Pengembangan Model Computer-Aided Diagnosis (CAD) Berbasis Deep Learning dan Explainable AI (XAI) untuk Diagnosis Knee Osteoarthritis dari Gambar X-ray

Shafrizaliansyah, Maulana Ilyasa (2025) Pengembangan Model Computer-Aided Diagnosis (CAD) Berbasis Deep Learning dan Explainable AI (XAI) untuk Diagnosis Knee Osteoarthritis dari Gambar X-ray. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5027211065-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Pengembangan Model Computer-Aided Diagnosis (CAD) Berbasis Deep Learning dan Explainable AI (XAI) untuk Diagnosis Knee Osteoarthritis dari Gambar X-ray bertujuan untuk menciptakan model berbasis kecerdasan buatan yang dapat membantu dokter dalam mendiagnosis Knee Osteoarthritis (KOA) dengan lebih akurat dan efisien. Model ini mengimplementasikan teknik Deep Learning menggunakan Convolutional neural networks (CNN) untuk menganalisis gambar X-ray dan memprediksi tingkat keparahan KOA berdasarkan KL Grading System. Dalam penelitian ini, digunakan beberapa arsitektur model CNN terbaru untuk mengoptimalkan performa model dalam ekstraksi fitur dari gambar medis. Selain itu, penelitian ini juga mengintegrasikan teknik Explainable AI (XAI), khususnya Grad-CAM, untuk memberikan penjelasan visual mengenai bagian-bagian gambar X-ray yang paling berpengaruh dalam pengambilan keputusan oleh model, sehingga meningkatkan transparansi dan kepercayaan terhadap hasil diagnosis. Data yang digunakan dibagi menjadi tiga set: training (70%), validation (20%), dan testing (10%), dengan evaluasi kinerja model dilakukan menggunakan metrik accuracy, F1-score, dan confusion matrix. Evaluasi ini bertujuan untuk mengukur akurasi dan performa model dalam membantu diagnosis Knee Osteoarthritis (KOA), serta untuk memastikan bahwa model dapat memberikan penjelasan yang transparan dan mudah dipahami oleh tenaga medis. Hasil penelitian menunjukkan bahwa model individu ResNet50 mencapai akurasi 91% dan DenseNet121 mencapai akurasi 95%, sementara VGG19 mencapai akurasi 91%; dengan pendekatan ensemble yang menggabungkan ketiga model, akurasi meningkat menjadi 93%, menegaskan bahwa model DenseNet121 serta kombinasi ensemble akan menghasilkan interpretabilitas dan akurasi melalui Explainable AI (XAI) dapat mendukung diagnosis klinis yang lebih andal dan lebih baik.
=======================================================================================================================================
Development of a Computer-Aided Diagnosis (CAD) Model Based on Deep Learning and Explainable AI (XAI) for Knee Osteoarthritis Diagnosis from X-ray Images aims to create an artificial intelligence-based model that can assist doctors in diagnosing Knee Osteoarthritis (KOA) more accurately and efficiently. This model implements Deep Learning techniques using Convolutional neural networks (CNN) to analyze X-ray images and predict the severity of Knee Osteoarthritis (KOA) based on the KL Grading System. In this research, several state-of-the-art CNN model architectures are used to optimize the model's performance in feature extraction from medical images. In addition, this research also integrates Explainable AI (XAI) techniques, specifically Grad-CAM, to provide visual explanations of the parts of X-ray images that most influence the model's decision-making, thereby increasing transparency and trust in the diagnostic results. The data used is divided into three sets: training (70%), validation (20%), and testing (10%), with model performance evaluation conducted using accuracy, F1-score, and confusion matrix metrics. This evaluation aims to measure the accuracy and performance of the model in assisting the diagnosis of Knee Osteoarthritis (KOA), as well as to ensure that the model can provide explanations that are transparent and easily understood by medical personnel. The research results show that the individual ResNet50 model achieved an accuracy of 91% and the DenseNet121 achieved an accuracy of 95%, while VGG19 achieved an accuracy of 91%; with an ensemble approach combining the three models, the accuracy increased to 93%, confirming that the DenseNet121 model and the ensemble combination would provide interpretability and accuracy through Explainable AI (XAI) to support more reliable and better clinical diagnosis.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Convolutional neural networks (CNN),Deep Learning, Explainable AI (XAI), Grad-CAM, X-ray.
Subjects:	T Technology > T Technology (General) > T385 Visualization--Technique T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.8 Nonlinear programming. Support vector machine. Wavelets. Hidden Markov models. T Technology > T Technology (General) > T58.62 Decision support systems T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User:	Maulana Ilyasa Shafrizaliansyah
Date Deposited:	23 Jul 2025 03:26
Last Modified:	23 Jul 2025 03:26
URI:	http://repository.its.ac.id/id/eprint/120753

Actions (login required)

View Item