We have developed an experimental cable for recording of chest and abdominal respiration effort. It is a silicone cable with a resistive silicone-carbon core. The recorded signal is roughly proportional to lung volume. 

Practical evaluations in several laboratories in the years 1995-2005 demonstrated that the cables:

• are reliable and hygenic and do produce very good respiration signals
  
• also pick up external electrical fields (50Hz mains, static-charge fields)
  
• 'roll' over the body despite being rather tightly strapped.
  

Based on that result we have made a theoretical redesign that:

• is based on the same materials
  
• is more elastic
• is virtually insensitive to external electrical fields
• can be more easily attached to the body
• stays better in position
  

We are now working with a company to realize this design.

---

The old evaluations were based on cables of 4.5mm and 3.5mm diameters. The 4.5mm types had lengths of 80cm, 100cm, 120cm, and 150cm. The 3.5mm prototypes were 30cm,  40cm,  50cm,  60cm, 80cm, 100cm, 120cm, 150cm. The gauges had a resistance typically between 300 and 1000 Ohm/cm. So, a 100cm cable was typically between 30kOhm and 100kOhm. Inspiration and expiration caused the resistance value to increase and decrease roughly proportional to respiratory volume by about 0.1 - 1%. The active part of the gauges ended in integrated silicone-insulated wires. The respiration gauges did not have connectors and electronics. Nor any special means to attach them to the body.

The circuit below was used to connect the gauge to an electrode input. The resistance bridge produces the respiration signal, superimposed on a DC voltage. DC is rejected by an RC network with long time-constant (R=1MOhm, C=10uF, results in TC=10s).
In some cases the DC component should not be rejected, for instance in diagnosing Rett syndrome or when you wish to standardize the tension by which the gauge is applied to the patient. In such cases, the DC component can be very large and thus drive the amplifiers out of range. A simple solution for this problem is on http://edfplus.info/specs/DCattenuator.html .