OBJECTIVE To comprehensively evaluate the 16 commonly used kinds of enteral nutrition preparations in Hebei province， aiming to provide a reference for the selection of drugs in medical institutions and clinical drug decision-making. METHODS Based on the Quick Guide for Drug Evaluation and Selection in Chinese Medical Institutions （the Second Edition）， evaluation evidence was collected， and the included drugs were scored and evaluated from four dimensions of pharmaceutical characteristics， clinical characteristics， economy and other attributes. RESULTS & CONCLUSIONS The scores for Enteral nutritional emulsion （TPF-T）， Enteral nutritional emulsion （TPF-D）， Enteral nutritional emulsion （TPF）， Enteral nutritional emulsion （TPF-HE）， Enteral nutritional emulsion （TP）， Enteral nutritional emulsion （SP）， Enteral nutritional suspension （TPF） （1.5 kcal/mL， 1 kcal＝4.184 kJ）， Enteral nutritional suspension （TPF） （1.0 kcal/mL）， Intact protein enteral nutrition （powder）， Enteral nutritional suspension （TPF-DM）， Enteral nutritional suspension （TPF-MCT）， Enteral nutritional suspension （SP）， Short- peptide enteral nutrition， Enteral nutritional powder （TP）， Enteral nutritional suspension （TPF-D） and Enteral nutritional suspension （TPF-FOS） were 82.9， 84.1， 84.1， 86.1， 78.4， 79.1， 82.6， 82.3， 82.4， 80.2， 83.0， 82.4， 82.1， 85.7， 76.0， 82.4 points， respectively. All medications scored above 70 points. In practice， appropriate drugs can be selected according to clinical requirements and patient needs.

ObjectiveHeadspace solid-phase microextraction-gas chromatography-mass spectrometry（HS-SPME-GC-MS） and GC-triple quadrupole MS（GC-QqQ-MS） in combination with non-targeted and targeted metabolomics were employed to systematically analyze the chemical composition differences of Xihuangwan prepared with natural musk and artificial musk， and establish an identification system for them. MethodsThe volatile components of 9 batches of Xihuangwan samples from 8 manufacturers were analyzed by HS-SPME-GC-MS non-targeted metabolomics， and identified by comparing their MS data with the National Institute of Standards and Technology（NIST） spectral library. Orthogonal partial least squares-discriminant analysis（OPLS-DA） was used to identify differential volatile components of Xihuangwan prepared with natural musk and artificial musk. Additionally， GC-QqQ-MS targeted metabolomics was applied to quantify the levels of α-pinene， β-elemene， muscone， dehydroepiandrosterone， bornyl acetate， and octyl acetate in 27 batches of samples from 9 manufacturers. Cluster analysis， principal component analysis（PCA）， and partial least squares-discriminant analysis（PLS-DA） were conducted to further explore the differences in volatile components between Xihuangwan samples prepared with natural musk and artificial musk. ResultsNon-targeted metabolomics identified 291 volatile compounds in Xihuangwan， including alkanes， esters， alkanes， alcohols， ketones， naphthalenes and others. OPLS-DA analysis revealed distinct separation between Xihuangwan samples containing artificial musk（A1， C1， D1， E1， F1， G1， I1） and those containing natural musk（H1， H3）. A total of 30 differential metabolites were identified. The relative contents of these 30 differential metabolites were visualized using a radar chart， revealing significant differences in the levels of octanol， borneol acetate and muscone. Cluster analysis and PCA results from targeted metabolomics indicated that Xihuangwan could be classified into two distinct groups：one composed of natural musk（H1， H3） and the other of artificial musk， sample H2. PLS-DA identified muscone， octyl acetate， and dehydroepiandrosterone as key differential volatile components. Although no significant difference was observed in the content of octyl acetate between the two groups， statistically significant differences were found for muscone and dehydroepiandrosterone（P<0.05）. ConclusionMuscone and dehydroepiandrosterone can be used for the differentiation of Xihuangwan samples containing natural musk from those containing artificial musk. This study systematically and comprehensively analyzed the differences in the types and contents of major volatile components in Xihuangwan prepared with natural musk and artificial musk， providing a scientific basis for quality evaluation and control of Xihuangwan.

