Pengembangan Sistem Neuronavigation Berbasis AR: Registrasi Fitur Wajah Dan Rekonstruksi Objek Medis Volumetrik

Fillah, Muhammad Rafi Insan (2025) Pengembangan Sistem Neuronavigation Berbasis AR: Registrasi Fitur Wajah Dan Rekonstruksi Objek Medis Volumetrik. Other thesis, Institut Teknilogi Sepuluh Nopember.

Text 5025211169-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Perkembangan kraniotomi stereotaktik telah mencapai implementasi AR (Augmented Reality), yang disebut sebagai ARN (Augmented Reality Neuronavigation). ARN memungkinkan rekonstruksi citra medis digital ke dalam format 3D yang lebih intuitif, sehingga membantu tenaga medis mendapatkan gambaran yang lebih jelas dan realistis mengenai struktur anatomi pasien. Namun ARN memiliki beberapa kelemahan, antara lain galat akurasi, Keterbatasan komputasi perangkat standalone AR, dan latensi pada proyeksi lapisan virtual. ARN berbasis PCVR (PC-based Virtual Reality) yang terkoneksi kabel dengan HMD (Head Mounted Device) disertai fitur passthrough dapat menjadi solusi terhadap kelemahan implementasi ARN menggunakan perangkat standalone. Tugas akhir ini memfokuskan pengembangan metode registrasi berbasis fitur wajah dan rekonstruksi volume citra medis digital dengan pendekatan direct volume rendering disertai fitur manipulasi volume dinamis pada aplikasi PCVR. Aplikasi yang dikembangkan menggunakan game engine Unity dengan Meta XR All-in-One SDK dan Unity Volume Rendering sebagai kerangka penyusun. Perangkat keras yang digunakan untuk menjalankan aplikasi adalah Meta Quest 3 yang terkoneksi dengan desktop melalui kabel. Pada aplikasi AR, pengguna mendapat visualisasi volume dari citra medis, melakukan registrasi model volume, dan melakukan interaksi manipulasi volume secara dinamis. Pengujian dilakukan dengan dua jenis pengujian, antara lain pengujian akurasi dan pengujian latensi proyeksi. Dihasilkan aplikasi ARN dengan akurasi proyeksi dengan euclidean offset rata-rata sebesar 2,516 ± 0,119 mm dan modus latensi motion-to-photon sebesar 47,045 ms, menunjukkan sistem ARN yang telah dikembangkan belum memenuhi syarat latensi maksimum 20 ms untuk mencegah terjadinya motion sickness.
===================================================================================================================================
The development of stereotactic craniotomy has progressed to the implementation of Augmented Reality (AR), referred to as Augmented Reality Neuronavigation (ARN). ARN enables the reconstruction of digital medical images into a more intuitive 3D format, thereby assisting medical professionals in obtaining a clearer and more realistic visualization of the patient's anatomical structures. However, ARN presents several limitations, including accuracy errors, computational constraints of standalone AR devices, and latency in virtual layer projection. A PCVR (PC-based Virtual Reality) implementation of ARN, connected via cable to a Head-Mounted Display (HMD) and equipped with a passthrough feature, may offer a solution to the limitations posed by standalone devices. This thesis focuses on the development of a facial feature-based registration method and volumetric reconstruction of digital medical images using a direct volume rendering approach, along with dynamic volume manipulation features in a PCVR application. The application is developed using the Unity game engine, leveraging the Meta XR All-in-One SDK and Unity Volume Rendering framework. The hardware used to run the application is Meta Quest 3, connected to a desktop via cable. In the AR application, users are able to visualize volumetric medical images, perform volume model registration, and dynamically manipulate the volume. Two types of evaluations were conducted: projection accuracy testing and latency testing. The resulting ARN application achieved a projection accuracy with an average euclidean offset of 2,516 ± 0,119 mm and a motion-to-photon latency mode of 47,045 ms, indicating that the developed ARN system has not yet met the maximum latency threshold of 20 ms required to prevent motion sickness.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	PC-based Virtual Reality (PCVR), Passthrough, Direct Volume Rendering (DVR), Unity
Subjects:	R Medicine > RC Internal medicine > RC78.7.N83 Magnetic resonance imaging. T Technology > T Technology (General) > T59.7 Human-machine systems.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Rafi Insan Fillah
Date Deposited:	25 Jul 2025 02:14
Last Modified:	25 Jul 2025 02:14
URI:	http://repository.its.ac.id/id/eprint/121273

Actions (login required)

View Item