Page 56 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

Article type Journal Journal Journal Journal Journal Journal

References [97] [98] [99] [100] [101] [102]

images set cases cases comprised and malignant digital were Dutch screening and contained by nodules

974 data 330 from 281 from set from views, training 1000 from screening a that of were 1729 of the of cancer acquired of nodules
containing proprietary images images data for roughly collected from full-field cases Permanente facilities non-PFNs scans center lung was Society (JSRT) lung without

set set a US MG validation mammographic used of cysts obtained Netherlands 840 of 210 Kaiser and baseline one that set Japanese Technology confirmed cases
data data cases, 408 digital and were validation 600 The from 9 California PFNs from from randomized data the with normal

on DDSM 512 646 used training 44,090 39,872 for consisted and mammograms in consisted mammograms at of participants (NELSON) standard from Radiological cases 93
Remark Public from containing and was The of which 4218 set Data masses program set Data collected Northern Samples collected eBelgian trial The used 154 and CT

imaging.dcontinued Modality CNN and US CNN mammography Mammography Mammography Mammography CT Chest radiography

medical of Type Multitask CNN Deep CNN Deep CNN Deep CNN AlexNet

to and in lesions solitary breast of
application detection lesions of computer-aided mammographic discriminate lesions of detection automatic pulmonary (PFNs) classification

CNN for description images for to tissue soft for calcifications for of nodules for

of designed designed of designed from designed designed classification perifissural designed nodule
Literature Applications Model semantic diagnostic Model detection Model cysts Model arterial Model Model lung

2.2 Year 2017 2015 2016
Table Body part Chest