ObjectiveHeadspace solid-phase microextraction-gas chromatography-mass spectrometry（HS-SPME-GC-MS） and GC-triple quadrupole MS（GC-QqQ-MS） in combination with non-targeted and targeted metabolomics were employed to systematically analyze the chemical composition differences of Xihuangwan prepared with natural musk and artificial musk， and establish an identification system for them. MethodsThe volatile components of 9 batches of Xihuangwan samples from 8 manufacturers were analyzed by HS-SPME-GC-MS non-targeted metabolomics， and identified by comparing their MS data with the National Institute of Standards and Technology（NIST） spectral library. Orthogonal partial least squares-discriminant analysis（OPLS-DA） was used to identify differential volatile components of Xihuangwan prepared with natural musk and artificial musk. Additionally， GC-QqQ-MS targeted metabolomics was applied to quantify the levels of α-pinene， β-elemene， muscone， dehydroepiandrosterone， bornyl acetate， and octyl acetate in 27 batches of samples from 9 manufacturers. Cluster analysis， principal component analysis（PCA）， and partial least squares-discriminant analysis（PLS-DA） were conducted to further explore the differences in volatile components between Xihuangwan samples prepared with natural musk and artificial musk. ResultsNon-targeted metabolomics identified 291 volatile compounds in Xihuangwan， including alkanes， esters， alkanes， alcohols， ketones， naphthalenes and others. OPLS-DA analysis revealed distinct separation between Xihuangwan samples containing artificial musk（A1， C1， D1， E1， F1， G1， I1） and those containing natural musk（H1， H3）. A total of 30 differential metabolites were identified. The relative contents of these 30 differential metabolites were visualized using a radar chart， revealing significant differences in the levels of octanol， borneol acetate and muscone. Cluster analysis and PCA results from targeted metabolomics indicated that Xihuangwan could be classified into two distinct groups：one composed of natural musk（H1， H3） and the other of artificial musk， sample H2. PLS-DA identified muscone， octyl acetate， and dehydroepiandrosterone as key differential volatile components. Although no significant difference was observed in the content of octyl acetate between the two groups， statistically significant differences were found for muscone and dehydroepiandrosterone（P<0.05）. ConclusionMuscone and dehydroepiandrosterone can be used for the differentiation of Xihuangwan samples containing natural musk from those containing artificial musk. This study systematically and comprehensively analyzed the differences in the types and contents of major volatile components in Xihuangwan prepared with natural musk and artificial musk， providing a scientific basis for quality evaluation and control of Xihuangwan.

ObjectiveThis paper aims to investigate the effects of Xixintang on aquaporin-4 （AQP4） polarity distribution， blood-brain barrier （BBB） function， and neuroinflammationin rats with Alzheimer's disease （AD）， thereby revealing the potential mechanism through which this formula protects the BBB by regulating AQP4 polarization. The aim is to provide a scientific basis for clinical treatment. MethodsSixty Sprague-Dawley （SD） rats were randomly divided into a normal group， a model group， a probiotic group， a donepezil group， and an Xixintang group. The model was established by intraperitoneal injection of D-galactose （D-Gal） combined with bilateral intracerebroventricular injection of amyloid-β_25-35 （Aβ_25-35）. The probiotic group （30.85 mg·kg^-1）， donepezil group （0.88 mg·kg^-1）， and Xixintang group （1.174 g·kg^-1） received daily gavage administration， while the normal and model groups received intragastric administration with an equal volume of normal saline for one month. Cognitive ability was assessed by using the Morris water maze. BBB permeability was detected via Evans blue extravasation. The contents of interleukin-6 （IL-6）， amyloid-β_1-42 （Aβ_1-42）， and tumor necrosis factor-α （TNF-α） in the hippocampal tissues were measured by enzyme-linked immunosorbent assay （ELISA）. The protein expressions of zonula occludens-1 （ZO-1）， occludin， tissue inhibitor of metalloproteinase-1 （TIMP-1）， matrix metalloproteinase-9 （MMP-9）， and AQP4 in the hippocampal tissues were detected by western blot. The expression and co-localization levels of Aβ_1-42， ionized calcium-binding adapter molecule 1 （IBA1）， and AQP4/platelet endothelial cell adhesion molecule 31 （CD31） in the hippocampal region were examined by immunofluorescence. ResultsCompared with the normal group， the model group exhibited a significant decline in cognitive ability （P<0.01） and a marked increase in Evans blue extravasation in the brain （P<0.01）. The expressions of ZO-1， occludin， and TIMP-1 were significantly decreased （P<0.01）， while the expressions of AQP4 and MMP-9 were significantly increased （P<0.01）. The