Iterated Region for Interactive Image Segmentation on Dental Panoramic Radiograph

Image Segmentation is a process to separate between foreground and background. Segmentation process in low contrast image such as dental panoramic radiograph image is not easily determined. Image segmentation accuracy determines the success or failure of the final analysis process. The process of segmentation can occur ambiguity. This ambiguity is due to an ambiguous area if it is not selected as a region so it may have occurred cluster errors. To solve this ambiguity, we proposed a new region merging by iterated region merging process on dental panoramic radiograph image. The proposed method starts from the user marking and works iteratively to label the surrounding regions. In each iteration, the minimal gray-levels value is merged so the unknown regions significantly reduced. This experiment shows that the proposed method is effective with an average of ME and RAE of 0.04% and 0.06%.

Download Full-text

Interactive Image Segmentation using Neighborhood Spatial Information and Statistical Grey Level on Dental Panoramic Radiograph

Jurnal Ilmu Komputer dan Informasi ◽

10.21609/jiki.v12i1.622 ◽

2019 ◽

Vol 12 (1) ◽

pp. 19

Author(s):

Shabrina Choirunnisa ◽

Ari Firmanto ◽

Agus Zaenal

Keyword(s):

Image Segmentation ◽

Spatial Information ◽

Panoramic Radiograph ◽

Spatial Distance ◽

Segmentation Result ◽

Grey Level ◽

Low Contrast ◽

Intensity Information ◽

Dental Panoramic Radiograph ◽

Novel Strategy

In dental panoramic radiographs, grey-level intensity information has been widely used for object segmentation in digital image. However, low contrast in the radiograph image causes high ambiguity that can cause the inconsistency of classification result. Since the grey-level intensity of background and object is almost similar, so in order to improve the segmentation result, the spatial distance on neighborhod region is applied. In this paper, we proposed a novel strategy to measure the distance using neighborhod spatial information and statistical grey level technique for image segmentation. The proposed method starts by calculating adjacency matrix and measured spatial distance on neighborhood region. Since the value of both distances are not in the same range, then the normalization is needed. The distances merging is approached by tuning the weight using several constant values. The experiment results show that our proposed merging strategy has better segmentation result based on Relative Foreground Area Error value.

Download Full-text

Segmentasi Tulang Kortikal pada Citra Dental Panoramic Radiograph

IJEIS (Indonesian Journal of Electronics and Instrumentation Systems) ◽

10.22146/ijeis.10769 ◽

2016 ◽

Vol 6 (1) ◽

pp. 37

Author(s):

Thohiroh Agus Kumala ◽

Agus Harjoko

Keyword(s):

Cortical Bone ◽

Active Contour ◽

Operating Characteristic ◽

Panoramic Radiograph ◽

Contour Method ◽

Initial Curve ◽

Active Contour Method ◽

Dental Panoramic Radiograph ◽

Contrast Stretching ◽

Radiograph Image

AbstrakPengolahan citra dalam dunia medis sudah banyak dikembangkan. Satu tahapan yang penting dalam pengolahan citra ini yaitu segmentasi. Ketepatan dalam menentukan diagnosis suatu penyakit ditentukan oleh ketepatan tahap segmentasi.Penelitian ini menggunakan citra dental panoramic radiograph dengan ukuran 2000x1000 piksel. Daerah sampel tulang kortikal diambil dari tulang kortikal rahang bawah bagian kanan dan kiri sekitar foramen mentalis dengan ukuran 128x128 piksel. Untuk mempermudah proses segmentasi maka dilakukan prapengolahan terhadap citra yaitu dengan contrast stretching dan grayscaling. Selanjutnya citra hasil prapengolahan dilakukan segmentasi menggunakan metode active contour. Metode ini diawali dengan pembentukan pembentukan mask sebagai kurva awal, dari kurva awal ini kemudian kurva akan bergerak keluar atau kedalam sesuai dengan bentuk tepi dari tulang kortikal. Ujicoba dilakukan dengan menggunakan metode ROC (Receiver Operating Characteristic). Segmentasi dari 21 data citra dental panoramic radiograph menggunakan metode Active Contour dapat melakukan segmentasi tulang kortikal kanan dengan prosentase akurasi 90,67%, sensitifitas 90,14% dan spesifitas 91,55%. Tulang kortikal kiri dengan prosentase akurasi 89,37%, sensitifitas 86,59% dan spesifitas 91,23%. Kata kunci— active contour, dental panoramic radiograph, tulang kortikal, segmentasi citra AbstractImage processing in the medical world has been developed. One important stage in the processing of this image is segmentation. The accuracy in determining the diagnosis of a disease is determined by the accuracy of the segmentation stage.This study used a dental panoramic radiograph image with the size of 2000x1000 pixels. The area of cortical bone samples taken from the cortical bone of the lower jaw right and left about the mental foramen with 128x128 pixels. To simplify the process of segmentation is carried out preprocessing on the image that is by contrast stretching and grayscale. Furthermore, image segmentation results of preprocessing conducted using active contour method. This method begins with the formation of the formation of the mask as the initial curve, from the initial curve is then the curve will move in or out according to the shape of the edge of the cortical bone.Tests performed using the ROC (Receiver Operating Characteristic). Segmentation of 21 dental panoramic radiograph image data using Active Contour method can perform with the right cortical bone segmentation accuracy percentage of 90.67%, 90.14% sensitivity and 91.55% specificity. Cortical bone is left with an accuracy percentage of 89.37%, 86.59% sensitivity and 91.23% specificity. Keywords— active contour, dental panoramic radiograph, cortical bone, image segmentation

