Background Medical artificial intelligence (AI) has entered the clinical implementation phase, although real-world performance of deep-learning systems (DLSs) for screening fundus disease remains unsatisfactory. Our study aimed to train a clinically applicable DLS for fundus diseases using data derived from the real world, and externally test the model using fundus photographs collected prospectively from the settings in which the model would most likely be adopted.Methods In this national real-world evidence study, we trained a DLS, the Comprehensive AI Retinal Expert (CARE) system, to identify the 14 most common retinal abnormalities using 207 228 colour fundus photographs derived from 16 clinical settings with different disease distributions. CARE was internally validated using 21 867 photographs and externally tested using 18 136 photographs prospectively collected from 35 real-world settings across China where CARE might be adopted, including eight tertiary hospitals, six community hospitals, and 21 physical examination centres. The performance of CARE was further compared with that of 16 ophthalmologists and tested using datasets with non-Chinese ethnicities and previously unused camera types. This study was registered with ClinicalTrials.gov, NCT04213430, and is currently closed.
FindingsThe area under the receiver operating characteristic curve (AUC) in the internal validation set was 0•955 (SD 0•046). AUC values in the external test set were 0•965 (0•035) in tertiary hospitals, 0•983 (0•031) in community hospitals, and 0•953 (0•042) in physical examination centres. The performance of CARE was similar to that of ophthalmologists. Large variations in sensitivity were observed among the ophthalmologists in different regions and with varying experience. The system retained strong identification performance when tested using the non-Chinese dataset (AUC 0•960, 95% CI 0•957-0•964 in referable diabetic retinopathy).Interpretation Our DLS (CARE) showed satisfactory performance for screening multiple retinal abnormalities in real-world settings using prospectively collected fundus photographs, and so could allow the system to be implemented and adopted for clinical care.
Canola (Brassica napus) is one of several important oil-producing crops, and the physiological processes, enzymes, and genes involved in oil synthesis in canola seeds have been well characterized. However, relatively little is known about the dynamic metabolic changes that occur during oil accumulation in seeds, as well as the mechanistic origins of metabolic changes. To explore the metabolic changes that occur during oil accumulation, we isolated metabolites from both seed and silique wall and identified and characterized them by using gas chromatography coupled with mass spectrometry (GC-MS). The results showed that a total of 443 metabolites were identified from four developmental stages. Dozens of these metabolites were differentially expressed during seed ripening, including 20 known to be involved in seed development. To investigate the contribution of tissue-specific carbon sources to the biosynthesis of these metabolites, we examined the metabolic changes of silique walls and seeds under three treatments: leaf-detachment (Ld), phloem-peeling (Pe), and selective silique darkening (Sd). Our study demonstrated that the oil content was independent of leaf photosynthesis and phloem transport during oil accumulation, but required the metabolic influx from the silique wall. Notably, Sd treatment resulted in seed senescence, which eventually led to a severe reduction of the oil content. Sd treatment also caused a significant accumulation of fatty acids (FA), organic acids and amino acids. Furthermore, an unexpected accumulation of sugar derivatives and organic acid was observed in the Pe- and Sd-treated seeds. Consistent with this, the expression of a subset of genes involved in FA metabolism, sugar and oil storage was significantly altered in Pe and Sd treated seeds. Taken together, our studies suggest the metabolite profiles of canola seeds dynamically varied during the course of oil accumulation, which may provide a new insight into the mechanisms of the oil accumulation at the metabolite level.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.