Magister Scientiae Dentium - MSc(Dent) / Objective: The objectives of this study were to evaluate the diagnostic capability of 5th year students using digital imaging, conventional bitewing radiographs (BW), printed film on paper and to compare the results with the observers’ experience level to detect early interproximal caries lesions on radiographs. To map the literature in regards to different diagnostic methods that students use globally in dental clinics by conducting a scoping review. Methods: A cross-sectional study was conducted with senior dental students (Reg No: BM 19/9/8). Three digital radiographs were shown individually to the students on a screen, 3 bitewing radiographs and 3 printed films on paper were passed on individually to all students with a questionnaire; with a viewing time of 2 mins per radiograph; thus 9 radiographs in total were viewed. A control group consisting of specialists from both the Restorative and Radiology Departments had finalized the answers prior to conducting the study regarding the presence/absence of caries and its depth on all 3 different radiographic images. The answers recorded by students were divided into 5 categories, R0: Intact surface, R1: Radiolucency in outer half of enamel, R2: Radiolucency in inner half of enamel, R3: Radiolucency in outer half of dentin, and R4: Radiolucency in inner half of dentin. A protocol specific for the objectives of this study was developed according to the criteria for a scoping reviews. Relevant databases (Pubmed, Scopus, Ebscohost, Science Direct, Wiley Online Library and Cochrane Library) were searched to identify evidence which was restricted to the English language for the period 2015–2021. Results: Comparison using an Anova test on the 3 different diagnostic methods showed that the type of method used affects the presence detection and size measuring accuracy, where the P-value confirms a significant difference. The Prisma flow diagram showed a result of the most common methods for diagnosing interproximal caries happen to be visual examination and bitewing radiography. Conclusions: Senior dental students have shown good accuracy in detecting the presence of interproximal caries. Although the accuracy in detecting the size of carious lesions was poor.

Este trabalho teve como objetivo modificar um algoritmo de esqueletização e criar outro para quantificar imagens radiográficas. Foi proposto um tratamento digital em radiografias panorâmicas de 50 voluntárias diagnosticadas como osteoporóticas e não osteoporóticas em um estudo anterior. Para verificar a validade deste recurso digital, foi elaborado um questionário que juntamente com as radiografias foi distribuído entre quinze examinadores radiologistas. Estes analisaram de forma subjetiva 99 imagens, respondendo quais os critérios adotados para a avaliação da arquitetura óssea trabecular. Os dados obtidos no questionário aplicado e no tratamento digital das imagens foram submetidos à análise descritiva para verificar o nível de concordância nas hipóteses de diagnóstico e definir um padrão ouro. Foi construída uma curva ROC, para verificar o ponto de corte e estabelecer qual filtro de Gauss era o melhor, com raio variável de 30, 35, 40, 45 ou 50. O filtro escolhido foi o de raio 30, pois apresentou melhor especificidade (94,3%), mesmo tendo uma baixa sensibilidade (50%). Os níveis de concordância entre os examinadores ficaram abaixo de 80%. Contudo foi registrada alta significância entre os dados percentuais de pontos pretos e de pontos terminais (p<0.001). O processamento de imagens através da esqueletização em radiografias panorâmicas pode tornar-se um valioso recurso para os cirurgiões-dentistas. / The purpose of the present work was to modify a skeletonization algorithm and to create an algorithm to quantify radiographic images. A panoramic radiographic digital treatment of 50 volunteers diagnosed as otheoporotics and non-ostheoporotics in a previous work was done. In order to verify the validity of the digital resource, a questionnaire was elaborated that along with the radiographies was distributed among 15 radiologists. The examiners analyzed in a subjective way 99 images, responding which criteria were adopted for the evaluation of the trabecular bone architecture. The data obtained in the applied questionnaire and in the digital image treatment were submitted to a descriptive analysis in order to verify the agreement in the diagnostic hypothesis and to define a golden standard. A ROC curve was built to check the cut point and to establish which Gauss filter was the best, with a variable ratio of 30, 35, 40, 45, or 50. The ratio 30 filter was chosen since it presented the best specificity (94.3%), even having a low sensibility (50%). The agreement levels among the examiners were below 80%. However a high significance among the percentage of black points was registered (p<0.001). The method of image processing trough the skeletonization in panoramic radiographies can become a valuable support for the dentist-surgeons.

