Sam`ani, Muhammad Yasir (2025) Rancang Bangun 3D Printed Orthosis Dengan Sistem Kontrol Gerakan Berbasis Motor Imagery (MI) Untuk Pasien Brachial Plexus Injury (BPI). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5023211067-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5023211067-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy |
Abstract
Disfungsi motorik ekstremitas atas, khususnya akibat Brachial Plexus Injury (BPI), berdampak signifikan terhadap kemandirian pasien dan kualitas hidup. Meskipun solusi bedah dan rehabilitasi konvensional tersedia, metode tersebut seringkali memiliki risiko komplikasi dan waktu pemulihan yang panjang. Sebagai alternatif, penelitian ini merancang dan mengimplementasikan 3D printed orthosis dengan sistem kontrol gerakan berbasis Motor Imagery (MI) sebagai alternatif rehabilitasi non-invasif. Sistem ini mengakuisisi sinyal Electroencephalogram (EEG) melalui modul EEG Click dan mikrokontroler STM32 Nucleo F411RE. Sinyal yang diakuisisi diproses secara digital melalui BPF 8-30 Hz, notch filter 50 Hz, dan normalisasi Z-score untuk standarisasi. Fitur-fitur diekstraksi menggunakan Filter Bank Common Spatial Pattern (FBCSP) dan Event Related Desynchronization/Event Related Synchronization (ERD/ERS), yang menunjukkan potensi diskriminatif. Model Random Forest (RF) digunakan untuk klasifikasi, dan output-nya secara otomatis menggerakkan servo DS3245 pada orthosis untuk menghasilkan gerakan menyerupai fleksi dan ekstensi siku (0˚-90˚). Pengujian sistem keseluruhan memverifikasi komunikasi dua arah sekuensial yang berhasil antara mikrokontroler dan PC. Namun, akurasi model RF dalam klasifikasi MI masih 47.33%, yang mengindikasikan bahwa model belum optimal dan fitur mungkin kurang robust. Meskipun demikian, hasil pengujian sistem keseluruhan pada subjek normal menunjukkan akurasi hingga 70% (Kanan) dan 80% (Kiri), sementara pada subjek BPI, akurasi untuk tangan kanan yang disfungsi hanya 10% berbanding 80% untuk tangan kiri yang berfungsi. Disparitas ini memperkuat potensi perangkat sebagai alat rehabilitasi untuk menstimulasi neuroplastisitas pasien BPI. Penelitian ini menyimpulkan bahwa sistem kontrol gerakan orthosis berbasis MI memiliki potensi besar, namun memerlukan optimasi lebih lanjut pada ekstraksi fitur, klasifikasi, perangkat keras, dataset, dan feedback loop otomatis untuk kontrol yang adaptif dan efektif.
====================================================================================================================================================================================
Upper extremity motor dysfunction, particularly due to BPI, significantly impacts patient independence and quality of life. Although conventional surgical and rehabilitation solutions are available, these methods often carry risks of complications and prolonged recovery times. As an alternative, this research designs and implement a 3D printed orthosis with a Motor Imagery (MI)-based motion control system as a non-invasive rehabilitation alternative. This system acquires Electroencephalogram (EEG) signals via an EEG Click module and an STM32 Nucleo F411RE microcontroller. The acquired signals are digitally processed through an 8-30 Hz BPF, a 50 Hz notch filter, and Z-score normalization for standardization. Features are extracted using Filter Bank Common Spatial Pattern (FBCSP) and Event Related Desynchronization/Event Related Synchronization (ERD/ERS), which demonstrate discriminative potential. A Random Forest (RF) model is then used for classification, and its output automatically drives a DS3245 servo on the orthosis to produce movements resembling elbow flexion and extension (0˚-90˚). Overall system testing verified successful sequential bidirectional communication between the microcontroller and the PC. However, the RF model's accuracy in MI classification is still 47.33%, indicating that the model is not yet optimal and features may lack robustness. Nevertheless, overall system test results on normal subjects showed accuracies up to 70% (Right) and 80% (Left), while for the BPI subject, accuracy for the dysfunctional right hand was only 10% compared to 80% for the functional left hand. This disparity reinforces the device's potential as a rehabilitation tool to stimulate neuroplasticity in BPI patients. This research concludes that the MI-based orthosis motion control system holds significant potential, but requires further optimization in feature extraction, classification, hardware, dataset, and automatic feedback loop for adaptive and effective control.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Brachial Plexus Injury (BPI), Orthosis, 3D print, Electroencephalogram (EEG), Motor Imagery (MI) |
| Subjects: | Q Science > QP Physiology > Q376.5 Electroencephalography (EEG) R Medicine > RC Internal medicine > RC386.5 Electroencephalography. R Medicine > RD Surgery > RD130 Artificial organs; Prosthesis R Medicine > RD Surgery > RD558 Elbow R Medicine > RM Therapeutics. Pharmacology > RM950 Rehabilitation technology. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK1007 Electric power systems control |
| Divisions: | Faculty of Electrical Technology > Biomedical Engineering > 11410-(S1) Undergraduate Thesis |
| Depositing User: | Muhammad Yasir Sam'ani |
| Date Deposited: | 04 Aug 2025 07:28 |
| Last Modified: | 04 Aug 2025 07:28 |
| URI: | http://repository.its.ac.id/id/eprint/126821 |
Actions (login required)
 |
View Item |