Hidayat, Mohammad Anwar (2018) Segmentasi Cortical Bone pada Citra Dental Panoramic Radiograph Menggunakan Dynamic Multi-scale Line Detection. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
05111440000137-Undergraduate_Thesis.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Dental panoramic radiograph adalah citra x-ray 2D dari rahang mulut seseorang. Selain informasi kondisi dari gigi, citra dental panoramic radiograph dapat digunakan sebagai diagnosis awal terhadap kemungkinan penyakit osteoporosis melalui ketebalan cortical bone. Dengan bantuan komputer, penghitungan ketebalan cortical bone dapat dilakukan dengan otomatis. Untuk menghitung ketebalan cortical bone, komputer harus memilih dan membedakan area cortical bone dari background. Dikarenakan komputer tidak dapat melakukan hal tersebut secara otomatis, dibutuhkan teknik segmentasi untuk memisahkan kedua area ini.
Beberapa teknik segmentasi telah dikembangkan. Salah satu teknik yang digunakan untuk segmentasi cortical bone adalah dengan menghitung fitur line strength menggunakan teknik line detection. Seperti teknik segmentasi lainnya, pengembangan teknik ini biasanya optimal untuk ukuran citra tertentu saja. Padahal, citra pada penerapan medis dapat memiliki ukuran yang berbeda. Oleh karena itu, dibutuhkan teknik segmentasi yang dapat secara dinamis menyesuaikan terhadap kondisi tersebut.
Tugas akhir ini mengusulkan metode segmentasi cortical bone menggunakan dynamic multi-scale line detection. Secara umum, sistem dibagi menjadi 5 (lima) tahap. Tahap pertama yaitu pemilihan ROI. Tahap kedua yaitu perhitungan parameter. Tahap ketiga yaitu perhitungan line strength. Tahap keempat yaitu segmentasi. Dan tahap kelima yaitu perbaikan citra. Uji coba pada tugas akhir ini dilakukan pada 30 citra grayscale dari dental panoramic radiograph. Hasil uji coba menunjukkan bahwa sistem mampu melakukan segmentasi pada citra dengan ukuran berbeda dengan nilai standard deviasi F1 score di bawah 5% atau 0,05. Hasil uji coba juga menunjukkan nilai ratarata F1 score adalah 0,890.
========================================================================================================
Dental panoramic radiograph is 2D x-ray image that captures entire structure of mouth. Beside information about the condition of teeth, dental panoramic radiograph image can be used for early diagnosis of the possibility of osteoporosis through cortical bone width. With help of computer, calculation of cortical bone width can be done automatically. To perform the calculation, computer has to distinct cortical bone from background. Because computer cannot do it automatically as it is, it requires segmentation techni-que to distinct both areas.
Several segmentation techniques have been proposed. One that perform segmentation for cortical bone is through calculating line strength features using line detection technique. Like any other technique, the development of this technique is usually optimal on images with fixed size. Whereas actual medical image can vary in size. Hence the requirement of segmentation technique that can dynamically adapt to this condition.
This research proposes segmentation technique of cortical bone using dynamic multi-scale line detection. In general, the system is divided into 5 (five) stages. First stage is choosing Region of Interest (ROI). Second stage is calculating parameter. Third stage is calculating line strength. Fourth stage is segmentation. And fifth stage is image refinement.
Experiments in this research are done on 30 grayscale images of dental panoramic radiograph. The results of experiment show that the system is capable of segmenting images with different sizes with standard deviation of F1 score with value under 5% or 0.05. It also shows the average F1 score is 0.890.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSIf 006.425 Hid s-1 3100018076690 |
| Uncontrolled Keywords: | Cortical Bone, Dental Panoramic Radiograph, Dynamic, Line Detection, Multi-scale |
| Subjects: | Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science) Q Science > QA Mathematics > QA9.58 Algorithms T Technology > T Technology (General) > T57.5 Data Processing T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques |
| Divisions: | Faculty of Information and Communication Technology > Informatics > 55201-(S1) Undergraduate Thesis |
| Depositing User: | Mohammad Anwar Hidayat |
| Date Deposited: | 18 Jan 2019 03:03 |
| Last Modified: | 08 Feb 2021 07:14 |
| URI: | http://repository.its.ac.id/id/eprint/58342 |
Actions (login required)
 |
View Item |