Aryatama, Muhammad Dimas Nugraha (2020) Deteksi Pneumothorax Pada Citra X-Ray Menggunakan Convolutional Neural Network. Other thesis, InstitutTeknologi Sepuluh Nopember.
Preview |
Text
07211640000008-Undergraduate_Thesis.pdf Download (5MB) | Preview |
Abstract
Pneumothorax adalah salah satu kondisi paru-paru dimana terkumpulnya udara yang pada rongga pleura, yaitu rongga tipis dibatasi dua selaput pleura diantara paru-paru dan dinding dada. Kondisi Pneumothorax termasuk kedalam kategori kondisi kritis pada paruparu yang memerlukan penanganan medis dari dokter atau ahli medis lain degan cepat. Jika tidak ditangani dengan cepat maka dapat
menyebabkan komplikasi hingga kematian. Maka dari itu deteksi dini dari kondisi Pneumothorax adalah suatu hal yang perlu di prioritaskan. Permasalahan yang sering terjadi saat ini adalah perbedaan analisis/penafsiran gambar medis antar dokter. Sementara metode diagnosis saat ini masih bersifat manual yaitu ahli radiologi perlu mengecek gambar secara langsung dengan bantuan Computer Aided Detection and Diagnosis (CAD). Dikarenakan tingkat
akurasi prediksi dari CAD masih belum signifikan maka diperlukan teknologi Deep Learning. Salah satunya metode yang digunakan untuk melakukan proses training deteksi adalah menggunakan Convolutional Neural Network. Untuk menjawab atas permasalahan tersebut, maka pada tugas akhir ini akan dikembangkan sebuah sistem yang menggunakan CNN agar dapat melakukan pendeteksian kondisi Pneumothorax, sehingga dapat membantu menegaskan
diagnosis yang dilakukan oleh dokter. Data citra yang digunakan adalah Dataset NIH Chest X-ray yang berjumlah 112.120 citra dan dipublikasikan oleh National Institutes of Health. Dataset tersebut terbagi menjadi tiga belas jenis kelas penyakit paru-paru.
===========================================================================================
Pneumothorax is a condition of the lungs where air is collected in the pleural cavity, which is a thin cavity bounded by two pleural membranes between the lungs and the chest wall. Pneumothorax condition is included in the category of critical conditions in the lungs that require medical treatment from a doctor or other medical expert quickly. If not treated quickly it can cause complications to death. Therefore early detection of the condition of pneumothorax is something that needs to be prioritized. The problem that often occurs at this time is the difference in analysis / interpretation
of medical images between doctors. While the diagnostic method is still manual, radiologists need to check the images directly with the help of Computer Aided Detection and Diagnosis (CAD). Because the level of prediction accuracy of CAD is still not significant, Deep Learning technology is needed. One of the methods used to carry out the classification training process is to use Convolutional Neural Network. To answer this problem, this final project will develop a system that uses CNN to be able to classify the condition of pneumothorax, so that it can help confirm the diagnosis made by a doctor. The image data used is the NIH Chest X-ray dataset, totaling 112,120 images and published by National Institutes of Health. The dataset is divided into thirteen types of lung disease class.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Pneumothorax, Deep Learning, Convolutional Neural Network |
| Subjects: | Q Science > Q Science (General) > Q337.5 Pattern recognition systems R Medicine > RB Pathology |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis |
| Depositing User: | Muhammad Dimas Nugraha Aryatama |
| Date Deposited: | 19 Aug 2020 02:55 |
| Last Modified: | 23 Jun 2023 07:57 |
| URI: | http://repository.its.ac.id/id/eprint/78932 |
Actions (login required)
 |
View Item |