AI for medical imaging goes deep

Ting, Daniel Sw; Liu, Yong; Burlina, Philippe; Xu, Xinxing; Bressler, Neil M.; Wong, Tien Yin

doi:10.1038/s41591-018-0029-3

Cited by 166 publications

(115 citation statements)

References 12 publications

Supporting

Mentioning

100

Contrasting

Unclassified

Order By: Relevance

“…However, several digital technologies that can be applied to tackle major clinical problems and diseases are now available. These digital technologies include: the Internet of Things, with next-generation telecommunication networks; 27 , 28 big-data analytics; 29 artificial intelligence that uses deep learning; 30 , 31 blockchain technology. 32 Of course, these technologies may work synergistically, enhancing the chance to manage health by using modified algorithms to ensure secured but traceable data.…”

Section: The Decaloguementioning

confidence: 99%

Covid-19 and the management of patients with inflammatory bowel disease: a practical decalogue for the post-pandemic phase

Papa

Lopetuso

et al. 2020

Therap Adv Gastroenterol

View full text Add to dashboard Cite

The coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has raised several concerns for patients with chronic immune-mediated diseases, including inflammatory bowel disease (IBD). As the outbreak appears to be in the descending phase, at least in some part of the world, as in most European countries, guidance is urgently needed to provide optimal care for our IBD patients in order to gradually and safely reduce the gap in care that has been accumulated in the months of lockdown and to face all the backlogs. Therefore, we have provided a decalogue of practical recommendations for gastroenterologists to manage patients with IBD in the post-peak phase of the COVID-19 pandemic. They include all the aspects of IBD care, not only pharmacological ones but also endoscopy, surgery, psychological treatment, telemedicine, diagnostics and educational tasks provided by doctors and patient associations.

show abstract

Section: The Decaloguementioning

confidence: 99%

Covid-19 and the management of patients with inflammatory bowel disease: a practical decalogue for the post-pandemic phase

Papa

Lopetuso

et al. 2020

Therap Adv Gastroenterol

View full text Add to dashboard Cite

show abstract

“…Costeffective strategies for DR management includes routine screening using retinal photographs and having referable cases (typically moderate or worse DR and/or diabetic macular edema) managed by eye care specialists [3][4][5] . Recently, deep learning (DL) using convolutional neural networks (CNNs) has sparked tremendous interest in medicine 6 . In ophthalmology, many DL algorithms and systems have been reported to achieve robust performances in detecting various ocular diseases from retinal photographs [7][8][9] , especially for DR [10][11][12][13] .…”

Section: Introductionmentioning

confidence: 99%

Technical and imaging factors influencing performance of deep learning systems for diabetic retinopathy

Yip

Lim

et al. 2020

npj Digit. Med.

Self Cite

View full text Add to dashboard Cite

Deep learning (DL) has been shown to be effective in developing diabetic retinopathy (DR) algorithms, possibly tackling financial and manpower challenges hindering implementation of DR screening. However, our systematic review of the literature reveals few studies studied the impact of different factors on these DL algorithms, that are important for clinical deployment in real-world settings. Using 455,491 retinal images, we evaluated two technical and three image-related factors in detection of referable DR. For technical factors, the performances of four DL models (VGGNet, ResNet, DenseNet, Ensemble) and two computational frameworks (Caffe, TensorFlow) were evaluated while for image-related factors, we evaluated image compression levels (reducing image size, 350, 300, 250, 200, 150 KB), number of fields (7-field, 2-field, 1-field) and media clarity (pseudophakic vs phakic). In detection of referable DR, four DL models showed comparable diagnostic performance (AUC 0.936-0.944). To develop the VGGNet model, two computational frameworks had similar AUC (0.936). The DL performance dropped when image size decreased below 250 KB (AUC 0.936, 0.900, p < 0.001). The DL performance performed better when there were increased number of fields (dataset 1: 2-field vs 1field-AUC 0.936 vs 0.908, p < 0.001; dataset 2: 7-field vs 2-field vs 1-field, AUC 0.949 vs 0.911 vs 0.895). DL performed better in the pseudophakic than phakic eyes (AUC 0.918 vs 0.833, p < 0.001). Various image-related factors play more significant roles than technical factors in determining the diagnostic performance, suggesting the importance of having robust training and testing datasets for DL training and deployment in the real-world settings.

show abstract

“…T he year 2020 should have been the start of an exciting decade in medicine and science, with the development and maturation of several digital technologies that can be applied to tackle major clinical problems and diseases. These digital technologies include the internet of things (IoT) with nextgeneration telecommunication networks (e.g., 5G) 1,2 ; big-data analytics 3 ; artificial intelligence (AI) that uses deep learning 4,5 ; and blockchain technology 6 . They are highly inter-related: the proliferation of the IoT (e.g., devices and instruments) in hospitals and clinics facilitates the establishment of a highly interconnected digital ecosystem, enabling real-time data collection at scale, which could then be used by AI and deep learning systems to understand healthcare trends, model risk associations and predict outcomes.…”

mentioning

confidence: 99%