co-localization level of AQP4/CD31 was significantly reduced （P<0.01）， and the expressions of Aβ_1-42， IL-6， TNF-α， and IBA1 were significantly elevated （P<0.01）. Compared with the model group， the Xixintang group showed significant improvement in cognitive ability （P<0.01） and a significant reduction in Evans blue extravasation in the brain （P<0.01）. The expressions of occludin， TIMP-1， and ZO-1 were significantly increased （P<0.05， P<0.01）， while the expressions of AQP4 and MMP-9 were significantly decreased （P<0.05）. The co-localization level of AQP4/CD31 was significantly enhanced （P<0.01）， and the expressions of Aβ_1-42， IL-6， TNF-α， and IBA1 were significantly reduced （P<0.05， P<0.01）. ConclusionXixintang may improve cognitive function and alleviate AD pathology in AD model rats by regulating AQP4 polarity distribution， thereby breaking the vicious cycle of "Aβ deposition-neuroinflammation-BBB damage" and restoring the homeostasis of the microenvironment in the brain.

ObjectiveTo explore the clinical efficacy of oral administration of modified Tuoli Xiaodusan on postoperative patients with perianal abscess， and its effects on related inflammatory factors and signal transducers and activators of transcription protein 3 （STAT3）/vascular endothelial growth factor （VEGF） signaling pathways. MethodsFrom January 2023 to December 2023 in Inner Mongolia hospital of traditional Chinese medicine， 60 postoperative patients with perianal abscess who met the inclusion criteria were selected. They were divided into a treatment group and a control group using the random number table method， with 30 cases in each group. The control group received conventional treatment， while the treatment group received additional treatment with modified Tuoli Xiaodusan on the basis of the control group. The course of treatment in both groups was three weeks. On the day of operation and on the 7^th， 14^th and 21^st day after operation， enzyme-linked immunosorbent assay （ELISA） was used to measure the expression levels of serum interleukin-1β （IL-1β）， interleukin-6 （IL-6）， and tumor necrosis factor-α （TNF-α）. Hematoxylin eosin （HE） staining was used to observe the pathological morphology of pathological tissue. Western blot was used to measure the levels of phosphorylated STAT3 （p-STAT3） and vascular endothelial growth factor （VEGF） proteins， and real-time fluorescence quantitative polymerase chain reaction （Real-time PCR） was used to determine the expression level of VEGF mRNA. The clinical efficacy of the two groups was compared according to the wound pain， secretion volume score， and healing rate of patients on the 3^rd， 7^th， 14^th， and 21^st day after operation. ResultsThe total effective rate of the treatment group was higher than that of the control group （P<0.05）. For intra-group comparison， the pain score of the control group decreased at each time period （P<0.05）， and the healing rate increased （P<0.05）. The secretion volume score decreased on the 14^th and 21^st days after operation （P<0.05）. The pain score and secretion volume score of the treatment group decreased at each time period （P<0.05）， and the healing rate increased （P<0.05）. The levels of various inflammatory factors decreased in both groups （P<0.05）. Compared with those on the surgical day， the levels of p-STAT3 and VEGF proteins in the wound tissue of the two groups were different on the 7^th and 21^st days after operation （P<0.05）. There were significant differences in VEGF mRNA levels in wound tissue between the two groups at each time period （P<0.01）. For inter-group comparison， on the 7^th and 14^th days after operation， the pain score in the treatment group was lower than that in the control group. On the 7^th， 14^th and 21^st days after operation， the secretion volume scores and healing rate of the treatment group were better than those of the control group （P<0.05）. The levels of various inflammatory factors in the treatment group were lower than those in the control group （P<0.05）， and the decline rate was faster （P<0.05）. On the 7^th day after operation， the levels of p-STAT3， VEGF protein， and VEGF mRNA in the wound tissue of the treatment group were higher than those in the control group （P<0.05）. HE staining showed that the inflammatory cell infiltration in the treatment group decreased faster. The cell arrangement was more orderly， and new blood vessel lumens were visible. There were no abnormalities in the safety observation indexes of all patients during the study period. ConclusionModified Tuoli Xiaodusan can relieve wound pain after perianal abscess surgery， reduce secretions， and improve wound healing rate. The mechanism may be reducing the levels of serum IL-1β， IL-6， and TNF-α， reducing the inflammatory response of the wound， upregulating the expression of p-STAT3 and VEGF proteins， and stimulating the STAT3/VEGF signaling pathway， thereby accelerating angiogenesis and promoting wound healing.