Este trabalho teve como objetivo modificar um algoritmo de esqueletização e criar outro para quantificar imagens radiográficas. Foi proposto um tratamento digital em radiografias panorâmicas de 50 voluntárias diagnosticadas como osteoporóticas e não osteoporóticas em um estudo anterior. Para verificar a validade deste recurso digital, foi elaborado um questionário que juntamente com as radiografias foi distribuído entre quinze examinadores radiologistas. Estes analisaram de forma subjetiva 99 imagens, respondendo quais os critérios adotados para a avaliação da arquitetura óssea trabecular. Os dados obtidos no questionário aplicado e no tratamento digital das imagens foram submetidos à análise descritiva para verificar o nível de concordância nas hipóteses de diagnóstico e definir um padrão ouro. Foi construída uma curva ROC, para verificar o ponto de corte e estabelecer qual filtro de Gauss era o melhor, com raio variável de 30, 35, 40, 45 ou 50. O filtro escolhido foi o de raio 30, pois apresentou melhor especificidade (94,3%), mesmo tendo uma baixa sensibilidade (50%). Os níveis de concordância entre os examinadores ficaram abaixo de 80%. Contudo foi registrada alta significância entre os dados percentuais de pontos pretos e de pontos terminais (p<0.001). O processamento de imagens através da esqueletização em radiografias panorâmicas pode tornar-se um valioso recurso para os cirurgiões-dentistas. / The purpose of the present work was to modify a skeletonization algorithm and to create an algorithm to quantify radiographic images. A panoramic radiographic digital treatment of 50 volunteers diagnosed as otheoporotics and non-ostheoporotics in a previous work was done. In order to verify the validity of the digital resource, a questionnaire was elaborated that along with the radiographies was distributed among 15 radiologists. The examiners analyzed in a subjective way 99 images, responding which criteria were adopted for the evaluation of the trabecular bone architecture. The data obtained in the applied questionnaire and in the digital image treatment were submitted to a descriptive analysis in order to verify the agreement in the diagnostic hypothesis and to define a golden standard. A ROC curve was built to check the cut point and to establish which Gauss filter was the best, with a variable ratio of 30, 35, 40, 45, or 50. The ratio 30 filter was chosen since it presented the best specificity (94.3%), even having a low sensibility (50%). The agreement levels among the examiners were below 80%. However a high significance among the percentage of black points was registered (p<0.001). The method of image processing trough the skeletonization in panoramic radiographies can become a valuable support for the dentist-surgeons.

No description available.

There is a need for an improved bone mineral density measurement procedure for neonates. Currently, measurements are made using single photon absorptiometry (SPA). The poor reproducibility of this method means that it has little direct clinical diagnostic application and is therefore not suitable for diagnosing disease in individual patients. A technique using digital radiography has been developed to measure bone mineral density. Digital images of phantoms and chicken bones were acquired at two kvp settings of a digital angiographic unit. Digital information from water, aluminum and lucite phantoms were used to calculate effective mass attenuation coefficients of the phantom materials. These values were subsequently used in bone mineral density calculations of sections of the chicken bones. The bone mineral densities of the chicken bones obtained from the digital radiography method were compared to SPA measurements. The digital radiography method gave consistently higher bone mineral densities for the bones than SPA. This could be due to the differences in measurement technique as SPA scans a single slice whereas digital radiography images a large area of the bone. / Thesis / Master of Science (MSc)

Yes / We investigated the potential for digital tomosynthesis (DT) to reduce pediatric x-ray dose while maintaining image quality. We utilized the DT feature (VolumeRadTM) on the GE DefiniumTM 8000 flat panel system installed in the Winnipeg Children’s Hospital. Facial bones, cervical spine, thoracic spine, and knee of children aged 5, 10, and 15 years were represented by acrylic phantoms for DT dose measurements. Effective dose was estimated for DT and for corresponding digital radiography (DR) and computed tomography (CT) patient image sets. Anthropomorphic phantoms of selected body parts were imaged by DR, DT, and CT. Pediatric radiologists rated visualization of selected anatomic features in these images. Dose and image quality comparisons between DR, DT, and CT determined the usefulness of tomosynthesis for pediatric imaging. CT effective dose was highest; total DR effective dose was not always lowest – depending how many projections were in the DR image set. For the cervical spine, DT dose was close to and occasionally lower than DR dose. Expert radiologists rated visibility of the central facial complex in a skull phantom as better than DR and comparable to CT. Digital tomosynthesis has a significantly lower dose than CT. This study has demonstrated DT shows promise to replace CT for some facial bones and spinal diagnoses. Other clinical applications will be evaluated in the future. / Medical Physics Division at CancerCare Manitoba and GE Healthcare (Waukesha, WI).

