Sign up Latest Topics
 
 
 


Reply
  Author   Comment  
Larry Flessland

Fortiori Design LLC
Registered:
Posts: 38
 #1 

In sensor firmware v0.1.2 we throw away scans that occur during wireless radio transmission because there appeared to be interference between the radio and detectors.  This interference occasionally created noisy SmO2 readings.

 

Another contributor to noisy readings appears to be sensor placement.

 

Lately I have been placing the sensor on my left thigh over the vastus lateralis muscle.  When the sensor is positioned correctly, I get steady SmO2 readings.  When shifting the sensor slightly it appears that the readings get noisy.  Look at the graphs in the two attached files.  In both examples the SmO2 reading is steady but gets noisy when the sensor shifts.

 

  • “NoisyRun.jpg”:  This is data from my Luce Line trail run.  You can see the SmO2 data start out smooth and gets noisy.  Once I reposition the sensor, the data is smooth again.
  • “NoisyInterval.jpg”:   This data is from interval training.  I could feel the sensor shift slightly but left it alone.  Once the sensor shifted, it felt like it stayed in position.  The SmO2 readings appear noisy especially at the lower readings.

Attached Images
Click image for larger version - Name: NoisyRun.jpg, Views: 77, Size: 105.16 KB  Click image for larger version - Name: NoisyInterval.jpg, Views: 73, Size: 109.21 KB 

Juerg Feldmann

Fortiori Design LLC
Registered:
Posts: 1,530
 #2 
Interesting observation.
 3 reasons why we see "noisy " readings:
  1. As mentioned, the placement of the  optodes. I mentioned it in another thread:  In muscle areas where you have major tendon sheets close  or tendon bundles  and you have receiver and transmitter ( R and T ) resp. one of them on the tendon sheet  and the other on the  actual fleshy muscle , than you see often  very noisy signals.
 Examples are placements  on vastus  lateralis and overlapping with ITB. Or in hamstrings area when placing on semimembranosus semitendinosus.
 In abdominal, rectus abdominalis area , when placing over intersections when doing  respiratory assessments. Sternocleidomastoideus , when using for  respiratory assessments and control
2. Depending on fixation of the sensor.  Circular  fixation  will create in  extreme ROM t often change in superficial pressure and therefor will influences the  "noise"
3. Eccentric load like in running due to the "stretch" mechanism of the fascia. See one of  many of our examples with first bike ( little to no eccentric load)  and than moving to running.

Attached Images
Click image for larger version - Name: JJ_bike_run_intervall.JPG, Views: 50, Size: 61.86 KB 

MTF Carter

Development Team Member
Registered:
Posts: 27
 #3 
Over the last year we have noticed variable tHb data in subjects with low adipose tissue thickness.

We get our thickness measurements from axial anatomical MRI scans before placing the NIRS device on the subject. That way we know the quality of the muscle as well as adipose tissue thickness before we interpret the data. However, most trainers and athletes do not have access or the money to send on an MRI so you can use skinfold calipers to do the same. If you are an athlete training with Moxy, chances are you do not have a muscle myopathy so your data should not be affected.  I have used several NIRS devices and so far Moxy seems to be the best when looking at tHb values.

In several cases, healthy subjects as well as patients with low adipose tissue thickness (<3.5 mm) can have tHb values ranging from 0.0 to 15.0 g/dL during exercise or cuff occlusion. If tHb = 0 than SmO2 = 0 and this data point must be excluded

The variability in tHb data is greater in subjects who have thin adipose + skin thickness (± 0.7 vs. ± 0.1).

figure 1:
   thick vs thin.JPG 
Fig. 1 is a comparison between two age and sex matched normal healthy subjects during a cardiovascular drug infusion. The open circle data points are from a subject with thin adipose tissue thickness (3.2 mm). The filled dot data points are from the second subject with thicker adipose tissue thickness (8.4 mm). 