ObjectiveTo characterize the quality differences among different germplasm and introduced varieties of Bupleurum scorzonerifolium roots（BSR）， and explore the underlying molecular mechanisms， providing a basis for high-quality production and quality control. MethodsWild BSR from Yulin（YLW） served as the quality reference， we conducted comparative analysis among YLW， locally domesticated wild germplasm in Yulin（YLC3）， Daqing germplasm introduced and cultivated in Yulin（YLDQC3）， and locally cultivated germplasm in Daqing（DQC3）. A combination of traditional pharmacognostic methods and modern multi-omics analyses was employed， including macroscopic traits（appearance， odor）， microscopic features（proportions of cork， phloem， xylem）， cell wall component contents（hemicellulose， cellulose， lignin）， carbohydrate contents（starch， water-soluble polysaccharides）， marker compound contents（ethanol-soluble extracts， total saponins， liposoluble extracts， and saikosaponins A， B₂， C， D）， metabolomics， and transcriptomics， in order to systematically characterize quality differences and investigate molecular mechanisms among these samples. ResultsMacroscopically， Yulin-produced BSR（YLW， YLC3， YLDQC3） exhibited significantly greater weight， length， and upper and middle diameters than Daqing-produced BSR（DQC3）. Odor-wise， YLW and YLC3 had a a fragrance taste， YLDQC3 had a rancid oil odor， and DQC3 had a sweet and fragrant taste. Microscopically， Yulin germplasm（YLW， YLC3） and Daqing germplasm（YLDQC3， DQC3） shared similar structural features， respectively. However， Yulin germplasm showed significantly higher proportions of cork and phloem， as well as stronger xylem vessel staining intensity compared to Daqing germplasm. Regarding various component contents， Yulin germplasm contained significantly higher levels of ethanol-soluble extracts， total saponins， and saikosaponins A， B₂， C， D， while Daqing germplasm had significantly higher levels of hemicellulose， starch， and liposoluble extracts. After introduction to Yulin， the Daqing germplasm（YLDQC3） showed increased starch， water-soluble polysaccharides and liposoluble extracts contents， decreased cell wall component content， but no significant difference in other component contents. Metabolomics revealed that saponins and terpenes accumulated significantly in Yulin germplasm， while alcohols and aldehydes accumulated predominantly in Daqing germplasm. Transcriptomics indicated similar gene expression patterns within the same germplasm but specificity between different germplasms. Integrative metabolomic-transcriptomic analysis identified 145 potential key genes associated with the saikosaponin biosynthesis pathway， including one acetyl-coenzyme A（CoA） acetyltransferase gene（ACAT）， one 3-hydroxy-3-methylglutaryl-coenzyme A synthase gene（HMGS）， two hydroxymethylglutaryl-CoA（HMG-CoA） reductase genes（HMG）， one phosphomevalonate kinase gene（PMK）， one 1-deoxy-D-xylose-5-phosphate synthase gene（CLA）， one hydroxymethylbuten-1-aldol synthase gene（HDR）， two farnesyl pyrophosphate synthase genes（FPPS）， one squalene synthase gene（SQS）， one β-amyrin synthase gene（BAS）， 102 cytochrome P450（CYP450） gene family members， and 32 uridine diphosphate-glucuronosyltransferase（UGT） gene family members. ConclusionAmong the three cultivated types， YLC3 most closely resembles YLW in appearance， microscopic features， contents of major bioactive constituents， metabolomic and transcriptomic profiles. Yulin germplasm exhibits superior saponin synthesis capability compared to Daqing germplasm， and Yulin region is more suitable for the growth of B. scorzonerifolium. Based on these findings， it is recommended that artificial cultivation in northern Shaanxi and similar regions utilize the local Yulin germplasm source cultivated for at least three years.