The dominant powder scintillator in most medical imaging modalities for decades is Gd2O2S:Tb due to the very good intrinsic properties and overall efficiency. Except for Gd2O2S:Tb there are alternative powder phosphor scintillators like Lu2SiO5:Ce and Gd2O2S:Eu that has been suggested for use in various medical imaging modalities. Gd2O2S:Eu emits red light and can be combined mainly with digital imaging devices like CCDs and CMOS based detectors. The purposes of the present thesis, is to investigate the fundamental imaging performance of a high resolution CMOS based imaging sensor combined with custom made Europium (Eu3+) activated Gd2O2S screens in terms of Modulation Transfer Function (MTF), Normalized Noise Power Spectrum (NNPS), Detective Quantum Efficiency (DQE), Noise Equivalent Quanta (NEQ) and Information Capacity (IC) covering the mammography and general radiography energy ranges. The CMOS sensor was coupled to two Gd2O2S:Eu scintillator screens with coating thicknesses of 33.3 and 65.1 mg/cm2, respectively, which were placed in direct contact with the photodiode array. The CMOS photodiode array, featuring 1200x1600 pixels with a pixel pitch of 22.5 m , was used as an optical photon detector. In addition to frequency dependent parameters (MTF, NPS, DQE) characterizing image quality, image information content was assessed through the application of information capacity (IC). The MTF was measured using the slanted-edge method to avoid aliasing while the Normalized NPS (NNPS) was determined by two-dimensional (2D) Fourier transforming of uniformly exposed images. Both parameters were assessed by irradiation under the RQA-5 protocol (70kVp digital-radiography) recommended by the International Electrotechnical Commission Reports 62220-1 and the W/Rh, W/Ag beam qualities (28kVp digital-mammography). The DQE was assessed from the measured MTF, NNPS and the direct entrance surface air-Kerma (ESAK) obtained from X-ray spectra measurement with a portable cadmium telluride (CdTe) detector. The spectral matching factor between the optical spectra emitted by the Gd2O2S:Eu and the Gd2O2S:Tb screens and the CMOS optical sensor, evaluated in the present study, was 1 and 0.95 respectively. The ESAK values ranged between 11.2-87.5 Gy , for RQA-5, and between 65.8-334 Gy , for W/Rh, W/Ag beam qualities. It was found that the detector response function was linear for the exposure ranges under investigation. Under radiographic conditions the MTF of the present system was found higher than previously published MTF data for a 48 m CMOS sensor, in the low up to medium frequency ranges. DQE was found comparable, while the NNPS appeared to be higher in the frequency range under investigation (0–10 cycles/mm). NEQ reached a maximum (73563 mm-2) in the low frequency range (1.8 cycles/mm), under the RQA 5 (ESAK: 11.2 Gy ) conditions. IC values were found to range between 1730-1851 bits/mm2. Under mammographic conditions MTF, NNPS and NEQ were found comparable to data previously published for the 48 m CMOS sensor while the DQE was found lower. The corresponding IC values were found ranging between 2475 and 2821 bits/mm2. The imaging performance of europium (Eu3+) activated Gd2O2S screens in combination to the CMOS sensor, investigated in the present study, was found comparable to those of Terbium (Tb) activated Gd2O2S screens (combined with the CMOS sensor). It can be thus claimed that red emitting phosphors could be suitably used in digital imaging systems, where the Silicon (Si) based photodetectors are more sensitive to longer wavelength ranges, and particularly in the red wavelength range. / -

