PRATAMA, REHAN DIAS (2026) RANCANG BANGUN SISTEM WEB DETEKSI KARDIOMEGALI PADA CITRA RONTGEN DADA MENGGUNAKAN ARSITEKTUR TIGA LAPIS BERBASIS FASTAPI, LARAVEL, DAN REACT.JS. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (410kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
BAB1.pdf Restricted to Registered users only Download (43kB) |
|
|
Text (BAB II)
BAB2.pdf Restricted to Registered users only Download (132kB) |
|
|
Text (BAB III)
BAB3.pdf Restricted to Registered users only Download (110kB) |
|
|
Text (BAB IV)
BAB4.pdf Restricted to Registered users only Download (341kB) |
|
|
Text (BAB V)
BAB5.pdf Restricted to Registered users only Download (30kB) |
|
|
Text (DAFTAR PUSTAKA)
DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (33kB) |
|
|
Text (LAMPIRAN)
LAMPIRAN.pdf Restricted to Registered users only Download (343kB) |
Abstract
This study aims to design and build a web application system named Thoraks Analytics to automatically and real-time detect Cardiomegaly abnormalities in chest X-Ray images through the integration of an object detection model provided by a FastAPI-based AI microservice into a web system built on a three-tier architecture. The variables examined include chest X-Ray images as the input variable, and bounding box coordinates, class labels, and confidence scores as output variables produced by the two models integrated into the system, namely YOLOv11 and SSD300-VGG16. The dataset consists of bounding-box-annotated chest X-Ray images with two target classes, Cardiomegaly and Normal, sourced from the Roboflow platform. Sampling was conducted using purposive sampling, with the dataset divided into training, validation, and test sets. The system was developed using a three-tier architecture that separates FastAPI as the AI microservice layer responsible for automatic CLAHE preprocessing and model inference, Laravel as the application layer managing user authentication, routing, and detection history storage in the database, and React.js as the presentation layer delivering an interactive user interface. The system analysis method employs functional testing across all features and inter-layer communication flows. Test results indicate that all system functionalities operate correctly, including chest X�Ray image uploading, REST API communication between layers, bounding box overlay display with confidence scores, dynamic model selection, and examination history storage. Thoraks Analytics proves to be an effective initial screening aid (second opinion) for medical personnel, bridging the gap between existing AI models and the need for a web system ready to be operated in healthcare facilities. Keywords: Web System, Three-Tier Architecture, Object Detection, Cardiomegaly, FastAPI, Laravel, React.js, Thoraks Analytics. Penelitian ini bertujuan merancang dan membangun sistem aplikasi web bernama Thoraks Analytics untuk mendeteksi kelainan Kardiomegali pada citra rontgen dada (Chest X-Ray) secara otomatis dan real-time melalui integrasi model object detection yang disediakan oleh layanan AI microservice berbasis FastAPI ke dalam sistem web berbasis arsitektur tiga lapis. Variabel yang diteliti meliputi citra rontgen dada sebagai variabel input, serta koordinat bounding box, label kelas, dan nilai confidence score sebagai variabel output yang dihasilkan oleh dua model yang diintegrasikan dalam sistem, yaitu YOLOv11 dan SSD300-VGG16. Dataset yang digunakan berupa citra rontgen dada beranotasi bounding box dengan dua kelas target yaitu Kardiomegali dan Normal, yang diperoleh dari platform Roboflow. Teknik pengambilan sampel dilakukan secara purposive sampling dengan pembagian dataset menjadi data latih, data validasi, dan data uji. Sistem dibangun menggunakan arsitektur tiga lapis yang memisahkan FastAPI sebagai AI microservice layer untuk menjalankan pra-pemrosesan CLAHE dan inferensi model secara otomatis, Laravel sebagai application layer untuk manajemen autentikasi pengguna, routing, dan penyimpanan riwayat deteksi ke database, serta React.js sebagai presentation layer untuk menyajikan antarmuka pengguna interaktif. Metode analisis sistem menggunakan pengujian fungsionalitas terhadap seluruh fitur dan alur komunikasi antar layer. Hasil pengujian menunjukkan bahwa seluruh fungsionalitas sistem berjalan dengan baik, meliputi unggah citra rontgen, komunikasi REST API antar layer, tampilan bounding box overlay beserta confidence score, pemilihan model secara dinamis, dan penyimpanan riwayat pemeriksaan. Thoraks Analytics terbukti layak sebagai alat bantu skrining awal (second opinion) yang efektif bagi tenaga medis, sekaligus menjembatani kesenjangan antara model AI yang telah ada dengan kebutuhan sistem web yang siap dioperasikan di fasilitas kesehatan. Kata Kunci : Sistem Web, Arsitektur Tiga Lapis, Object Detection, Kardiomegali, FastAPI, Laravel, React.js, Thoraks Analytics.
Actions (login required)
 |
View Item |