ObjectiveThis study aims to investigate whether Xixintang could ameliorate cognitive dysfunction in an Alzheimer's disease （AD） rat model induced by D-galactose and β-amyloid （Aβ_25-35）， by means of repairing the colonic mucosal barrier， regulating the Toll-like receptor 4 （TLR4）/nuclear factor-κB p65 （NF-κB p65） signaling pathway， and intervening in the pathological process mediated by the gut-brain axis. MethodsSixty specific pathogen-free （SPF） male Sprague-Dawley （SD） rats were randomly divided to five groups （n=12）： A control group， a model group， a donepezil group， an Xixintang group， and a probiotic group. Except for those in the control group， rats in all other groups received daily intraperitoneal injections of D-galactose for six consecutive weeks. Subsequently， aggregated Aβ_25-35 was injected stereotactically into the bilateral ventricles to establish the AD model. During the intervention periods， the rats in all groups were administered their respective drugs and normal saline by gavage. The Morris water maze test was used to assess the capacity for spatial learning and memory. Hematoxylin-eosin （HE） staining was employed to observe the histopathological changes in the colon tissues. Immunofluorescence was used to detect Aβ_1-41 deposition in the hippocampal region and Mucin 2 （MUC2） expression in the colonic mucosa. Western blot was performed to measure the protein expression levels of FFAR2，TLR4， NF-κB p65， occludin （OCLN）， zonula occludens-1 （ZO-1）， and MUC2 in the colonic tissues. Enzyme-linked immunosorbent assay （ELISA） was used to determine the contents of interleukin-6 （IL-6）， tumor necrosis factor-α （TNF-α）， serum amyloid A （SAA）， and Aβ_1-42 in the hippocampal region from the colonic tissues. The lipopolysaccharide （LPS） concentrations in colon tissues of rats were measured by using a dynamic chromogenic limulus assay. ResultsCompared with those in the control group， the rats in the model group exhibited a significantly prolonged escape latency and a markedly shorter duration in the target quadrant （P<0.01）. The integrity of the colonic mucosal structure was compromised， with disordered gland arrangement and a reduced number of goblet cells. The Aβ_1-42 deposition in the hippocampal region was significantly increased （P<0.01）. The protein expression levels of TLR4 and NF-κB p65 in colonic tissues were significantly upregulated （P<0.01）， while those of occludin and ZO-1 were downregulated （P<0.01）. The contents of inflammatory factors such as IL-6， TNF-α， and SAA were significantly elevated （P<0.01）， and the LPS level in the serum was markedly increased （P<0.01）. In comparison to those in the model group， the rats in the Xixintang group showed a significantly shortened escape latency and a prolonged duration in the target quadrant （P<0.01）. The colonic mucosal structure was ameliorated， with neat gland arrangement and an increased number of goblet cells. The Aβ_1-42 deposition in the hippocampal region was reduced （P<0.01）. The protein expressions of TLR4 and NF-κB p65 in the colon tissues were decreased （P<0.05，P<0.01）， while the protein levels of occludin and ZO-1 were increased （P<0.01）. The contents of IL-6， TNF-α， and serum amyloid A （SAA） were decreased （P<0.01）， and the LPS level was reduced （P<0.01）. ConclusionXixintang can significantly ameliorate cognitive dysfunction of AD model rats， by means of restoring the colonic mucosal barrier structure， reducing cerebral Aβ deposition， and suppressing peripheral and central inflammatory response. Its mechanism of action may be closely associated with the suppression of the TLR4/NF-κB signaling pathway activation， reduction of endotoxin levels， and regulation of the gut-brain axis.

