RESUMEN
OBJECTIVES: To evaluate the performance of a commercially available Generative Pre-trained Transformer (GPT) in describing and establishing differential diagnoses for radiolucent lesions in panoramic radiographs. MATERIALS AND METHODS: Twenty-eight panoramic radiographs, each containing a single radiolucent lesion, were evaluated in consensus by three examiners and a commercially available ChatGPT-3.5 model. They provided descriptions regarding internal structure (radiodensity, loculation), periphery (margin type, cortication), shape, location (bone, side, region, teeth/structures), and effects on adjacent structures (effect, adjacent structure). Diagnostic impressions related to origin, behavior, and nature were also provided. The GPT program was additionally prompted to provide differential diagnoses. Keywords used by the GPT program were compared to those used by the examiners and scored as 0 (incorrect), 0.5 (partially correct), or 1 (correct). Mean score values and standard deviation were calculated for each description. Performance in establishing differential diagnoses was assessed using Rank-1, -2, and - 3. RESULTS: Descriptions of margination, affected bone, and origin received the highest scores: 0.93, 0.93, and 0.87, respectively. Shape, region, teeth/structures, effect, affected region, and nature received considerably lower scores ranging from 0.22 to 0.50. Rank-1, -2, and - 3 demonstrated accuracy in 25%, 57.14%, and 67.85% of cases, respectively. CONCLUSION: The performance of the GPT program in describing and providing differential diagnoses for radiolucent lesions in panoramic radiographs is variable and at this stage limited in its use for clinical application. CLINICAL RELEVANCE: Understanding the potential role of GPT systems as an auxiliary tool in image interpretation is imperative to validate their clinical applicability.
Asunto(s)
Diagnóstico Diferencial , Radiografía Panorámica , ConsensoRESUMEN
OBJECTIVES: To evaluate the feasibility of frozen soft tissues in simulating fresh soft tissues of pig mandibles using cone beam CT (CBCT). METHODS: Two fresh pig mandibles with soft tissues containing 2 tubes filled with a radiopaque homogeneous solution were scanned using 4 CBCT units and 2 field-of-view (FOV) sizes each. The pig mandibles were deep-frozen and scanned again. Three cross-sections were exported from each CBCT volume and grouped into pairs, with one cross-section representing a fresh and one a frozen mandible. Three radiologists compared the pairs and attributed a score to assess the relative image quality using a 5-point scale. Mean grey values and standard deviation were obtained from homogeneous areas in the tubes, compared using the Wilcoxon matched-pair signed-rank test and subjected to Pearson correlation analysis between fresh and frozen physical states (α = .05). RESULTS: Subjective evaluation revealed similarity of the CBCT image quality between fresh and frozen states. The distribution of mean grey values was similar between fresh and frozen states. Mean grey values of the frozen state in the small FOV were significantly greater than those of the fresh state (P = .037), and noise values of the frozen state in the large FOV were significantly greater than those of the fresh state (P = 0.007). Both mean grey values and noise exhibited significant and positive correlations between fresh and frozen states (P < 0.01). CONCLUSIONS: The freezing of pig mandibles with soft tissues may serve as a method to prolong their usability and working time when CBCT imaging is planned.