Muitos pacientes submetidos à exodontia de terceiros molares relatam, no pósoperatório, que percebem alterações na oclusão. Por isso é interessante avaliar se realmente existe esta possível movimentação para distal dos segundos molares inferiores após a exodontia dos terceiros molares. Os objetivos deste trabalho foram analisar e verificar se existe variação da angulação do longo eixo do segundo molar em relação ao longo eixo do primeiro molar por meio de tomadas radiográficas periapicais digitais pelo sistema Digora®. A amostra constituiu-se de 20 pacientes entre 15 e 25 anos de idade que possuíam terceiros molares inferiores não irrompidos, bilateralmente, com indicação para extração e que se apresentavam de maneira simétrica com relação ao seu posicionamento. Os dentes foram extraídos, sendo um dos lados mantido apenas o coágulo (grupo controle) e o outro lado (grupo experimental) recebeu o enxerto do material proposto e membrana de cortical óssea bovina desmineralizada (Gen-derm). No Adobe Photoshop 7.0 foi possível realizar os traçados dos longos eixos dos primeiros e segundos molares inferiores e a intersecção entre eles. O programa ImageJ 1.41 foi responsável pela obtenção dos valores das angulações. Os exames foram realizados utilizando medidas em imagens obtidas no controle pós-operatório imediato, de 2, 6 e 24 meses nos dois grupos estudados. Os resultados mostraram que não houve diferença estatisticamente significante entre os grupos experimental e controle, e também não houve diferença ao longo do tempo, sugerindo que não existe movimentação para distal dos segundos molares inferiores após a extração dos terceiros molares e alterações na oclusão. Unitermos: Terceiro molar, Radiografia digital, Movimentação dentária. / Many patients submitted to the extraction of third molars, report, postoperatively, that they feel alterations in their occlusion. Thus, it is interesting to assess whether this possible movement to the distal of lower second molars, following the removal of third molars, really exists. This work aimed at analyzing and verifying, through digital periapical radiographies of the Digora® System, whether the angulation of the second molar´s long axis varies in relation to that of the first molar. The sample was comprised of 20 patients between 15 and 25 yrs, who presented non-erupted lower third molars, bilaterally, referred to extraction and in a symmetrical manner, in relation to their position. The teeth were extracted, being maintained, in one of the sides, only the clot (control group), while the other side (experimental group) received the graft of the proposed material and demineralized cortical bone bovine membrane (Gen-derm). In the Adobe Photoshop 7.0, it was possible to perform the outlining of the long axes of lower first and second molars and the intersection between them. The 1.41 ImageJ Program allowed the angulations values, and the exams were performed by utilizing measures in images obtained in the immediate postoperative control of 2, 6 and 24 months, in the groups studied. No statistically significant difference was seen between the experimental and control groups, as there was no difference as time passed, suggesting that there is no movement towards the distal of lower second molars, following the extraction of third molars and alterations in the occlusion.

Avaliar a influência no diagnóstico de cáries proximais das imagens geradas por dois escâneres e quatro máquinas fotográficas digitais, em formato JPEG e TIFF, e comparar o desempenho destes equipamentos ao das radiografias convencionais em relação à presença e profundidade de lesões de cárie proximais. 56 dentes extraídos foram radiografados com filme Kodak Insight em um aparelho de raios X Yoshida Kaycor, operando com 70kv e 7mA, com tempo de exposição de 0,40s. As radiografias resultantes foram digitalizadas pelos escaneres CanonScan e Genius ColorPage, e máquinas fotográficas digitais Canon Powershot G2, Canon RebelXT, Nikon Coolpix 8700 e Nikon D70s, nos formatos JPEG e TIFF. As radiografias convencionais e as imagens digitais obtidas foram avaliadas por três examinadores em duas sessões de observação. Classificaram a presença de acordo com uma escala de três níveis e a profundidade de acordo com uma escala de três níveis, quando julgassem que havia lesão. Os dentes foram seccionados no sentido mésiodistal e observados ao microscópio em luz polarizada com aumento de 10x, para a verificação da presença e profundidade da lesão de cárie. Foram determinados concordância intra e interexaminadores, probabilidade de não diagnóstico, sensibilidade, especificidade e acurácia para cada equipamento, em ambos os formatos, e para as radiografias convencionais. Os resultados foram submetidos à análise de variância a dois critérios, demonstrando que não houve diferenças estatisticamente significantes entre os formatos. A concordância intraexaminadores foi de moderada a satisfatória, e a interexaminadores de regular a moderada. A probabilidade de não diagnóstico variou de 1,34% (Convencional) a 52,83% (Escaner CanonScan/JPEG). A sensibilidade variou de 0,24 (Canon RebelXT/JPEG) a 0,53 (Convencional), a especificidade de 0,93 (Nikon Coolpix/JPEG; Canon Powershot/TIFF; Cânon RebelXT/JPEG e TIFF) a 0,97 (Escaner CanonScan/TIFF e JPEG), e a acurácia variou de 0,82 (Canon RebelXT/JPEG) a 0,91 (Escaner CanonScan/JPEG). O teste de proporções indicou diferenças estatisticamente significantes para o Escaner CanonScan, para ambos os formatos, quando comparado aos demais equipamentos e à radiografia convencional. Foi possível concluir que os formatos não influenciaram o desempenho no diagnóstico de cárie para qualquer equipamento usado. O Escaner CanonScan apresentou o pior desempenho e não se mostrou adequado para digitalização de radiografias para o diagnóstico de lesões de cárie. / To evaluate the performance of conventional and indirect digital radiography and the effect of JPEG and TIFF file formats for detecting proximal surface caries. Fifty-six extracted teeth were mounted in blocks and radiographed with Insight conventional film. Films were digitized by two flatbed scanners (Canon CanonScan and Genius ColorScan) and four digital cameras (Canon Powershot G2, Canon Rebel XT, Nikon Coolpix 8700 and Nikon D70), in both JPEG and TIFF file formats. Three observers recorded the presence and depth of caries on 112 proximal surfaces for the conventional and digitized radiographs. They scored lesion presence in a 3-point confidence scale and depth on another 3-point scale. Radiographs were evaluated twice. The observer\'s scores were compared with the results from a histological examination. Intra-observer agreement was moderate to good and interobserver agreement was fair to moderate. Sensitivities varied from 0.24 (Canon RebelXT/JPEG) to 0.53 (Convencional) and specificities from 0.93 (Nikon Coolpix/JPEG; Canon Powershot/TIFF; Canon RebelXT/JPEG e TIFF) to 0.97 (Escaner CanonScan/TIFF e JPEG). The accuracy varied from 0.82 (Canon RebelXT/JPEG) to 0.91 (Escaner CanonScan/JPEG). Images from CanonScan, both JPEG and TIFF, showed significant statistically differences compared with digital radiographs from the others devices and conventional film radiographs. File formats did not influences the detection of proximal caries. Digitizing conventional radiographs using the scanner CanonScan did not produce images of diagnostic quality.