When using PeriPedal to view data , it is not unrealistic to see tHb values = 0 however this is not a real physiological data and the tHb and SmO2 data need to be excluded. This is seen in figure 2 where the tHb goes to 0. Note that the data points within normal range are picked up every few seconds so that data can be used. Thus it is possible to have an SmO2 almost equal to 0 while the tHb remains within normal range.

tHb Noise.JPG

We call this "tHb dropout" in the Sinoway Lab. It is very easy to see as the "bottom" or lower tHb values fall to 0 or appear to drop out of the window. Hence the name tHb dropout. Only the upper level tHb values are physiologically relevant. SmO2 data with corresponding tHb values = 0 are not interpretable. 

You will loose some SmO2 data however the trend should still remain the same.   

juergfeldmann

Development Team Member
Registered:
Posts: 1,501
 #4 
carter  super interesting  and thanks   much  for sharing. We never stop learning and I will take much more care in the future on  assessing  skin fold better  than I  did in the past. Again thanks  so  much.
MTF Carter

Development Team Member
Registered:
Posts: 27
 #5 
Hello Moxy users and developers,

You all know I am about data quality on the front end of the study and making sure what I am seeing is real! 

Does anyone know of a surface thermometer that is ant+ ??

We have one that will feed analog signals into our PowerLab but it is a wire nightmare so we can't take it out on the road [frown] any ideas?? we will use this data to support Juerg's idea that skin temp/blood flow can influence the NIRS signal and that Moxy is very good at filtering that out vs. Portamon. 

I would like to get this into ant+ so we can do a big multi center gps NIRS tracking study on vascular patients. Thats my long term goal at least [smile]

I will start collecting this pilot data in 2 weeks hopefully...once the nurses get all the consents and screening visits in order. so much red tape [nono] haha!

Thanks,

Carter

 
juergfeldmann

Development Team Member
Registered:
Posts: 1,501
 #6 
Carter , thanks  for the   feedback.
 I was   doing this idea  with MOXY  and Portamon  as you know  well back and I used  my old Bio-harness as it measures  surface skin temperature  over Infrared  sensor on the back of the  unit.
 We did some interesting studies  to see, how  the surface  temperature  my  react in an on ice interval on ice hockey player  and whether we  would see a  difference in  Portamon TSI  %  and MOXY SmO2  %.
 Unfortunately  very small study  samples  of  only   8  so not  worse  to even start thinking on a  direction.
. The result  was  very similar a the one you did    with  Portamon and MOXY. If  we apply  surface  heat  TSI  goes up  and  MOXY  SmO2  goes  down. If  we applied  ice  than TSI  dropped  and SmO2   goes up.
 If  we  applied   surface pressure lie with pressure stockings  we had a  drop in TSI  %  but an increase in SmO2  MOXY.
 So  it  show  towards  a situation, where  Portamon TSI is a  result  of  different lases  of  reactions including surface  where   MOXY seem to juts show  reaction in the deeper  muscle   situation.
 This  than explains , when we  see TSI % going up  during a step test the skin temperature  went  very high   so blood  moved to the surface  so less in  muscle  and SmO2  drops.
 We had this in a love demonstration in Switzerland  with an athlete  and I  can dig out some   test results  where we  can show that TSI    increase during the step test but actually dropped  during the 1 min rest and  skin temperature  did  the same.
 So  bioharness may be one options  with their Skin temperature  sensor.
Roger

Moderator
Registered:
Posts: 266
 #7 
This is the only ANT+ temperature sensor that I know of.

https://buy.garmin.com/en-US/US/shop-by-accessories/fitness-sensors/tempe-/prod107335.html

It's meant to be worn on the outside of a jacket as an environmental temperature sensor.  I'm not sure if you could adapt it for skin surface temperature or not.
Roger

Moderator
Registered:
Posts: 266
 #8 
This is a calorimeter that I think is based on precise skin temperature measurements.  It deals with some of the issues of heat transfer occuring with evaporative cooling too so it might be better than just a temperature sensor.

http://www.metalogicscorp.com/

I think it uses ANT in some way.
juergfeldmann

Development Team Member
Registered:
Posts: 1,501
 #9 
One interesting idea  Per lundstrom  from Red Bull research used in his incredibility   work is the  tablet the   for core temperature  they  used. Fancy  but interesting.
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.

HTML hit counter - Quick-counter.net