Orbital wall fractures were noted in 57 patients (58.8% of cases) with the frontal sinus wall affected in 49 patients (50.5% of cases) and were accompanied by

pneumo-orbit in 15 patients (15.5%) and muscle incarceration in 2 patients. The jawbone (Figure 1(c)) was the site of a fracture in 50 patients (51.5% of cases) and Lefort fractures were noted in 19 patients (19.5%) with 7 cases of Lefort I fracture (7.2%), 4 cases of Lefort II fracture (4.1%) and 8 cases of Lefort III fracture (8.2%). A hemosinus was observed in 60 patients (61.8% of cases). Fractures in the lower floor were found in 18 patients (18.6%). In the vault, fractures concomitant with those of the bones of the face were observed (part belonging to the vault of the frontal bone: 35 (36.1% of cases), temporal bone: 18 patients (18.6% In our study, an association with traumatic brain injury was noted: there were 40 patients with cerebral contusion (Figure 1(a)) (41.2%) and 15 patients (15.5%) had pneumoencephaly. Extradural hematoma was present in 16 patients (16.5%) and subdural hematoma (Figure 1(a)) affected 13 patients (13.4%).

Imaging is an essential tool in the management of facial trauma to reveal lesions that threaten the patient’s life, therefore requiring urgent treatment [1] [4] . In low-income countries, such as Burundi, there is the problem of access to CT-scan and MRI which would allow a complete lesion assessment to be carried out. Few patients benefit from CT scans due to lack of sufficient financial resources. Facial trauma remains a crucial public health problem in our country. In fact, in our study, trauma to the face represented 27.1% of cases of trauma to the head. In the study by Nóbrega LM et al. [6] , facial trauma concerned 16.4% of all trauma cases and 15.6% of road accidents in the study by Caldas IM et al. [7] . Facial trauma, in our series, concerned more young adults of the fourth to fifth decades and males and occurred during road accidents and many of our patients (50.66%) resided in the capital of Burundi, Bujumbura. As also noted by most of the authors like Khalatbari S et al. [8] , Gurung US et al. [9] , Yadav SK et al. [10] and Holmgren EP et al. [11] , male subjects, through their socio-behavioral habits, are more exposed to trauma from road accidents, brawls, assaults, and violent games. They also showed that the occurrence of road accidents was linked to non-compliance with the highway code, the increase in the number of motorcycles in public transport and the poor condition of certain roads.

In our study, soft tissue lesions of the head were noted in all our patients. Their presence and importance would motivate the realization of a tomodensitometric examination. This exploration made it possible to note that the trauma caused fractures of several bones at the same time in most cases and the lesions were limited to the soft tissues only in 6.7% of cases. The presence of several fractures in a single patient indicates the severity of the trauma [12] . The mid-level of the face, more precisely at the level of the zygomatico-maxillary complex, presented more fractures than the other two floors as also noted by some authors [13] [14] [15] [16] . The middle floor, being made up of a mosaic of bones of varying size and shape, the trauma can lead to fractures of several bones at the same time. Their tomodensitometric analysis after trauma therefore requires multiplanar and volume reconstructions to reduce any unrecognized fractures [2] [5] [17] [18] [19] . Orbital fractures, more found in our series (58.8%), are rarely isolated. In fact, they were accompanied by fracture of the wall of the frontal sinus and pneumencephaly in 50.5% and 15.5% of patients, respectively. This observation was made by Avery L L et al. [2] who noted most orbital fractures were associated with those of the frontal sinus and the base of the skull and were accompanied by intra-orbital damage and the eyeball. The fractures of the jawbone found in our series concerned 51.5% of patients. These fractures, as well as those of the orbit, are seen in severe trauma that causes complex fractures [20] [21] [22] . Lefort fractures were noted in 19.5% of cases and were dominated by Lefort III fractures (8.2%). The latter achieve complete cranio-facial disjunction and are rarely isolated [19] . The predominance of Lefort III fractures was noted by Scheyerer MJ et al. [3] . Our finding is different from that made by most of the other authors who noted that there was a predominance of Lefort II fractures among those of Lefort [9] [19] [23] [24] . Lefort II and Lefort III fractures are frequently associated with potentially life-threatening brain damage [25] [26] . Facial fractures were accompanied by hemosinus in 61.8% of patients in our series. Hemosinus occurs when the mid-upper or upper level of the face is involved with a fractured sinus wall [15] . In our series, there were cranioencephalic lesions associated with fractures of the face. Fractures of the vault bones were dominated by fractures of the frontal bone (36.1% of cases). These fractures are often accompanied by damage to the brain parenchyma. This impairment of the cerebral parenchyma was marked, in our series, by contusions (41.2% of patients), pneumoencephaly (15.5% of patients) and extra-dural hematomas (16.5% of patients). The concomitant occurrence of facial mass lesions and brain lesions is recognized and is indicative of high energy trauma [19] [27] . Brain lesions were noted by several authors as the most frequently associated with facial fractures in 50% of cases and especially during trauma to the upper face of the face [2] [3] [21] [28] [29] . In a traumatic context, the tomodensitometric exploration of the face must imperatively extend to the brain so as not to ignore the lesions the brain lesions, some of which can put the patient’s vital prognosis at risk [2] [3] [15] [17] .

The limitations of our study are linked to its retrospective nature. Furthermore, more patients are referred by various hospitals that do not have scanner equipment. Access to their complete medical file becomes difficult and some data, especially epidemiological and clinical, are not accessible.

Nevertheless, this study sheds light on the extent of facial trauma and the types of lesions that are important to expect in Burundi and in another country with the same socio-demographic and economic aspects.