PENERAPAN ALGORITMA MACHINE LEARNING DALAM DETEKSI KANKER MELALUI RADIOGRAFI : PENDEKATAN SYSTEMATIC LITERATURE REVIEW

Abstract

Early cancer detection is key to improving patient survival rates; however, manual interpretation of radiographic images still faces challenges such as limited medical personnel and diagnostic subjectivity. With the advancement of artificial intelligence, machine learning (ML) algorithms have emerged as a potential solution to enhance the accuracy and efficiency of cancer detection using radiographic images such as X-rays, CT scans, and MRIs. This study aims to systematically map the implementation of ML algorithms in radiographic-based cancer detection through a Systematic Literature Review (SLR) approach. Following the PRISMA protocol, an initial screening of 100 articles was conducted, resulting in 30 relevant articles published between 2020 and 2025. The findings reveal that convolutional neural networks (CNNs) and their derivatives (e.g., VGG, ResNet, EfficientNet) dominate implementations, showing high accuracy, particularly in lung and breast cancer detection. Key challenges identified include limited datasets, low model interpretability, and poor generalization to unseen data. In response, several solutions have been proposed, including GAN-based data augmentation, multimodal data integration, and the use of Explainable AI techniques to improve model transparency. This study highlights the importance of integrating technology and clinical context to accelerate the adoption of ML in developing more reliable and human-centered cancer diagnostic systems.