Artificial intelligence in tongue diagnosis: classification of tongue lesions and normal tongue images using deep convolutional neural network.

OBJECTIVE: This study aims to classify tongue lesion types using tongue images utilizing Deep Convolutional Neural Networks (DCNNs). METHODS: A dataset consisting of five classes, four tongue lesion classes (coated, geographical, fissured tongue, and median rhomboid glossitis), and one healthy/normal tongue class, was constructed using tongue images of 623 patients who were admitted to our clinic. Classification performance was evaluated on VGG19, ResNet50, ResNet101, and GoogLeNet networks using fusion based majority voting (FBMV) approach for the first time in the literature. RESULTS: In the binary classification problem (normal vs. tongue lesion), the highest classification accuracy performance of 93,53% was achieved utilizing ResNet101, and this rate was increased to 95,15% with the application of the FBMV approach. In the five-class classification problem of tongue lesion types, the VGG19 network yielded the best accuracy rate of 83.93%, and the fusion approach improved this rate to 88.76%. CONCLUSION: The obtained test results showed that tongue lesions could be identified with a high accuracy by applying DCNNs. Further improvement of these results has the potential for the use of the proposed method in clinic applications.

Diagnosis of common intraosseous lesions of the dentomaxillofacial region by chemometry-assisted FT-IR spectroscopy in dental tissue samples.

PURPOSES: This study aimed to determine the differential diagnosis of three intraosseous lesions (odontogenic keratocyst (OKC), central giant cell granuloma (CGCG), and dentigerous cyst (DC)) of the dentomaxillofacial region with very similar radiological and clinical features by using chemometrics assisted FT-IR (Fourier transform infrared) spectroscopy in tissue samples. METHODS: Tissue samples (exposed to formaldehyde for a similar time) of 20-micron thickness belonging to 19 intraosseous lesions diagnosed histopathologically were obtained from the pathology laboratory. The samples were analyzed by FT-IR spectroscopic method using the 400-4000 cm-1 wavenumber range, and the obtained spectra of the samples were evaluated using the orthogonal partial least squares discriminant analysis (OPLS-DA) algorithm. RESULTS: The intraosseous lesions with different histopathological diagnoses were accurately and precisely clustered with different FT-IR bands corresponding to the main molecular vibrations, especially the phosphodiester region, of the tissue components using the proposed model with 3 latent variables. CONCLUSIONS: The model showed high sensitivity and specificity. The present study is the first to report the elucidation of clear spectral differences between similar lesions in the maxillofacial region. In the future, the FT-IR method may be used in the non-destructive classification of similar lesions in the maxillofacial region as an alternative to histopathological evaluation.

Ultrasonography in determining pubertal growth and bone age.

OBJECTIVE:: The purpose of this study is to evaluate the compatibility of ultrasonographic data with hand-wrist radiographs taken to determine the extent of pubertal growth and bone age in patients and investigate the usability of ionizing radiation-free ultrasonography instead of conventional radiography. METHODS:: In this study, a total of 120 children from 10 to 17 years old (mean age was 168 months ± 27.5 months) were treated with routine radiographs before orthodontic treatment, and ultrasonographic imaging was performed on the wrists the same day. Researchers examined the phalanges, sesamoid bone, and radial bone distal epiphysis-diaphysis comparatively in each patient by both imaging methods and statistical evaluation. RESULTS:: There was no statistically significant difference between conventional radiography and ultrasonography values at 13 points except for PP1 (proximal phalanges of the first finger), PP2 (proximal phalanges of the second finger), and radial epiphysis (p > 0.05). PP1, PP2, and radial epiphysis showed a statistically significant difference (p < 0.05). The CBA (bone age obtained from conventional radiographs) of the females was found to be larger than their CA (chronological age) and their UBA (ultrasonographic bone age). For males; the means of the CBA, UBA and CA values close to each other. In females and males; there was a strong correlation between the CA, the UBA and the CBA (p < 0.01). CONCLUSION:: Ultrasonography gives detailed information about epiphyseal diaphysis relations. It can be used as an alternative to conventional radiography in the detection of bone age and pubertal growth, owing to the absence of ionizing radiation.