Comparative study of a novel machine learning augmented mobile acoustic uroflowmetry and conventional uroflowmetry

Rangganata, Ervandy (2024) Comparative study of a novel machine learning augmented mobile acoustic uroflowmetry and conventional uroflowmetry. Masters thesis, Universitas Pelita Harapan.

[thumbnail of Title] Text (Title)
cover - Ervandy Rangganata.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (90kB)
[thumbnail of Abstract] Text (Abstract)
abstract - Ervandy Rangganata-10.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (103kB)
[thumbnail of ToC] Text (ToC)
toc - Ervandy Rangganata-9.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (135kB)
[thumbnail of Chapter 1] Text (Chapter 1)
chapter 1 - Ervandy Rangganata-8.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (114kB)
[thumbnail of Chapter 2] Text (Chapter 2)
chapter 2 - Ervandy Rangganata-7.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (516kB)
[thumbnail of Chapter 3] Text (Chapter 3)
chapter 3 - Ervandy Rangganata-6.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (147kB)
[thumbnail of Chapter 4] Text (Chapter 4)
chapter 4 - Ervandy Rangganata-5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (824kB)
[thumbnail of Chapter 5] Text (Chapter 5)
chapter 5 - Ervandy Rangganata-4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (30kB)
[thumbnail of Bibliography] Text (Bibliography)
bibliography - Ervandy Rangganata-3.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (115kB)
[thumbnail of Appendices] Text (Appendices)
appendices - Ervandy Rangganata-2.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (871kB)

Abstract

Uroflowmetri adalah pengukuran volume urin non-invasif yang dikeluarkan dari waktu ke waktu. Uroflowmetri konvensional telah menjadi modalitas utama pengukuran aliran urin dalam beberapa waktu. Namun, metode ini mengharuskan pasien untuk hadir di rumah sakit atau tempat perawatan kesehatan beberapa kali, terkadang membuat pasien merasa tidak nyaman menjalani pemeriksaan dan membuat profil berkemih harian yang tidak akurat. Uroflowmetri akustik (sonouroflowmetri) telah diusulkan sebagai metode alternatif pengukuran aliran urin karena portabilitasnya. Penelitian ini bertujuan untuk mengevaluasi akurasi dan reliabilitas sono-uroflowmetri dibandingkan dengan uroflowmetri konvensional. Teknik pembelajaran mendalam untuk analisis gambar memungkinkan pembuatan alat analisis data eksplorasi yang mudah digunakan. Menggunakan toolbox pemrograman visual Orange (http://orange.biolab.si), studi ini menyederhanakan analisis gambar dengan menggabungkan deep learning, proses pembelajaran mesin, dan visualisasi data. Orange memfasilitasi pengembangan alur kerja untuk analisis data dengan menyusun komponen untuk pretreatment data, visualisasi, dan pemodelan. Studi ini menggunakan Orange dengan komponen yang memprofilkan foto dengan vektor fitur menggunakan jaringan konvolusional dalam yang telah dilatih sebelumnya. Vektor ini digunakan dalam pengelompokan dan klasifikasi citra di dalam kerangka kerja yang memfasilitasi penambangan koleksi citra oleh pengguna pemula dan berpengalaman. Volume urin, durasi pengosongan, laju aliran maksimum, dan laju aliran rata-rata diidentifikasi dan digunakan untuk menentukan hasil pengukuran. Sonouroflowmetri menunjukkan korelasi yang signifikan dibandingkan dengan uroflowmetri konvensional. Oleh karena itu dapat digunakan sebagai alternatif untuk uroflowmetri konvensional. Studi ini mendemonstrasikan kegunaan alat ini dalam analisis citra rekaman aliran urin dalam membedakan pola berkemih normal dan abnormal pasien. / Uroflowmetry is a non-invasive measurement of the volume of urine excreted over time. Conventional uroflowmetry has become the main modality of urine flow measurement within times. However, this method requires the patient to be present in the hospital or healthcare setting on multiple occasions, sometimes making patients feel uncomfortable undergoing the examination and inaccurately profiling daily micturition. Mobile acoustic uroflowmetry (sono-uroflowmetry) has been proposed as an alternative method of urine flow measurement due to its portability. This study aims to evaluate the accuracy and reliability of sono-uroflowmetry compared to conventional uroflowmetry. Deep learning techniques for image analysis allow for the creation of exploratory data analysis tools that are userfriendly. Using the visual programming toolbox Orange (http://orange.biolab.si), we simplify image analysis by combining deep-learning embedding, machine learning processes, and data visualization. Orange facilitates the development of workflows for data analysis by assembling components for data pretreatment, visualization, and modeling. We equipped Orange with components that profile photos with vectors of features using deep convolutional networks that have been pre-trained. These vectors are utilized in image clustering and classification inside a framework that facilitates the mining of image collections by both novice and seasoned users.. Voided volume, voiding duration, maximum flow rate, and average flow rate were identified and used to determine measurement outcomes. Sonouroflowmetry showed significant correlations compared to conventional uroflowmetry. Hence it can be used as an alternative to conventional uroflowmetry. We demonstrate the utility of the tool in image analysis of urinary streamrecordings in differing the normal and abnormal voiding pattern of the patients.
Item Type: Thesis (Masters)
Creators:
Creators
NIM
Email
ORCID
Rangganata, Ervandy
NIM01679210005
UNSPECIFIED
UNSPECIFIED
Contributors:
Contribution
Contributors
NIDN/NIDK
Email
Thesis advisor
UNSPECIFIED
UNSPECIFIED
UNSPECIFIED
Uncontrolled Keywords: Uroflowmetri konvensional ; uroflowmetri akustik ; pembelajaran mesin ; deep learning
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: University Subject > Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Master of Informatics
Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Master of Informatics
Depositing User: Phillips Iman Heri Wahyudi
Date Deposited: 26 Mar 2025 02:11
Last Modified: 26 Mar 2025 02:11
URI: http://repository.uph.edu/id/eprint/67952

Actions (login required)

View Item
View Item