ObjectiveThis study aims to investigate whether Xixintang could ameliorate cognitive dysfunction in an Alzheimer's disease （AD） rat model induced by D-galactose and β-amyloid （Aβ_25-35）， by means of repairing the colonic mucosal barrier， regulating the Toll-like receptor 4 （TLR4）/nuclear factor-κB p65 （NF-κB p65） signaling pathway， and intervening in the pathological process mediated by the gut-brain axis. MethodsSixty specific pathogen-free （SPF） male Sprague-Dawley （SD） rats were randomly divided to five groups （n=12）： A control group， a model group， a donepezil group， an Xixintang group， and a probiotic group. Except for those in the control group， rats in all other groups received daily intraperitoneal injections of D-galactose for six consecutive weeks. Subsequently， aggregated Aβ_25-35 was injected stereotactically into the bilateral ventricles to establish the AD model. During the intervention periods， the rats in all groups were administered their respective drugs and normal saline by gavage. The Morris water maze test was used to assess the capacity for spatial learning and memory. Hematoxylin-eosin （HE） staining was employed to observe the histopathological changes in the colon tissues. Immunofluorescence was used to detect Aβ_1-41 deposition in the hippocampal region and Mucin 2 （MUC2） expression in the colonic mucosa. Western blot was performed to measure the protein expression levels of FFAR2，TLR4， NF-κB p65， occludin （OCLN）， zonula occludens-1 （ZO-1）， and MUC2 in the colonic tissues. Enzyme-linked immunosorbent assay （ELISA） was used to determine the contents of interleukin-6 （IL-6）， tumor necrosis factor-α （TNF-α）， serum amyloid A （SAA）， and Aβ_1-42 in the hippocampal region from the colonic tissues. The lipopolysaccharide （LPS） concentrations in colon tissues of rats were measured by using a dynamic chromogenic limulus assay. ResultsCompared with those in the control group， the rats in the model group exhibited a significantly prolonged escape latency and a markedly shorter duration in the target quadrant （P<0.01）. The integrity of the colonic mucosal structure was compromised， with disordered gland arrangement and a reduced number of goblet cells. The Aβ_1-42 deposition in the hippocampal region was significantly increased （P<0.01）. The protein expression levels of TLR4 and NF-κB p65 in colonic tissues were significantly upregulated （P<0.01）， while those of occludin and ZO-1 were downregulated （P<0.01）. The contents of inflammatory factors such as IL-6， TNF-α， and SAA were significantly elevated （P<0.01）， and the LPS level in the serum was markedly increased （P<0.01）. In comparison to those in the model group， the rats in the Xixintang group showed a significantly shortened escape latency and a prolonged duration in the target quadrant （P<0.01）. The colonic mucosal structure was ameliorated， with neat gland arrangement and an increased number of goblet cells. The Aβ_1-42 deposition in the hippocampal region was reduced （P<0.01）. The protein expressions of TLR4 and NF-κB p65 in the colon tissues were decreased （P<0.05，P<0.01）， while the protein levels of occludin and ZO-1 were increased （P<0.01）. The contents of IL-6， TNF-α， and serum amyloid A （SAA） were decreased （P<0.01）， and the LPS level was reduced （P<0.01）. ConclusionXixintang can significantly ameliorate cognitive dysfunction of AD model rats， by means of restoring the colonic mucosal barrier structure， reducing cerebral Aβ deposition， and suppressing peripheral and central inflammatory response. Its mechanism of action may be closely associated with the suppression of the TLR4/NF-κB signaling pathway activation， reduction of endotoxin levels， and regulation of the gut-brain axis.

