Search papers, labs, and topics across Lattice.
Medical imaging is fundamental for modern medicine, providing non-invasive techniques for visualizing structures and pathologies that are used for diagnosis, treatment planning, and monitoring of diseases. The year 2025 marks the 130th anniversary of Roentgen’s X-ray discovery, which was the inception point of modern medical visualization. This paper examines the historical development, core methodologies, and emerging trends in medical imaging modalities. A detailed examination of key imaging modalities including X-ray, Computed Tomography, Magnetic Resonance Imaging, Ultrasound, Positron Emission Tomography, and Single Photon Emission Computed Tomography is provided, highlighting their features, clinical applications, inherent limitations, and future development directions. Subsequently, the paper provides a comprehensive survey of computer vision-based methods adapted for medical image analysis. It begins by examining the evolution of foundational computer vision models, followed by a discussion of convolutional neural networks (CNNs) and Vision Transformers (ViTs) for image classification tasks. U-Net architectures and transformer-based approaches are examined in relation to image segmentation. In conclusion, current state-of-the-art deep learning models are summarized and potential directions for future research are outlined.