OBJECTIVE To introduce the development of an intelligent prescription review decision-support system for anti-tumor drugs and assess its clinical application outcomes. METHODS Relevant data sources， including national and local pharmaceutical administration policies， clinical practice guidelines/consensus， hospital information systems data， and genetic testing results， were integrated. Adhering to the principles of structure， standardization and dynamic updating， a knowledge base covering chemotherapeutic， targeted and immunotherapeutic agents was constructed using a dual-dimensional modeling approach that combined “drug attributes” and “clinical contexts”. This knowledge base was then embedded into the hospital’s electronic medical order system to establish the prescription review decision-support system. The application and performance of the system were evaluated at Shanghai Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. RESULTS A knowledge base containing 18 318 prescription review rules for anti-tumor drugs was constructed， and a closed-loop prescription review system was successfully established， encompassing pre-prescription real-time intervention， in-process interactive review， and post-prescription evaluation and analysis. From 2021 to 2024， the system generated a total of 57 879 alerts for prescriptions of five typical categories of anti-tumor drugs. For platinum-containing prescriptions， 22 577 alerts were generated， with Cisplatin for injection （lyophilized） being the most frequently alerted drug （13 445 alerts）， and “ototoxicity risk due to combined use” alerts remained high （7 682 alerts）. For methotrexate-containing prescriptions， 3 721 alerts were recorded， primarily related to “precaution-related issues” （76.4%， 2 843/3 721）. For doxorubicin-containing prescriptions， 17 301 alerts were triggered， primarily related to “dosage and administration” （14 315 alerts）. For human epidermal growth factor receptor 2-targeted agents-containing prescriptions， 1 007 alerts were issued， mostly related to “reimbursement restrictions” （956 alerts）. For programmed death-1/programmed death-ligand 1 inhibitors-containing prescriptions， the alerts increased year by year， totaling 13 273 alerts， primarily related to “inappropriate indication” （9 118 alerts）. Over the 4 years， the physician response rates to system alerts were 21.4%， 27.1%， 33.5% and 51.6%， respectively. CONCLUSIONS An intelligent decision-support system for anti-tumor drug prescription review， encompassing a closed-loop process of “real-time pre-event intervention， interactive in-event prescription review， post-event evaluation and analysis”， has been successfully constructed and implemented throughout the entire workflow. There is a discernible trend in this hospital， where the focus on monitoring anti-tumor drugs is shifting towards immunotherapy drugs. Additionally， the acceptance rate of physicians regarding prescription review opinions has been steadily increasing year by year.

BACKGROUND:Clinical studies suggest that both bone quality and excessive crown-to-implant ratio are factors that affect the success rate of implant surgery,but there is no consensus on how large the crown-to-implant ratio under each bone quality will affect the prognosis of implant repair.OBJECTIVE:To analyze the stress and strain of bone tissue around implant restorations with different crown-to-implant ratios under different bone types after stress loading using three-dimensional finite element method.Combined with the Frost bone mechanics regulation system theory,the physiological limits of bone strain were observed for each bone type within a certain range of crown to implant ratios.METHODS:Cone beam CT data were selected from a patient to establish solid bone block models with four different bone types(class Ⅰ bone:cortical bone thickness 3 mm+dense cancellous bone,class Ⅱ bone:cortical bone thickness 2 mm+relatively dense cancellous bone,class Ⅲ bone:cortical bone thickness 1 mm+relatively low-density cancellous bone,class Ⅳ bone:cortical bone thickness 1 mm+low-density cancellous bone).Implant restorations with five different crown-to-implant ratios(1,1.5,2,2.5,and 3)were created on each bone block model,for a total of 20 models.Vertical and oblique forces were applied to the dental crown to observe the von Mises stress values,strains,and displacement of the cortical and cancellous bones,as well as the degree of implant displacement.RESULTS AND CONCLUSION:(1)In the class Ⅰ bone model,when the crown-to-implant ratio reached 3 under oblique loading,the cortical bone strain exceeded the physiological limit.In the class Ⅱ bone model,under oblique loading,when the crown-to-implant ratio reached 2.5,the cortical bone strain exceeded the physiological limit.In the class Ⅲ bone model,under oblique loading,cortical bone exceeded the physiological limit when the crown-to-implant ratio reached 2.5,while cancellous bone reached the critical physiological limit at 1.In the class Ⅳ bone model,under oblique loading,when the crown-to-implant ratio reached 1.5,2,2.5,and 3,the cortical bone strain exceeded the physiological limit.In the class Ⅳ bone model,all five crown-to-implant ratios of cancellous bone exceeded physiological limits.