In squamous cell carcinoma of the tongue, complete surgical resection of the tumor generally represents the primary treatment strategy. Among factors determining the disease prognostic, the surgical margins status is essential for the continued management of the patient, with or without complementary therapies. Generally, surgical margins are defined using visual and tactile elements. Unfortunately, in 15 to 30 percent of cases, they are defined as positive after histopathological examination, on average 5 to 7 days post-surgery.

Techniques have recently been developed to analyze the cancerous and non-cancerous nature of tissues, using the spectral properties of tumor cells. SpiderMass™ is an instrument equipped with a probe for tissue micro-sampling and real-time molecular profile analysis. It is thus ultimately a tool for anatomopathological decision support, for diagnosis and prognosis of the disease. Used in the operating room, SpiderMass™ could also become an innovative surgical tool to precisely define cancerous areas in situ and the margins status in intraoperative conditions.

Based on the principle of artificial intelligence, the SpiderMass™ machine must initially learn to recognize pathological tissues versus healthy. This project aims to accomplish this first phase of machine learning for squamous cell carcinoma of the tongue.

Inclusion Criteria:

• age : 18 years and older
• squamous cell carcinoma of the tongue
• surgical resection as first-line treatment
• tumor classification T1 to T4.

Exclusion Criteria:

• metastatic disease
• neoadjuvant therapies

Patients treated in the Maxillofacial Surgery Department of the CHU Amiens-Picardie for a squamous cell carcinoma of the tongue and for whom the first-line treatment is surgical resection