AIM:To explore the effectiveness of predictive models based on quantitative characteristics of multimodal fundus images in assessing the risk of progression of non-proliferative diabetic retinopathy(NPDR).METHODS:A single-center retrospective cohort study was conducted. Patients diagnosed with moderate to severe NPDR between January 2021 and December 2023 in Wuhan Ninth Hospital were selected. Baseline color fundus photography(CFP), optical coherence tomography(OCT), and fundus autofluorescence(FAF)images were collected. Microaneurysm count, hard exudate area, central retinal thickness, retinal nerve fiber layer thickness, and total area of hyperautofluorescent spots were extracted and quantified. All patients were followed up regularly. The primary endpoint was the progression to treatment-requiring proliferative diabetic retinopathy(PDR)or clinically significant diabetic macular edema(DME)during the follow-up period. According to the occurrence of endpoint events during follow-up, the patients were divided into the progression group and the non-progression group. Combined with clinical significance and inter-group comparison results, candidate quantitative imaging parameters were included in the multivariate logistic regression analysis to construct a prediction model. The model was visualized using a Nomogram. The discriminative ability and calibration degree of the model were evaluated via receiver operating characteristic(ROC)curves and calibration curves.RESULTS:A total of 300 patients(300 eyes)completed follow-up, with 93 in the progression group, including 47 males and 46 females, with an average age of 59.02±8.21 y. There were 207 cases in the non-progression group, including 105 males and 102 females, with an average age of 58.01±8.10 y. Multivariate logistic regression showed that the total area of FAF hyperautofluorescent spots(OR=1.580, 95%CI: 1.061-1.718, P=0.002),OCT retinal nerve fiber layer thickness(OR=1.520, 95%CI: 1.307-6.212, P=0.013), and CFP microaneurysm count(OR=1.459, 95%CI: 1.089-2.095, P=0.015)were risk factors for DR progression. The AUC of the Nomogram for predicting progression risk during the follow-up period was 0.889(95%CI: 0.826-0.952), and the calibration curve showed good consistency between predicted probability and actual risk.CONCLUSION: The prediction model integrating quantitative imaging features of CFP, OCT and FAF can effectively identify NPDR patients with high progression risk and exhibits good predictive efficacy, which can serve as a quantitative tool for individualized clinical risk management.

Objective:To explore the application effectiveness of the large language model DeepSeek in the cultivation of clinical thinking skills for medical laboratory technology students.Methods:A non-randomized controlled study was conducted. In the 2024-2025 academic year, two classes of second-year medical laboratory technology students from Hubei University of Chinese Medicine were selected and divided into a DeepSeek-assisted teaching group (Class A, n=53) and a traditional teaching control group (Class B, n=53), totaling 106 students. Both groups followed a 20-week problem-based learning (PBL) framework with identical teaching content, instructors, and class hours. Class A utilized DeepSeek via the"Learning Pass AI"platform for case diagnosis reasoning, prompt construction training, test plan formulation, and result analysis, while Class B received traditional PBL instruction. Paired t-tests were used to compare pre-and post-teaching scores in clinical thinking skills, AI interaction literacy, and prompt construction in Class A. Independent samples t-tests and chi-square ( χ2) tests were used to evaluate differences in case reasoning scores, etiology analysis accuracy, and teaching satisfaction between groups. Structured questionnaires supplemented the evaluation of model-assisted teaching processes. Results:The comparison of pre-and post-teaching scores in Class A showed that post-teaching scores significantly improved in clinical thinking skills[(4.02±0.45) points vs. (3.09±0.50) points, t=2.23)] and AI interaction literacy [(4.62±0.41) points vs. (3.27±0.54) points, t=2.18]. Compared to Class B, Class A demonstrated superior performance in case reasoning scores [(81.1±3.8) points vs.(74.3±4.2) points, t=8.97], etiology analysis accuracy [94.3% (50/53) vs. 81.1% (43/53), χ2=4.29], and teaching satisfaction [(95.6±3.2)points vs. (82.6±4.8) points, t=11.86] ( P<0.05). The results of questionnaires indicated that during model application, the prompt construction improved in logic [(2.85±0.58) points to (4.25±0.50) points, t=14.23, P<0.01] and innovation [(2.60±0.53) points to (4.05±0.46) points, t=11.57, P<0.05], but question clarity (77.4%, 41/53) and medical terminology accuracy (43.4%, 23/53) remained primary shortcomings. Conclusion:Integrating large language models into AI-teacher collaborative learning pathways can effectively promote students′ autonomous inquiry and clinical reasoning skills, thereby enhancing medical laboratory technology students′ clinical thinking skills.

