Recognition of eye diseases based on deep neural networks for transfer learning and improved D-S evidence theory.

BACKGROUND: Human vision has inspired significant advancements in computer vision, yet the human eye is prone to various silent eye diseases. With the advent of deep learning, computer vision for detecting human eye diseases has gained prominence, but most studies have focused only on a limited number of eye diseases. RESULTS: Our model demonstrated a reduction in inherent bias and enhanced robustness. The fused network achieved an Accuracy of 0.9237, Kappa of 0.878, F1 Score of 0.914 (95% CI [0.875-0.954]), Precision of 0.945 (95% CI [0.928-0.963]), Recall of 0.89 (95% CI [0.821-0.958]), and an AUC value of ROC at 0.987. These metrics are notably higher than those of comparable studies. CONCLUSIONS: Our deep neural network-based model exhibited improvements in eye disease recognition metrics over models from peer research, highlighting its potential application in this field. METHODS: In deep learning-based eye recognition, to improve the learning efficiency of the model, we train and fine-tune the network by transfer learning. In order to eliminate the decision bias of the models and improve the credibility of the decisions, we propose a model decision fusion method based on the D-S theory. However, D-S theory is an incomplete and conflicting theory, we improve and eliminate the existed paradoxes, propose the improved D-S evidence theory(ID-SET), and apply it to the decision fusion of eye disease recognition models.

Can platelet-rich plasma enhance the effect of meniscus repair? A meta-analysis of randomized controlled trials Platelet-rich plasma and meniscus repair.

BACKGROUND: Studies have shown that platelet-rich plasma (PRP) can enhance the effect of meniscus repair, but some studies have suggested different views on the role of PRP. PURPOSE: To determine whether PRP can enhance the effect of meniscus repair with respect to pain reduction and improved functionality and cure rate in patients with meniscus injury. METHODS: By searching PubMed, EMBASE, Cochrane Library databases, clinicaltrials.gov, and the CNKI database from their inception till December 1, 2020, we performed a meta-analysis of RCTs reporting the results of the Pain Visual Analog Scale (VAS), the pain of Knee injury and Osteoarthritis Outcome Score (KOOS), Lysholm score, the International Knee Documentation Committee (IKDC), healing rate, and adverse events. The risk of bias is assessed using Cochrane's collaborative tools. The summary results are expressed with effect size and 95% confidence interval, and sensitivity were performed. RESULTS: The meta-analysis included 9 RCTs and 345 patients. In general, compared with the control group, used of PRP during meniscus surgery significantly improved the pain (SMD: -0.95, p < 0.00001,95% CI: -1.22 to -0.69, I2 = 42%) and knee joint function (SMD: 1.00, p = 0.01.95% CI: 0.22 to 1.79, I2 = 89%) of patients with meniscus injury at 6 months after treatment. However, both PRP and non-PRP showed improvements in the pain and knee joint function, with no significant difference between the groups at 1 months and beyond 12 months. The PRP enhancement technique showed benefit in improving the cure rate of meniscus repair (RR:1.44; p < 0.0001, 95% CI: 1.20-1.73). No serious adverse events were reported in any study. CONCLUSION: As an enhancement program for meniscus repair, PRP is worthy of further consideration in improving the function and pain of patients during the mid-term follow-up after surgery, and PRP can further improve the healing rate of meniscus repair. However, the evidence still needs to be interpreted carefully because of the quantity and quality of the included studies.