ObjectiveTo enhance teaching in postoperative cancer rehabilitation, this study developed an integrative Chinese-Western medicine postoperative oncology rehabilitation system, termed the medical oncology generative pre-trained transformer (MedOncoGPT). By introducing MedOncoGPT as an intelligent assistant, an integrated teaching model combining Chinese and Western medicine was established. The study evaluated its impact on students’ integrative clinical reasoning and practical abilities, providing support for instructional reform in related courses. MethodsUsing teaching resources as the knowledge base, MedOncoGPT was built upon the open-source ChatGLM model and incorporated Low-Rank Adaptation (LoRA) fine-tuning and retrieval-augmented generation (RAG) techniques to address postoperative integrative oncology scenarios. The system was applied in courses and clinical clerkships related to integrative oncology. In alignment with course objectives, a five-stage instructional process—pre-class preparation, in-class inquiry, simulated multidisciplinary consultation, clinical reinforcement, and teaching reflection—was designed to guide students in completing syndrome differentiation, comprehensive assessment, and follow-up planning within real or simulated case contexts. Comparative analyses of student engagement, syndrome differentiation thinking, evidence-based awareness, and interdisciplinary integration skills before and after the teaching reform were conducted using questionnaires, course assessments, classroom observations, and semi-structured interviews. ResultsFollowing the implementation of MedOncoGPT, students demonstrated improved performance in case analysis, prescription formulation, and integrative Chinese-Western medical evaluation compared with those receiving traditional instruction. Classroom participation and the relevance of student inquiries also increased. Self-assessment results indicated high levels of satisfaction with respect to clarity of integrative clinical reasoning, ability to retrieve and apply guideline-based evidence, and awareness of appropriate use of intelligent tools in clinical decision-making. More than 92% of students reported that the system facilitated understanding of abstract theoretical concepts presented in textbooks. Instructors noted that the system helped reduce lesson preparation time, enriched typical case materials and discussion scenarios, and promoted the translation of research findings into classroom teaching. Pilot data showed that, with MedOncoGPT assistance, the mean time for initial syndrome differentiation decreased from 18.4 min to 12.1 min, and the agreement rate increased from 68.3% to 82.5%. In the teaching pilot, the experimental group achieved a higher mean score on the final case analysis assessment than the control group (82.6 vs. 74.3). ConclusionThe integration of MedOncoGPT into teaching on postoperative integrative cancer rehabilitation enabled the establishment of a stable instructional process within existing curricula and enhanced students’ integrative clinical reasoning and evidence-based practice capabilities. The approach demonstrates positive potential for advancing the integration of research, clinical practice, and education and represents a valuable exploratory strategy for instructional reform in courses on integrative Chinese-Western medicine.

ObjectiveTo enhance teaching in postoperative cancer rehabilitation, this study developed an integrative Chinese-Western medicine postoperative oncology rehabilitation system, termed the medical oncology generative pre-trained transformer (MedOncoGPT). By introducing MedOncoGPT as an intelligent assistant, an integrated teaching model combining Chinese and Western medicine was established. The study evaluated its impact on students’ integrative clinical reasoning and practical abilities, providing support for instructional reform in related courses. MethodsUsing teaching resources as the knowledge base, MedOncoGPT was built upon the open-source ChatGLM model and incorporated Low-Rank Adaptation (LoRA) fine-tuning and retrieval-augmented generation (RAG) techniques to address postoperative integrative oncology scenarios. The system was applied in courses and clinical clerkships related to integrative oncology. In alignment with course objectives, a five-stage instructional process—pre-class preparation, in-class inquiry, simulated multidisciplinary consultation, clinical reinforcement, and teaching reflection—was designed to guide students in completing syndrome differentiation, comprehensive assessment, and follow-up planning within real or simulated case contexts. Comparative analyses of student engagement, syndrome differentiation thinking, evidence-based awareness, and interdisciplinary integration skills before and after the teaching reform were conducted using questionnaires, course assessments, classroom observations, and semi-structured interviews. ResultsFollowing the implementation of MedOncoGPT, students demonstrated improved performance in case analysis, prescription formulation, and integrative Chinese-Western medical evaluation compared with those receiving traditional instruction. Classroom participation and the relevance of student inquiries also increased. Self-assessment results indicated high levels of satisfaction with respect to clarity of integrative clinical reasoning, ability to retrieve and apply guideline-based evidence, and awareness of appropriate use of intelligent tools in clinical decision-making. More than 92% of students reported that the system facilitated understanding of abstract theoretical concepts presented in textbooks. Instructors noted that the system helped reduce lesson preparation time, enriched typical case materials and discussion scenarios, and promoted the translation of research findings into classroom teaching. Pilot data showed that, with MedOncoGPT assistance, the mean time for initial syndrome differentiation decreased from 18.4 min to 12.1 min, and the agreement rate increased from 68.3% to 82.5%. In the teaching pilot, the experimental group achieved a higher mean score on the final case analysis assessment than the control group (82.6 vs. 74.3). ConclusionThe integration of MedOncoGPT into teaching on postoperative integrative cancer rehabilitation enabled the establishment of a stable instructional process within existing curricula and enhanced students’ integrative clinical reasoning and evidence-based practice capabilities. The approach demonstrates positive potential for advancing the integration of research, clinical practice, and education and represents a valuable exploratory strategy for instructional reform in courses on integrative Chinese-Western medicine.