(2)Under vertical load,the cancellous bone strain exceeded the physiological limit when the crown-root ratio reached 1,2,2.5,and 3 in the class Ⅲ bone model,and when the cancellous bone strain of the five crown-root ratios in the class Ⅳ bone model exceeded the physiological limit.(3)Under oblique or vertical load,the implant displacement of the 20 groups of models did not exceed 100 μm.(4)From the perspective of biomechanics,when patients with insufficient bone height choose implant restoration,the crown-root ratio that class Ⅰ bone can tolerate is up to 2.5 times,the crown-root ratio that class Ⅱ bone can tolerate is up to 2 times,and the crown-root ratio that class Ⅲ and class Ⅳ bones can tolerate is up to 1 times due to excessive strain of cancellous bone;but the crown-root ratio that cortical bone in class Ⅲ bone can tolerate is up to 2 times.Whether class Ⅲ bone can tolerate implant restoration with a high crown-root ratio and whether cancellous bone can tolerate higher strains needs further study.

BACKGROUND:Clinical studies suggest that both bone quality and excessive crown-to-implant ratio are factors that affect the success rate of implant surgery,but there is no consensus on how large the crown-to-implant ratio under each bone quality will affect the prognosis of implant repair.OBJECTIVE:To analyze the stress and strain of bone tissue around implant restorations with different crown-to-implant ratios under different bone types after stress loading using three-dimensional finite element method.Combined with the Frost bone mechanics regulation system theory,the physiological limits of bone strain were observed for each bone type within a certain range of crown to implant ratios.METHODS:Cone beam CT data were selected from a patient to establish solid bone block models with four different bone types(class Ⅰ bone:cortical bone thickness 3 mm+dense cancellous bone,class Ⅱ bone:cortical bone thickness 2 mm+relatively dense cancellous bone,class Ⅲ bone:cortical bone thickness 1 mm+relatively low-density cancellous bone,class Ⅳ bone:cortical bone thickness 1 mm+low-density cancellous bone).Implant restorations with five different crown-to-implant ratios(1,1.5,2,2.5,and 3)were created on each bone block model,for a total of 20 models.Vertical and oblique forces were applied to the dental crown to observe the von Mises stress values,strains,and displacement of the cortical and cancellous bones,as well as the degree of implant displacement.RESULTS AND CONCLUSION:(1)In the class Ⅰ bone model,when the crown-to-implant ratio reached 3 under oblique loading,the cortical bone strain exceeded the physiological limit.In the class Ⅱ bone model,under oblique loading,when the crown-to-implant ratio reached 2.5,the cortical bone strain exceeded the physiological limit.In the class Ⅲ bone model,under oblique loading,cortical bone exceeded the physiological limit when the crown-to-implant ratio reached 2.5,while cancellous bone reached the critical physiological limit at 1.In the class Ⅳ bone model,under oblique loading,when the crown-to-implant ratio reached 1.5,2,2.5,and 3,the cortical bone strain exceeded the physiological limit.In the class Ⅳ bone model,all five crown-to-implant ratios of cancellous bone exceeded physiological limits.(2)Under vertical load,the cancellous bone strain exceeded the physiological limit when the crown-root ratio reached 1,2,2.5,and 3 in the class Ⅲ bone model,and when the cancellous bone strain of the five crown-root ratios in the class Ⅳ bone model exceeded the physiological limit.(3)Under oblique or vertical load,the implant displacement of the 20 groups of models did not exceed 100 μm.(4)From the perspective of biomechanics,when patients with insufficient bone height choose implant restoration,the crown-root ratio that class Ⅰ bone can tolerate is up to 2.5 times,the crown-root ratio that class Ⅱ bone can tolerate is up to 2 times,and the crown-root ratio that class Ