Prolonged mechanical ventilation has been shown to induce diaphragm dysfunction. Temporary diaphragm pacing is an interesting lead to halt or slow diaphragm dysfunction. A detailed description of the distal portion of the phrenic nerve would be the start for developing a new surgical approach for diaphragm pacing. Its in that perspective that we are launching a clinical observational study of the distal portion of the phrenic nerve in patients who undergo minimally invasive thoracic surgey. We hypothetize that the density of the fatty tissue surrounding the distal portion of the phrenic nerve might influence efficiency of the electrode during diaphragm pacing. The objective of this study is to find a correlation between the amount of fatty tissue around the nerve and the body mass index of individuals.

Implanted phrenic nerve stimulation is a validated technique to produce lung ventilation when the neural drive to breathe is abolished (e.g. congenital central alveolar ventilation during sleep, or during wakefulness in severe forms of the disease) or defective (e.g. periodic breathing due to chronic heart failure), or when it cannot reach the phrenic spinal motoneurons (e.g. high cervical spinal cord lesions). Several implantation techniques have been described. They include the surgical implantation of contact electrodes in intimate exposure with the trunk of the nerve, the laparoscopic implantation of hook electrodes within the diaphragm in the vicinity of the phrenic nerve endings (ref), and transvenous approaches that can be permanent (pacemaker like device) or temporary (central venous catheter-borne electrodes). The surgical implantation of electrodes in contact with the phrenic nerve ("periphrenic electrodes") is the better established of these techniques, having been described in the 1970s and continuously used since. Periphrenic electrodes have been implanted at the bottom of the cervical path of the nerve. This cervical approach is quick and easy to implement, but the results can be compromised by system damages resulting from neck movements and by the fact that the phrenic nerve is often anatomically incomplete at this level (namely above the junction with the accessory phrenic nerve). For these reasons, periphrenic electrodes are typically implanted within the thorax, at the level of the vena cava on the right and of the pulmonary hilum on the left. These locations are readily accessible by use of video-assisted thoracoscopic surgery or thoracotomy. Careful dissection of the phrenic nerve is necessary to separate it from the underlying anatomical plane and create the necessary access to position the electrodes. The possibility to position periphrenic electrodes over a segment of the phrenic nerve where it would be fully constituted and to do so without the need for nerve dissection would render intrathoracic phrenic nerve stimulation easier and safer to implement. It would also open the possibility of temporary implantation. From the anatomical description of the terminal branching of the phrenic nerve toward different portions of the diaphragm , we hypothesized that such "free" segment would exist between the point where the phrenic nerve pathway leaves the anterolateral angle of the pericardial base in direction of the diaphragm and the point where the phrenic nerve branches before entering the diaphragm muscle mass. This study was therefore designed to determine whether or not a free phrenic segment exists between the cardiophrenic angle and the diaphragm, to describe the anatomical characteristics of this segment, and to compare the nature, number and size of its constituting fibers with the corresponding characteristics as identified at the usual site of implantation of intrathoracic periphrenic electrodes. To this aim, we first performed a human cadaver study. This study allowed us to describe a fatty tissue surrounding the phrenic nerve which density vary from one patient to another, and from one side to the other. During routine surgical interventions by minimally invasive approach, we want to complete our observation by searching for a correlation between patients body mass index (BMI) and the density of fatty tissue surrounding the phrenic nerve.

The patients, depending on their BMI will be divided in three categories: - Category 1: 18,5 kg/m2 < BMI < 24,9 kg/m2 / - Category 2: 25 kg/m2 < BMI < 29,9 kg/m2 / - Category 3: 30,0 kg/m2 < BMI < 34,9 kg/m2 Fifteen patients will be included in each category. A fifteen seconds recording of each phrenic nerve during minimally invasive surgery will be reviewed by 2 independent observers in order to classify the phrenic nerve in three categories: - type 1: pericadiophrenic bundle is free of any relevant surrounding fatty tissue / - Type 2: pericardiophre-nic bundle is surrounded by one fringe of fatty tissue / - Type 3: pericardiophrenic bundle is surrounded by multiple fringes of fatty tissue

Inclusion Criteria:

• patients older than 18 years-old and not under a protection measure
• patients undergoing lung resection for a malignant or benign lesion
• surgical approach by minimally invasive technique (video-assisted thoracoscopic surgery or robotic-assisted thoracoscopic surgery)

Exclusion Criteria:

• surgical approach by thoracotomy
• intense pleural adhesions between the lung parenchyma and the mediastinal pleura.

The study population will be patients who consult for a benign or malignant pulmonary lesion. Resection of the lesion would have to be done by a minimally invasive approach (video-assisted thoracoscopic surgery or robotic-assisted thoracoscopic surgery).