Fast Convolutional Neural Network Training Using Selective Data Sampling: Application to Hemorrhage Detection in Color Fundus Images

By M. V. Grinsven, B. van Ginneken, Carel B. Hoyng, Thomas Theelen, Clara I. Sánchez. IEEE Transactions on Medical Imaging, 2016

420 citations recorded by the source corpus.

Subfields: Neurology, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/226df2c0315fcd6ac45413e70d3a0bac3e1b1072

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Neurology

B. van Ginneken

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Clinically Applicable AI System for Accurate Diagnosis, Quantitative Measurements, and Prognosis of COVID-19 Pneumonia Using Computed Tomography (2020, background): “Recent studies demonstrated the potential of integrating AI into both the eye and childhood-disease diagnostic systems and found that it significantly improved clinical diagnostic efficiency and accuracy (Burlina et al., 2017; Kermany et al., 2018; Long et al., 2017; van Grinsven et al., 2016).”
Guest Editorial Deep Learning in Medical Imaging: Overview and Future Promise of an Exciting New Technique (2016, background): “[24] attempt to improve and speed-up the CNN training for medical image analysis tasks by dynamically selecting misclassified negative samples during training.”
Convolutional neural networks based transfer learning for diabetic retinopathy fundus image classification (2017, methodology): “For the hemorrhage detection, this paper [50] proposes a dynamic CNN training strategy.”
IDRiD: Diabetic Retinopathy - Segmentation and Grading Challenge. (2020, methodology): “In another approach researchers used two-dimensional (2D) image patches as an input instead of full-sized images for lesion detection ( Tan et al., 2017b; van Grinsven et al., 2016; Lam et al., 2018; Chudzik et al., 2018; Khojasteh et al., 2018 ), and OD and fovea detection ( Tan et al., 2017a ).”
Automatic segmentation of nine retinal layer boundaries in OCT images of non-exudative AMD patients using deep learning and graph search. (2017, methodology): “Recent works have extended the CNN framework to complex medical image analysis, such as retinal blood vessels segmentation [39,40], retinal hemorrhage detection [41], brain tumor segmentation [42], and cerebral microbleeds detection [43].”
CABNet: Category Attention Block for Imbalanced Diabetic Retinopathy Grading (2020, methodology): “Van Grinsven et al. [18] proposed a selective sampling method to speed-up the training for the detection of hemorrhages in colored fundus images by dynamically selecting misclas-siﬁed negative samples during training.”
CANet: Cross-Disease Attention Network for Joint Diabetic Retinopathy and Diabetic Macular Edema Grading (2019, methodology): “[29] sped up model training by dynamically selecting misclassified negative samples for hemorrhage detection.”
A survey on deep learning in medical image analysis (2017, background): “Some researchers have specifically looked into tackling this imbalance by incorporating intelligence in the training process itself, by applying selective sampling (van Grinsven et al., 2016) or hard negative mining (Wang et al., 2016b).”
Deep Learning and Its Applications in Biomedicine (2018, methodology): “Medical images analysis CNN Brain tumor segmentation, won top 2 in BRATS [108] Segmentation of pancreas in CT [113] Knee cartilage segmentation [114] Segmentation of hippocampus [117] Predict semantic descriptions from medical images [118] Segmentation of MR brain images [121] Anatomy-specific class”