralis and vastus medialis muscles on both legs to enable the MyoWave muscle assessment device to calculate the amplitude of effort generation throughout the ride.

MyoWave collects about 60 data points per second so the amount of data that had to be analyzed over that 5 minute period was rather substantial. MyoWave spent about 2 seconds generating a report after the ride was stopped. The results demonstrated that there is an opportunity to use MyoWave to analyze individual muscle contribution and muscle balance in a dynamic exercise over a protracted period of time.

Since bicycle riding causes muscles to exert a varying degree of effort as the leg spins through each cycle we focused on the overall average amplitude of muscle effort.  However, MyoWave does permit the user to zoom in on individual sections of the data so it would be possible to analyze each section of the stroke and consider the ratio of peak effort between Muscle A and Muscle B.

So what did this data tell us?

Tim’s dominant leg is his left. From the data, it appears that as the five minute test progressed, he relied more and more on only one muscle – the vastus medialis  – to carry the predominant load.  This was a surprise – we did not expect to see such single muscle dominance.

The left vastus lateralis, that had performed virtually equally from the outset, really diminished in effort as time passed and it seemed clear that an imbalance between the two medialis muscles was evident.

Sara, on the other hand, used both vastus lateralis muscles to carry the load and their output remained quite constant throughout the entire exercise. It would appear that fatigue showed, as evidenced by the decline in both of her vastus medialis muscle fiber contractions over time.  This is a distinct difference between VMG and electromyography, which counter-intuitively tends to show increased amplitude as muscles fatigue.

Bottom line – we learned that MyoWave can digest data from lengthy exercises which opens the tool up to be used in a broader range of activities.  For example, serious biking enthusiasts might learn a great deal about their performance and be able to tailor training regimens to their own specific muscle balance metrics. For racers, this will provide a wealth of data that can translate into valuable training edges.

Since we have architected the device to try to minimize the effect of impact spikes, one of our next experiments will be on the treadmill. Imagine evaluating running styles with different shoe types, body position, etc.  The world is now that much bigger.

The purpose of the database is to help establish both what “healthy” knee muscle behavior looks like, and correspondingly, the nature of knee muscle imbalance in those subjects who already complain of knee discomfort. One widely accepted factor in the etiology of osteoarthritis (OA) of the knee is the existence of knee muscle imbalances which can compromise the proper function of the knee joint, and over the course of many years lead to joint damage.

If we can identify the normal balance of flexor and extensor muscle activity of healthy knees during these activities, we will be well positioned to identify those individuals in whom muscle behavior is imbalanced.  Our goal is to help reduce the age-related onset of OA in those individuals with muscle imbalance at an early age, which would then permit early intervention (appropriate exercise) to correct the muscle imbalances.

These muscle assessments are being undertake using MyoWave, a remarkable device that converts muscle vibration to absolute effort via a technology called vibromyography (VMG).

Jonathan was one of the subjects who responded to our invitation to perform a simple “step-up” routine which we utilized in this database development project.

Jonathan is a previous college football player with numerous knee injuries in his youth, but was functioning well until the last year.

At the time of testing, Jonathan was struggling just to get up the stairs to the room where the evaluation was taking place but what transpired that day and over the following three weeks was quite inspiring and deepened our appreciation for the sometimes magical information this technology can provide.

Below is a video we did with him:

Bob “Kingpin” Eres developed ProSportsBroadcasting.Com and hosts interviews with companies describing new technology and Sonostics was selected to be the interview target this morning.

To support the interview about vibromyography in general and MyoWave in particular we took the opportunity to film a video in preparation that tested the flexors and extensors of Adam, one of our software architects, during a full golf swing. This seemed to be a useful experiment given the type of audience Bob gets on his website.

Here’s the video:

Not knowing what was under the hood, we spent weeks trying to reverse engineer the application, applying band-aid fixes and patches wherever we could.  However, as almost any programmer who has worked on somebody else’s antiquated code can tell you, this was no simple task.

MyoWave, a product designed to measure muscle activity and detect muscle imbalance, had taken many different twists and turns since its initial inception and as a result, the code reeked of disarray. There were single files thousands of lines long and any flow which may once have existed had long since vanished.

With looming deadlines only weeks away, we were faced with a choice.  Should we try to salvage the existing code base or should we toss it aside in favor of a totally new, untested approach?

In a larger company the bean counters surely would have told us that the risk was much too high, that we had invested too much to simply throw away something that “worked.”  Something, even something unpleasant, is always better than the unknown.  Or is it?

Throwing caution and conventional wisdom to the wind, we decided to attempt the unthinkable and started architecting a new application which would share the same MyoWave name, but be an entirely different application from the ground up.  Looking back one month later, I can say without a doubt that we made the right choice.

This is what we came up with.

We wanted the MyoWave application to be extremely elegant, simple, fast, and we wanted it to run on practically any platform we could conceive of with a minimal amount of rework.  We also wanted the MyoWave software to be as unique and revolutionary as the Vibromyography technology it supports.  The application architecture we came up with set out to do exactly that.

Most desktop applications rely on a desktop framework for handling UI elements… Aqua, GTK, Qt, Tk, and the list goes on and on.  However, none of those frameworks really works well across all platforms, especially once you get into the mobile space.  You might be able to snag a handful, but certainly not the whole group.  As a result, we decided to build our entire interface around HTML, CSS, and JavaScript (an approach similar to HP’s revolutionary webOS).  With HTML and CSS, we would have a limitless canvas on which to draw and to do whatever we pleased in a way no desktop framework could offer.

However, we still wanted this to be a desktop-only application which would not rely on any type of network connection or dedicated webserver.  As a result, we built a custom WebKit (.NET) “web browser” with APIs allowing for bidirectional communication between C# and the JavaScript running therein. In this way we were able to open files, communicate over USB, and accomplish other feats which would not normally be possible using JavaScript alone.

…but we still had a problem.

Vibromyography relies upon enormous computations which would not be feasible in Javascript – or any interpreted language for that matter.  C# might be able to keep up, but would force us to abandon our goal of a cross-platform existence.

As a result, we set out to write an entirely separate application in straight C which would be extremely fast, run natively on just about any platform, and handle all of the computations we required.  It would be controlled entirely by the master MyoWave app, and so, a simple command line interface would suffice.  The computational results would be output in a JSON string, communicated via stdout through the C# handler, and passed onto the javascript code where they would be displayed to the screen. The end result was as beautifully elegant as it was innovative.

The resulting application (written in one month’s time) is blazingly fast and sports a slick UI not easily possible in your run-of-the-mill C# or GTK app.  Additionally, by swapping out the few hundred lines of C# glue for some Objective-C (OS X/iOS) or Java (Android), MyoWave will be able to run on any platform which supports WebKit (which these days is pretty much everybody!).  Furthermore, if we decide one day to make the transition to SaaS (Software as a Service), our code reuse will be close to 100%.

I hope this example helps to serve as a testament to the notion that frequently throwing out software that “works” can allow something far greater to be born.  Just as a farmer prunes back the branches of an apple tree, so too should software developers cut out bad code so they can focus all their efforts on something which is simpler, cheaper, and better.