Loading...
Publication
Desenvolvimento e validação de um software de diagnóstico por scanner intraoral
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 3.11 MB | Adobe PDF |
Advisor(s)
Abstract(s)
Introdução: A cárie dentária é uma doença que compromete a saúde oral, causando desmineralização dos tecidos dentários. O diagnóstico precoce é essencial na prevenção. A utilização de um software especializado melhora a precisão e eficácia no diagnóstico.
Objetivos: Desenvolver e validar um software de diagnóstico (CavityFinder) para deteção de lesões de cárie como meio auxiliar ao diagnóstico clínico, através de Scanner intraoral.
Materiais e métodos: O desenvolvimento do software para análise de lesões cariosas envolveu a utilização de um modelo intraoral Frasaco A3® (frasaco GmbH™, Baden-Wurttemberg, Germany), canetas de acetato para simular diferentes tipos de cáries, dois scanners intraorais; iTero® (Align Technology™, California, USA), TRIOS 3® (3SHAPE™, Copenhagen, Denmark), e ferramentas de visualização 3D; Paint 3D® (Microsoft™, Washington ,USA), Exocad view 1.6.2® (Exocad™, Darmstadt, Germany) e MeshLab 2023.12® (Meshlab™, Karnataka, India). O algoritmo, implementado em Python 3.12.0® (Python™, Delaware, USA), identifica a arcada, cada dente e suas lesões, e foi integrado ao Visual Studio IDE 2022 v17.10® (Microsoft™, Washington, USA) para compilação e otimização. O software foi validado por meio de testes com cenários simulados de lesões de cárie.
Resultados: Os testes com o TRIOS 3® e Paint 3D® identificaram corretamente todas as cáries simuladas, sem erros, exceto em um caso com um falso positivo. O Exocad view 1.6.2® igualmente. O iTero® e MeshLab 2023.12® detetaram todas as cáries simuladas, mas apresentaram dois falsos positivos.
Conclusões: O projeto CavityFinder atingiu todos os objetivos, demonstrando eficiência na deteção de cáries em Closed Beta. A IA baseada em CNN é promissora. Este programa gratuito é adaptável, atendendo estudantes de medicina dentária, e é promissor para a educação e prática clínica.
Introduction: Dental caries is a disease that compromises oral health by causing demineralization of dental tissues. Early diagnosis is essential in preventing severe complications. The use of specialized software improves the accuracy and efficiency of dental caries diagnosis. Objectives: To develop and validate diagnostic software (CavityFinder) for detecting carious lesions as an aid to clinical diagnosis through intraoral scanning. Materials and Methods: Materials and methods: The development of the software for caries lesion analysis involved the use of an intraoral model Frasaco A3® (frasaco GmbH™, Baden-Wurttemberg, Germany), acetate pens to simulate different types of cavities, two intraoral scanners; iTero® (Align Technology™, California, USA), TRIOS 3® (3SHAPE™, Copenhagen, Denmark), and 3D visualization tools; Paint 3D® (Microsoft™, Washington, USA), Exocad view 1.6.2® (Exocad™, Darmstadt, Germany), and MeshLab 2023.12® (Meshlab™, Karnataka, India). The algorithm, implemented in Python 3.12.0® (Python™, Delaware, USA), identifies the arch, each tooth, and its lesions, and was integrated into Visual Studio IDE 2022 v17.10® (Microsoft™, Washington, USA) for compilation and optimization. The software was validated through tests with simulated caries lesion scenarios Results: The tests with TRIOS 3® and Paint 3D® correctly identified all simulated cavities without errors, except for one case with a false positive. Exocad view 1.6.2® did the same. iTero® and MeshLab 2023.12® detected all simulated cavities but presented two false positives. Conclusions: The CavityFinder project achieved all its objectives, demonstrating efficiency in detecting cavities in Closed Beta. The CNN-based AI showed promise. This free program is adaptable, catering to dental students, and holds promise for education and clinical practice.
Introduction: Dental caries is a disease that compromises oral health by causing demineralization of dental tissues. Early diagnosis is essential in preventing severe complications. The use of specialized software improves the accuracy and efficiency of dental caries diagnosis. Objectives: To develop and validate diagnostic software (CavityFinder) for detecting carious lesions as an aid to clinical diagnosis through intraoral scanning. Materials and Methods: Materials and methods: The development of the software for caries lesion analysis involved the use of an intraoral model Frasaco A3® (frasaco GmbH™, Baden-Wurttemberg, Germany), acetate pens to simulate different types of cavities, two intraoral scanners; iTero® (Align Technology™, California, USA), TRIOS 3® (3SHAPE™, Copenhagen, Denmark), and 3D visualization tools; Paint 3D® (Microsoft™, Washington, USA), Exocad view 1.6.2® (Exocad™, Darmstadt, Germany), and MeshLab 2023.12® (Meshlab™, Karnataka, India). The algorithm, implemented in Python 3.12.0® (Python™, Delaware, USA), identifies the arch, each tooth, and its lesions, and was integrated into Visual Studio IDE 2022 v17.10® (Microsoft™, Washington, USA) for compilation and optimization. The software was validated through tests with simulated caries lesion scenarios Results: The tests with TRIOS 3® and Paint 3D® correctly identified all simulated cavities without errors, except for one case with a false positive. Exocad view 1.6.2® did the same. iTero® and MeshLab 2023.12® detected all simulated cavities but presented two false positives. Conclusions: The CavityFinder project achieved all its objectives, demonstrating efficiency in detecting cavities in Closed Beta. The CNN-based AI showed promise. This free program is adaptable, catering to dental students, and holds promise for education and clinical practice.
Description
Dissertação para obtenção do grau de Mestre no Instituto Universitário Egas Moniz
Keywords
Diagnóstico Cárie dentária Software Cor