SVPT — Next Generation devices for biofeedback in Physical Therapy (Article 2 in series)

Authors: Kalpana S Mair (B.P.T), Sushrut J Mair.

In these series of articles we would like to present and discuss a modular physical therapy biofeedback platform developed by us, which, in our opinion, can be an invaluable daily practice tool for therapists. The platform has been prototyped using the latest mature and proven technologies in the field of electronics and mobile device software. This is the second article where we introduce the actual platform and details around it. The first article in the series is available here.

In this particular article we introduce details of the new platform and how it stacks up against the 6 key attributes (which were also outlined in the previous article). We also discuss what has already been prototyped and our future plans. The device platform has been designed to generally be spec’d to the following guidelines:

The base unit is capable enough to be extended for future sensors to be added as needed. Sensor assemblies are usually targeted to more than one condition.

Here are images of the current platform that is under use. This is actually the 3rd version of the prototype. In another article we’ll discuss about the prior two and how we evolved the platform (If enough readers are interested).

The Base Unit:

Images of the various sensors follow:

The Load Pressure Sensor (LPS) — The LPS is designed to assist in various therapies like:

• Related to knee and foot conditions, one example being rehabilitation of ACL injury patients
• For Lumbrical grip analysis
• Anytime a gross analysis of how much pressure can the patient apply is needed

The Multi Utility Sensor (MUS) is for -

• Finger grip analysis (for traumatic , neurological hand rehabilitation)
• Pain threshold monitoring, much like a dolorimeter (can analyze and have objective measure of myofascial pain, trigger point or any tenderness)

Finally, a screenshot of the mobile app driving all these (only the app home screen for now — app details shall be treated in a later article):

Moving on, we have 3 different videos of the LPS and MUS sensors in action. Each video is followed by a brief description to help understand what is going on. Future articles shall dissect and discuss each parts of the system (base unit, sensors and the mobile app) in detail.

Video 1 — LPS

At the very start of the video, we are shown the base unit, the LPS and the mobile app home screen. The app has already connected to the base unit. At 00:07 in the video, the therapist loads the patient profile screen. From there, a patient is selected (“ksm” is a dummy patient used for demos only). Then a sensor selection is made (at 00:13), in this case it is the LPS. The therapist then presses “Start Therapy” which loads up the live therapy screen. At this point, the base unit verifies that the sensor is connected and this is indicated by 3 red LED’s blinking on it and the therapy screen graph settling to zero. At 00:30, the therapist places the sensor beneath the patient’s knee joint. The exercise being demonstrated in this video is known as static quadriceps exercise. Typically, post ACL surgery patients experience discomfort and pain and are not able to apply normal levels of pressure on the LPS. However as the therapy progresses, a gradual increase in the LPS readings are expected. See the graph of a real patient with ACL injury below to get a better idea. Coming back to the video, you can see that as the patient does isometric contraction of quadriceps, the LPS captures the readings and displays it on the app (00:32–00:55).

The image above shows a 25 year old male, in otherwise good health that had experienced an ACL injury. He underwent surgery and post-surgical rehabilitation for the ACL. The graph above is for the complete rehabilitation period. A positive trend is clearly visible in the graph. The red line marked as ‘Threshold’ is essentially a target set by the therapist for the patient to achieve. It is also visible in the live session data too (revisit the video). The patient has to strive to at least meet that goal. Depending on the progress, the threshold can be moved to a different value. All threshold change history is also stored with the other persistent data so that it can be used for later analysis.

Video 2 — MUS for Finger Grip Analysis

This video shows the MUS with the base unit and app, at the start. The MUS is further modularized and can accept a variety of attachments (hence the name ‘Multi Utility Sensor’). At 00:04, you can see the therapist fixing an attachment to the main MUS body (blue). This attachment is typically used for finger grip analysis and dolorimeter therapies. At 00:35, you can see the therapist zooming into the live graph so as to be able to better view smaller values — this can be changed at any point in time during a live session. At 00:45 the therapist demonstrates applying pressure on the MUS via their index finger and the relevant readings are displayed on the screen. This can be done for all fingers or as needed. As their therapy progresses, the improvement in their ability to apply pressure is seen which suggests improvement in grip strength.

The image above is of a 50 yr old female having polyneuropathy. The graphs span about 3+ months and show how her finger grip has slowly improved over that period of time. It goes to demonstrate how a degenerative neurological condition affects humans and how can disciplined therapy show results that also contribute towards improving the patient’s quality of life. The larger spikes in the graph (reaching at between 3 and 4 on the Y axis) are recordings of the patient’s stronger hand. This particular patient had visibly more weakness in one hand than the other. As you sweep across the graph, you can see how the weaker hand is gradually catching up with the stronger one.

Video 3 — MUS sensor as Smart Dolorimeter

This video shows the use of the MUS as a dolorimeter — an instrument that can help quantize the ‘degree’ of pain a patient experiences. The patient had complaints of tenderness or localized pain and the therapist has marked viable trigger points on the patient’s back (The MUS is used to quantify how much pressure the patient can bear at a certain point in time. As therapy progresses (e.g. via trigger point release, manual therapy or electrotherapy), the patients ability to bear pressure should increase. This would be evident in the graph.

The image above captures data of a 32 yr old female complaining of Rt Trapezitis. It can be seen from the graph that as the trigger point is being released, the patient’s ability to bear sustained pressure at higher levels increases.

The intent of this article was not to go into detail about the base unit or each sensor and the app (that will, of course be covered in future articles), but to introduce the system and assess it against the six attributes covered in the first article:

• Provide extremely objective therapy data that can be correlated across sessions — The system gives very objective data for each therapy as can be seen in the videos above. The data can be correlated on day to day basis and the correlation is evident from historical charts (see historical data related attribute below)
• Provide an intuitive and easy way for patients to view their activity during sessions, preferably real time during a session — For each session the app is highly visible with all its details and the graph can be zoomed in and out at any point in time. The patient can easily correlate the graph with the therapy they are undergoing — in fact we’ve had patients letting us know that the live graph was like having a positive reinforcement for them. In absence of such data, it was difficult for them to comprehend progress in any meaningful manner
• Be modular so that various modalities are served as far as possible with a single device or fairly limited number of devices — Both the LPS and MUS sensors are designed to assist in various different therapies involving a multitude of conditions
• Persistently and reliably store patient therapy data across any number of sessions and allow therapists to call up historical sessions data at any point in time (as an example, for trend analysis) — The images after each video (above) are essentially historical graphs pulled off persistently stored data from therapy sessions, spanning across months. It is pretty clear how they can help in overall trend analysis of the patient’s progress
• Link patient data with their medical condition so that this context can be leveraged to provide high value information to therapists — Each patient has a profile with details of their condition(s) stored persistently in the app. This helps open up an advisory for therapists based on analytics. Something like, “Patient A’s condition X has been seen by 76% of therapists alongside condition Y”, or, “Therapy regimen ABC has helped 64% patients with condition Z to a 81% faster recovery”. In addition, we have also done early POC’s of predictive analytics based off SVPT data and results are encouraging enough for us to delve deeper into assessing feasibility
• Be small form factor and low cost. Be preferably battery powered — The base unit and the sensors are small and light enough that they can be carried in a small backpack or even a ladies handbag. A single rechargeable Li-Ion battery powers the base unit and the sensors. If needed, the base unit and sensors can also be run off a wall wart (using a 9V power adapter). The system uses readily availble components, enclosures and rides on top of mature ecosystems — all these help in keeping its cost low enough

Current Status:

• The system has been in use since the past almost 2 years (base unit + LPS + mobile app). The MUS has been in use from the past 6+ months. The app supports all Android OS versions KitKat onwards (4.4 and greater). All the graphs above are real patient data represented anonymously
• We’re prototyping a 3rd sensor, called the ExtFlex sensor — this sensor is used for all therapies where highly objective metrics are needed for ROM of the patient anatomy. This sensor will be introduced in the same base unit and mobile app combo. It is in advanced stages of testing and being stabilized

Future Plans:

• We have good ideas on how would we introduce foot pressure analysis as well as gait analysis in this system. We have not started building it out yet but have plans to do so in the coming months
• Once we gather enough data, we would like to run analytics over it to hopefully be able to generate advisories for therapists
• As discussed earlier, we have POC’d ( Proof of Concept) predictive analytics on data from the system with encouraging results. We plan to dig deeper into this and a future article will cover the same
• We have concrete thoughts on enhancing the system with remote monitoring capabilities. For e.g., the therapist can leave the system at the patient’s site (with a patient version of the mobile app) under the expectation that the patient would regularly use it for therapy. The therapy details would be synced with the therapist’s version of the app so that they would be on top of the patient’s progress (or lack of it) at all times. This is of course not applicable to all conditions but still has a wide enough audience to consider it
• While we do all of the above, i.e., introduce new capabilities, we are also continuously imbibing learning from what has already been prototyped and is being used. We are continuously working to bring in fixes and refinements based on those learnings. Some of them will be discussed in future articles

We hope that you have liked what you have seen of SVPT so far and are excited by its possibilities as a companion to physical therapists and their patients. Please do send us a note on any questions you may have! And do standby by for further articles on this system!