Artificial Intelligence for the Diagnosis of Oral Lesions
Trial statusNot yet recruiting
Trial phaseNot listed
Trial typeObservational
Biological sexAll
Age18+
SponsorAssistance Publique - Hôpitaux de Paris
About this trial
Squamous cell carcinomas of the upper aerodigestive tract are among the most common cancers worldwide, with the oral cavity being the most frequent site. Oral cavity squamous cell carcinomas (OCSCC) represent a major cause of morbidity and mortality, mainly due to high rates of locoregional or metastatic recurrence and the frequent occurrence of second primary tumors. Unlike oropharyngeal squamous cell carcinomas, human papillomavirus (HPV) is not involved in the carcinogenesis of OCSCC.
In some cases, OCSCC develop from oral potentially malignant disorders (OPMDs), such as leukoplakia and erythroplakia, which have a worldwide incidence of 3-5%. The malignant transformation rate of OPMDs ranges from 3% to 50%, reflecting their marked heterogeneity. Although several clinical, histological, and molecular factors have been proposed to identify patients at high risk of malignant transformation, none have demonstrated sufficient clinical utility to date. In other cases, OCSCC arise from clinically normal oral mucosa in patients with OPMDs located at a distance and/or with established risk factors, particularly tobacco and alcohol use.
Currently, no chemopreventive or preventive strategy has been established as a standard of care to prevent malignant transformation of OPMDs. Improving the prognosis of OCSCC therefore requires the development of tools to better identify high-risk OPMDs and to enable the earliest possible diagnosis. Early detection of OPMDs is essential for secondary prevention of OCSCC. However, conventional oral examination based on visual inspection and palpation has limited sensitivity, and clinical recognition of OPMDs remains challenging. Consequently, there is a clear need for improved methods to enhance early detection and risk stratification of OPMDs.
Main objective:
To develop a tool to aid in the diagnosis of cancerous lesions in the oral cavity using Artificial Intelligence (AI). This tool appears promising in meeting the current needs of the oral cavity practitioner community.
Eligibility criteria
This trial does not accept healthy volunteersQualifiers
Patient aged ≥ 18 years
Patients followed up in the Oral Mucosa Pathology Department (Maxillofacial Surgery and Stomatology Department, Pitié-Salpêtrière Hospital, AP-HP, Paris) between January 1, 1970, and December 31, 2023, with a diagnosis of potentially malignant oral lesion and/or oral cavity cancer.
Disqualifiers
Photograph of the lesion unavailable (in standard care)
Trial population
Patients followed in the Department of Oral Mucosal Pathology (Maxillofacial Surgery and Stomatology Department, Pitié-Salpêtrière Hospital, AP-HP, Paris) from January 1, 1970, to December 31, 2023, with a diagnosis of potentially malignant oral lesions and/or oral cavity cancer.
Trial design
Design model
Cohort
Time perspective
Retrospective
Treatments tested in this trial
Not listed
Trial groups
No trial groups listed
Trial outcomes
Primary outcomes
1
Develop a tool to aid in the diagnosis of cancerous lesions in the oral cavity using Artificial Intelligence (AI)
To develop a tool to aid in the diagnosis of cancerous lesions in theoral cavity using Artificial Intelligence (AI). This tool appears promisingin meeting the current needs of the oral cavity practitioner community. To achieve this objective, anonymized clinical photographs of oral lesions, along with relevant clinical data routinely recorded in medical charts, will be collected. All photographs will undergo retrospective review by experienced specialists in oral and maxillofacial surgery. The experts will independently assess the images and establish a reference diagnosis. In cases of disagreement, a consensus diagnosis will be reached. The complete dataset, including image data, associated clinical variables, and reference diagnoses, will be used to develop and internally validate a machine learning algorithm for the automated classification of oral lesions
Time frame
Through study completion, an average of 9 monthsLocations
Sponsors and contacts
Click on the lead sponsor to view all of their trials.
Assistance Publique - Hôpitaux de Paris
Lead sponsor
BPIfrance
Collaborator
Institut Universitaire de Cancérologie, Sorbonne University
Collaborator
Health Data Hub (France)
Collaborator
Source ClinicalTrials.gov
This trial is not recruiting at the moment. You can still explore other options: