Nainggolan, Maya Maria (2024) Seleksi Model Terbaik untuk Analisis Survival dengan Antarmuka Grafis berbasis Python (Studi Kasus: Data Survival untuk Evaluasi Terapi Komplementer di RS Nur Hidayah Bantul). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5003201028_Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5003201028_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (6MB) |
Abstract
Seiring dengan berkembangnya pengobatan medis, semakin beragam pula pilihan pengobatan salah satunya telah berkembang pengobatan komplementer dan alternatif. Pengobatan komplementer dan alternatif atau complementary and alternative medicine (CAM) adalah pengobatan yang mencakup beragam pendekatan mulai dari pendekatan kuno hingga modern yang dimaksudkan untuk mencegah maupun mengobati penyakit yang dapat membantu perbaikan kondisi pasien medis dimana analisis survival dapat diaplikasikan. Analisis survival digunakan untuk mempelajari waktu perbaikan kondisi mental pasien terapi komplementer dengan menggunakan pustaka lifelines dari Python. Data yang digunakan adalah data sekunder dari rekam medis 100 pasien terapi komplementer Al-Quran di Rumah Sakit Nur Hidayah Bantul. Perbaikan kondisi mental pasien dianalisis menggunakan variabel Sikap (X1), Gangguan Ibadah (X2), Riwayat Interaksi (X3), Jenis Kelamin (X4), Usia (X5), Status Perkawinan (X6), Pendidikan (X7), Keluhan Utawa (X8), Riwayat Penyakit Kronis (X9) dan Reaksi Terapi (X10). Kaplan-Meier dan uji log-rank menunjukkan variabel Sikap (X1), Gangguan Ibadah (X2), Riwayat Interaksi (X3) dan Reaksi Terapi (X10) memiliki perbedaan kurva survival antara kategori masing-masing variabel. Variabel Riwayat Interaksi tidak memenuhi asumsi proportional hazard sehingga model Cox yang digunakan adalah stratified Cox dan time-dependent Cox fungsi waktu. Evaluasi model menunjukkan model time-dependent Cox fungsi waktu memiliki nilai ukuran kebaikan paling baik dengan nilai C-index 0,92 dan rata-rata Dynamic-AUC 0,84. Analisis survival yang dianalisis menggunakan Python, dapat dikembangkan menjadi antarmuka grafis menggunakan framework Streamlit. Antarmuka grafis ini dinamakan Sur.M, bertujuan untuk mengatasi kompleksitas penerapan analisis survival, terutama bagi peneliti dengan keterbatasan pemahaman pemrograman. Sur.M yang dikembangkan masih memiliki keterbatasan dalam faktorisasi data kategorik dalam pemodelan Cox proportional hazard namun, sudah berhasil mengakomodasi dengan baik proses loading data, visualisasi serta interpretasi untuk kurva Kaplan-Meier dan uji log-rank tiap variabel kategorik. Sur.M diharapkan memberikan kontribusi terhadap metode analisis survival yang lebih mudah diakses, terutama dalam studi kasus terapi komplementer Al-Quran di Rumah Sakit Nur Hidayah Bantul.
========================================================================================================================
As the development of medical treatment, there are also a variety of treatment options, one of which has developed complementary and alternative medicine. Complementary and alternative medicine (CAM) is a treatment that includes a variety of approaches ranging from ancient to modern approaches intended to prevent or treat diseases that can help improve the condition of medical patients where survival analysis can be applied. Survival analysis is used to study the time to improvement of mental state of complementary therapy patients using the lifelines library of Python. The data used is secondary data from the medical records of 100 patients of complementary Al-Quran therapy at Nur Hidayah Bantul Hospital. The improvement of patients' mental state was analyzed using variables of Attitude (X1), Worship Disorder (X2), Interaction History (X3), Gender (X4), Age (X5), Marital Status (X6), Education (X7), Chief Complaint (X8), History of Chronic Disease (X9) and Therapy Reaction (X10). Kaplan-Meier and log-rank tests showed the variables of Attitude (X1), Worship Disorder (X2), History of Interaction (X3) and Therapeutic Reaction (X10) had different survival curves between the categories of each variable. The Interaction History variable did not fulfil the proportional hazard assumption so that the Cox model used was stratified Cox and time-dependent Cox time function. Model evaluation showed the time-dependent Cox model had the best goodness of fit with a C-Index value of 0.92 and an average Dynamic-AUC of 0.84. Survival analysis, which is analyzed using Python, can be developed into a graphical interface using the Streamlit framework. This graphical interface, called Sur.M, aims to overcome the complexity of applying survival analysis, especially for researchers with limited understanding of programming. Sur.M still has limitations in factorizing categorical data in Cox proportional hazard modelling, however, it has successfully accommodated the data loading process, visualization and interpretation for Kaplan-Meier curves and log-rank test for each categorical variable. Sur.M is expected to contribute to a more accessible survival analysis method, especially in the case study of Al-Quran complementary therapy at Nur Hidayah Bantul Hospital.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Analisis Survival, Antarmuka Grafis, Evaluasi Model, Regresi Cox, Streamlit Survival Analysis, Graphic Interface, Model Evaluation, Cox Regression, Streamlit |
| Subjects: | Q Science > QA Mathematics > QA76.76.A63 Application program interfaces R Medicine > R Medicine (General) > R853.S7 Survival analysis (Biometry) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Maya Maria Nainggolan |
| Date Deposited: | 10 Aug 2024 06:22 |
| Last Modified: | 10 Aug 2024 06:22 |
| URI: | http://repository.its.ac.id/id/eprint/114995 |
Actions (login required)
 |
View Item |