Convolutional neural networks detect cells in densely packed images at performance levels similar to human readers.
Madeleine S. DurkeeNevaeh PetrieKyle LlerasJunting AiRebecca AbrahamJ. Cy ChittendenChasity KasirFiona ClarkGabriel CasellaMarcus R. ClarkMaryellen L. GigerPublished in: Medical Imaging: Image Perception, Observer Performance, and Technology Assessment (2023)
Keyphrases
- densely packed
- image data
- image database
- convolutional neural networks
- image retrieval
- input image
- image analysis
- human subjects
- image features
- three dimensional
- image collections
- automatically detected
- image classification
- edge detection
- image registration
- ground truth
- convolutional network
- human retina
- image regions
- microscope images
- human observers
- object recognition
- image set
- test images
- detection algorithm
- coarse to fine
- visually similar
- cell nuclei
- image matching
- cancer cells
- cell segmentation
- segmentation method
- feature points
- images depicting
- high speed