Download Full-text

Identification of molar and premolar teeth in dental panoramic radiograph image

2017 Second International Conference on Informatics and Computing (ICIC) ◽

10.1109/iac.2017.8280614 ◽

2017 ◽

Cited By ~ 1

Author(s):

Romi Fadillah Rahmat ◽

Silviani Silviani ◽

Erna Budhiarti Nababan ◽

Opim Salim Sitompul ◽

Rina Anugrahwaty ◽

...

Keyword(s):

Panoramic Radiograph ◽

Dental Panoramic Radiograph ◽

Radiograph Image

Download Full-text

Multi-Task Deep Learning Model for Classification of Dental Implant Brand and Treatment Stage Using Dental Panoramic Radiograph Images

Biomolecules ◽

10.3390/biom11060815 ◽

2021 ◽

Vol 11 (6) ◽

pp. 815

Author(s):

Shintaro Sukegawa ◽

Kazumasa Yoshii ◽

Takeshi Hara ◽

Tamamo Matsuyama ◽

Katsusuke Yamashita ◽

...

Keyword(s):

Deep Learning ◽

Dental Implant ◽

Area Under The Curve ◽

Treatment Phase ◽

Panoramic Radiograph ◽

Radiographic Images ◽

Treatment Stage ◽

Prefectural Central Hospital ◽

Dental Panoramic Radiograph

It is necessary to accurately identify dental implant brands and the stage of treatment to ensure efficient care. Thus, the purpose of this study was to use multi-task deep learning to investigate a classifier that categorizes implant brands and treatment stages from dental panoramic radiographic images. For objective labeling, 9767 dental implant images of 12 implant brands and treatment stages were obtained from the digital panoramic radiographs of patients who underwent procedures at Kagawa Prefectural Central Hospital, Japan, between 2005 and 2020. Five deep convolutional neural network (CNN) models (ResNet18, 34, 50, 101 and 152) were evaluated. The accuracy, precision, recall, specificity, F1 score, and area under the curve score were calculated for each CNN. We also compared the multi-task and single-task accuracies of brand classification and implant treatment stage classification. Our analysis revealed that the larger the number of parameters and the deeper the network, the better the performance for both classifications. Multi-tasking significantly improved brand classification on all performance indicators, except recall, and significantly improved all metrics in treatment phase classification. Using CNNs conferred high validity in the classification of dental implant brands and treatment stages. Furthermore, multi-task learning facilitated analysis accuracy.

Download Full-text

Fast Pixel-Superpixel Region Merging for SAR Image Segmentation

IEEE Transactions on Geoscience and Remote Sensing ◽

10.1109/tgrs.2020.3041281 ◽

2020 ◽

pp. 1-17

Author(s):

Deliang Xiang ◽

Fan Zhang ◽

Wei Zhang ◽

Tao Tang ◽

Dongdong Guan ◽

...

Keyword(s):

Image Segmentation ◽

Sar Image ◽

Region Merging ◽

Sar Image Segmentation

Download Full-text

FAST INTERACTIVE REGIONAL PATTERN MERGING FOR GENERIC TISSUE SEGMENTATION IN HISTOPATHOLOGY IMAGES

Biomedical Engineering Applications Basis and Communications ◽

10.4015/s1016237221500125 ◽

2021 ◽

pp. 2150012

Author(s):

Kuo-Lung Lor ◽

Chung-Ming Chen

Keyword(s):

Image Segmentation ◽

Color Image ◽

Linear Time ◽

Mean Shift ◽

Texture Features ◽

Color Image Segmentation ◽

Image Size ◽

Region Merging ◽

Tissue Segmentation ◽

Regional Pattern

The image segmentation of histopathological tissue images has always been a challenge due to the overlapping of tissue color distributions, the complexity of extracellular texture and the large image size. In this paper, we introduce a new region-merging algorithm, namely, the Regional Pattern Merging (RPM) for interactive color image segmentation and annotation, by efficiently retrieving and applying the user’s prior knowledge of stroke-based interaction. Low-level color/texture features of each region are used to compose a regional pattern adapted to differentiating a foreground object from the background scene. This iterative region-merging is based on a modified Region Adjacency Graph (RAG) model built from initial segmented results of the mean shift to speed up the merging process. The foreground region of interest (ROI) is segmented by the reduction of the background region and discrimination of uncertain regions. We then compare our method against state-of-the-art interactive image segmentation algorithms in both natural images and histological images. Taking into account the homogeneity of both color and texture, the resulting semi-supervised classification and interactive segmentation capture histological structures more completely than other intensity or color-based methods. Experimental results show that the merging of the RAG model runs in a linear time according to the number of graph edges, which is essentially faster than both traditional graph-based and region-based methods.

Download Full-text