I would suggest an especially desgined respi-trace, connected with a sensitive data aquisition software, able to provide you with a good quality tidal volume signal. 

Thanks Andreas. That's the system I currently use, with labchart software. But, since the tidal volume of mice is extremely small (~ 10 ml / kg) and frequency very high (~ 280 resp / min), and considering the average wight of mice 25 g, the data are a bit inconsistent. That's why I was looking for a more sensitive system. 

This is a good question.   There is no safe non-invasive protocol for testing 'emphysema' (static compliance, lung volume subdivisions) in mice.   Respiratory pattern has not been well-described in elastase or transgene induced emphysema in mice.   We use repeated orotracheal intubation (without tracheostomy) to obtain static lung compliance and lung volume subdivisions.  The mice were repeatedly tested and recovered well.  The key is to use a proper catheter for orotracheal intubation and a lot of practice to get this to work.   As well, we developed a method for measuring specific airway resistance (sRaw = Raw * FRC) that should demonstrate increased sRaw after elastase injury or similar emphysema in mice (see Lofgren et al 2006) which is totally non-invasive and permits longitudinal study.    Finally, Wayne Mintzer's group from Johh Hopkins published a paper showing that using gas chromatography one can get measures of single breath DLco (DFCO in their paper, Fallica et al 2011) that would work to quantify diffusion limitation in murine emphysema.  Hope this helps to think about it.