**5. Conclusions**

The identification of the volume condition is important for the clinical managemen<sup>t</sup> of many patients in the emergency room or in wards of general medicine and cardiology. We propose an automated approach for the classification of the volemic status, based on the processing of B-mode US video-clips

of the IVC and on the extraction of pulsatility features. The presented results sugges<sup>t</sup> that this approach may be useful to ge<sup>t</sup> more reliable clinical indication from the US monitoring of IVC. Investigation over a larger dataset will however be necessary to test the actual effectiveness of the proposed method. Moreover, our results hold true in conditions of normal respiratory function and cardiac rhythm. It is reasonable that our classifier will not apply to patients with more complex cardio-respiratory conditions; however, the same approach could be applied to develop models fitting their conditions.