O desenvolvimento tecnológico busca fatores que possibilitem uma menor exposição do paciente e profissional aos raios ionizantes, maior agilidade no atendimento ao paciente e maior auxilio no desenvolvimento do diagnóstico final. Assim a associação entre informática e radiologia básica, traduzidas como radiologia digital, fez-se conseguir os fatores buscados. Uma ferramenta dos sistema digitais é a calibração prévia às medidas. Com essas observações avaliamos tais ferramentas de três sistemas de radiografias digitais, objetivando validar e comprovar sua eficácia nas metodologias dos trabalhos presentes e futuros. Para tanto utilizamos 20 dentes humanos secos provindos do Banco de Dentes Permanentes da FOUSP, os quais foram radiografados com o auxílio de três sistemas de radiografia digital, direto e semidireto (Digora, RVG Trophy e Visualix). O padrão ouro das medidas dos dentes foi a verificação das dimensões dos dentes por meio de um paquímetro digital e posteriormente comparados os seus padrões de medidas com os softwares dos respectivos sistema. Os resultados foram analisados estatisticamente pelo ANOVA (métodos de Tukey e Dunnett), chegando às seguintes conclusões: existe diferença significante entre os aparelhos p<0,001; a média das medidas obtidas houve diferença significante (p<0,001) com o aparelho Digora foi menor que as médias obtidas com os aparelhos das marcas Visualix e RVG; considerando-se o padrão A e o padrão B de calibragem, não existe diferença estatística significante entre os avaliadores; considerando-se a calibragem sem padrão, um dos avaliadores obteve média mais elevada que os demais avaliadores no caso das medidas realizadas com o aparelho Digora, porém este fato não se repetiu quando da utilização dos outros aparelhos; comparando-se as medidas de referência (paquímetro) observa-se que mais uma vez o aparelho Digora (padrão A, padrão B e sem padrão) apresentou diferença significativa; os aparelhos das marcas Visualix e RVG não mostraram diferenças significativas quando comparados com o valor de referência. / Technological development aims at reducing exposure of both the professional and the patient to ionizing radiation, improving patient treatment and aiding final diagnosis. The combination of computer technology and radiology, i.e. digital radiography, has made this possible. One of the tools of digital radiography systems is calibration prior to measurement. The purpose of this study was to assess the calibration tools of three digital radiography systems to verify their efficacy. Twenty human teeth obtained from the Human Teeth Bank of the School of Dentistry, University of São Paulo, were used. The dry teeth were radiographed using three different direct digital radiography systems Digora, RVG Trophy and Visualix. The measurements of the teeth obtained with a digital caliper were considered the gold standard. These measurements were later compared with those obtained using the digital systems. The results were submitted to ANOVA (methods of Tukey and Dunnett), and the following conclusions were drawn: there was a significant difference between the systems used (p < 0.001); the averages of the measurements obtained with the three systems were significantly different (p<0,001). Visualix and RVG were more accurate than Digora considering A and B calibration standards, no significant difference was observed between the examiners; when there was no calibration standard and the Digora system was used, one examiner obtained more high averages measurements than the others; when there was no calibration standard and Visualix or RVG systems were used, no differences were observed between the examiners; the measurements obtained with Digora system (standards A, B or no standard) were statistically different from the gold standard the measurements obtained with Visualix and RVG systems presented no significant difference when compared with the gold standard.