Objective:To explore the application effectiveness of the large language model DeepSeek in the cultivation of clinical thinking skills for medical laboratory technology students.Methods:A non-randomized controlled study was conducted. In the 2024-2025 academic year, two classes of second-year medical laboratory technology students from Hubei University of Chinese Medicine were selected and divided into a DeepSeek-assisted teaching group (Class A, n=53) and a traditional teaching control group (Class B, n=53), totaling 106 students. Both groups followed a 20-week problem-based learning (PBL) framework with identical teaching content, instructors, and class hours. Class A utilized DeepSeek via the"Learning Pass AI"platform for case diagnosis reasoning, prompt construction training, test plan formulation, and result analysis, while Class B received traditional PBL instruction. Paired t-tests were used to compare pre-and post-teaching scores in clinical thinking skills, AI interaction literacy, and prompt construction in Class A. Independent samples t-tests and chi-square ( χ2) tests were used to evaluate differences in case reasoning scores, etiology analysis accuracy, and teaching satisfaction between groups. Structured questionnaires supplemented the evaluation of model-assisted teaching processes. Results:The comparison of pre-and post-teaching scores in Class A showed that post-teaching scores significantly improved in clinical thinking skills[(4.02±0.45) points vs. (3.09±0.50) points, t=2.23)] and AI interaction literacy [(4.62±0.41) points vs. (3.27±0.54) points, t=2.18]. Compared to Class B, Class A demonstrated superior performance in case reasoning scores [(81.1±3.8) points vs.(74.3±4.2) points, t=8.97], etiology analysis accuracy [94.3% (50/53) vs. 81.1% (43/53), χ2=4.29], and teaching satisfaction [(95.6±3.2)points vs. (82.6±4.8) points, t=11.86] ( P<0.05). The results of questionnaires indicated that during model application, the prompt construction improved in logic [(2.85±0.58) points to (4.25±0.50) points, t=14.23, P<0.01] and innovation [(2.60±0.53) points to (4.05±0.46) points, t=11.57, P<0.05], but question clarity (77.4%, 41/53) and medical terminology accuracy (43.4%, 23/53) remained primary shortcomings. Conclusion:Integrating large language models into AI-teacher collaborative learning pathways can effectively promote students′ autonomous inquiry and clinical reasoning skills, thereby enhancing medical laboratory technology students′ clinical thinking skills.

Objective:To explore the regulatory effects of Danzhi Jiangtang Capsules on glucose metabolism and mitochondrial oxidative stress levels in db/db mice.Methods:Totally 32 db/db mice were randomly divided into model group, positive control group, and Danzhi Jiangtang Capsule low-, medium- and high-dosage groups; at the same time, 8 same-sex C57BL/6 mice of the same week age were selected as the blank group. The metformin group was filled with metformin solution 0.1 g/kg, and the Danzhi Jiangtang Capsule low-, medium- and high-dosage groups were injected with 0.99, 0.49 and 0.25 g/kg Danzhi Jiangtang Capsule solution respectively. The model group and the blank group received the same volume of physiological saline solution for the gavage, 1 time/d, continuous 12 weeks. The weight and fasting blood sugar (FBG) changes in each group of mice were detected; the serum insulin (FINS) level was detected by ELISA method and the insulin resistance index (HOMA-IR) was calculated; kits were used to detect the pancreatic tissue SOD and MDA levels of each group of mice; HE staining was used to observe pathological morphology of pancreatic tissue. Transmission electron microscopy was used to observe the ultrastructure of pancreatic mitochondria; Western blot method was used to detect the expressions of p-AMPK and AMPK proteins in pancreatic tissue.Results:Compared with the model group, after 8 or 12 weeks of drug administration, the weight of mice in the Danzhi Jiangtang Capsule high-dosage group decreased ( P<0.05 or P<0.01), and the level of FBG decreased ( P<0.01); FINS and HOMA-IR in the Danzhi Jiangtang Capsule high- and medium-dosage groups decreased ( P<0.01), the SOD activity in Danzhi Jiangtang Capsule high- medium- and low-dosage groups increased, and the level of MDA decreased ( P<0.01); the expressions of p-AMPK and AMPK in Danzhi Jiangtang Capsule high-and medium-dosage groups increased ( P<0.01), the expression of AMPK in Danzhi Jiangtang Capsule low-dosage group increased ( P<0.01). Conclusion:Danzhi Jiangtang Capsule can improve SOD activity in pancreatic tissue of db/db mice, reduce MDA content, and activate the AMPK signaling pathway, suggesting that Danzhi Jiangtang Capsule can improve insulin resistance and restore glucose homeostasis by inhibiting oxidative stress levels and improving mitochondrial function.