Sukmanawati, Andri (2017) Klasifikasi Respon Otot Lengan Bawah Pada Penderita Stroke Berdasarkan Sinyal EMG Menggunakan Metode Naive Bayes Classifier. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
2913100007-undergraduate-thesis.pdf - Published Version Download (14MB) | Preview |
Abstract
Saat ini banyak pengembangan wearable device menggunakan variasi pola aktifasi otot sebagai media kontrol, sehingga perkembangan ini dapat dimanfaatkan untuk membantu manusia yang memiliki keterbatasan gerak. Pada Tugas Akhir ini, dilakukan proses klasifikasi terhadap ekstraksi fitur-fitur dari sinyal EMG menggunakan metode Naive Bayes Classifier untuk mengontrol electric device. Ekstraksi fitur yang digunakan adalah Time-domain. Fitur-fitur yang dipilih yaitu MAV, IEMG, WL dan VAR. Pengukuran sinyal EMG pada respon otot lengan bawah (forearm) pasien stroke dilakukan dengan bantuan Myo Armband yang menggunakan 8 channel input. Pengujian dilakukan terhadap 4 subjek dengan rincian subjek 1 dan 2 merupakan reponden stroke, sedangkan subjek 3 dan 4 merupakan responden normal. Gerakan yang diteliti pada penelitian ini adalah gerakan hand-close, hand-open, jempol keatas, lima jari lurus kedepan (dengan posisi telapak tangan merapat), wave-right, kelingking, kelingking-jempol, kelingking-telunjuk, telunjuk, telunjuk jempol, tiga-jari. Masing-masing responden melakukan 11 variasi gerakan, setiap gerakan dilakukan sebanyak sepuluh kali. Hasil akurasi dari pengujian klasifikasi dua gerakan diperoleh 100 % untuk subjek 1, subjek 3 dan subjek 4. Sedangkan subjek 2 memperoleh 83,35 %. Pada pengujian klasifikasi lima gerakan hasil akurasi tertinggi pada responden stroke diperoleh pada subjek 1 dengan nilai akurasi 86,68 %, sedangkan pada responden normal diperoleh pada subjek 4 dengan nilai akurasi 100 %. Pada pengujian klasifikasi 11 gerakan hasil akurasi tertinggi pada responden stroke diperoleh pada subjek 1 dengan nilai akurasi 63,63 %, sedangkan pada responden normal diperoleh pada subjek 4 dengan nilai akurasi 87,87 %.
=========================================================
Nowdays, there are many developments of wearable devices that use pattern variation of muscle activation as media control, so that these developments can be used to help people with limited mobility. In this final project, a classification process of EMG signal features extraction is done using Naive Bayes Classifier method to control electric device. This project uses Time-domain feature.The chosen features are MAV, IEMG, WL and VAR. Measurment of EMG signals in response to the forearm muscles was done using Myo Armband with 8 channels as the input. There were be 4 subjects included in this project, in which subject 1 and 2 are respondents with stroke disease, while subject 3 and 4 are normal respondents. The movements which were analyzed are hand-close, hand-open, thumbs up, five fingers straight (with palms firmly held), wave-right, pinkie, pinkie-thumb, pinkie-index finger, index finger, and fingers three. Each respondent does eleven kinds of movements, each of which was done ten times. The accuracy of two-movement classification of subject 1, 3, and 4 were 100 %, while subject 2 obtained 83,35 %. From five-movement classification test, the highest result obtained from respondent with stroke (subject 1) with accuracy of 86,68 %, while the from the normal respondent, subject 4 got the highest accuracy of 100 %. From eleven-movement classification test, the highest result obtained from respondent with stroke (subject 1) with accuracy of 63,63 %, while the from the normal respondent, subject 4 got the highest accuracy of 87,87 %.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSKom 616.74 Suk k |
| Uncontrolled Keywords: | Electromyography, Naive Bayes Classifier, Feature Extraction. |
| Subjects: | Q Science > QA Mathematics > QA76 Computer software Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) |
| Divisions: | Faculty of Industrial Technology > Multimedia and Network Engineering > (S1) Undergraduate Theses |
| Depositing User: | Andri Sukmanawati . |
| Date Deposited: | 11 Jan 2018 08:13 |
| Last Modified: | 05 Mar 2019 03:08 |
| URI: | http://repository.its.ac.id/id/eprint/48685 |
Actions (login required)
 |
View Item |