2024 · 12 citations · 11 references
ComputingBiomedical EngineeringComputer ScienceBiomedical ImagingDiagnostic ImagingMachine LearningDeep LearningImage RegistrationMedical Image AnalysisMedical ImagingNeural NetworksReinforcement LearningDeep Reinforcement LearningHyperparameter TuningLandmark DetectionLesion DetectionModel-free AlgorithmsNeural Architecture SearchOptimization AlgorithmsView Plane
Deep reinforcement learning (DRL), which learns a set of behaviors that maximize the projected reward, combines the representational power of deep neural networks with the reinforcement learning paradigm. DRL holds great promise for the future of healthcare and medicine, according to recent studies.An overview of the research on DRL in medical imaging is provided in this article. We start with a comprehensive DRL course that covers both the most recent model-based and model-free approaches. The tasks covered in the next section of this article are loosely divided into three main categories: (i) parametric medical image analysis tasks like landmark detection,object/lesion detection, registration, and view plane localization; optimization tasks like hyperparameter tuning, augmentation strategy selection, and neural architecture search; and (iii) other applications like surgical gesture segmentation, person tracking, and perso The study finishes with thoughts of potential future directions.
11
Going deeper with convolutions
Christian Szegedy, Wei Liu, Yangqing Jia et al. · 2015 · 46.8K citations
Haibo He, Edwardo A. Garcia · IEEE Transactions on Knowledge and Data Engineering · 2009 · 9.7K citations
ADADELTA: An Adaptive Learning Rate Method
Matthew D. Zeiler · arXiv (Cornell University) · 2012 · 5.5K citations · Full text
Geoffrey E. Hinton · Scholarpedia · 2009 · 868 citations · Full text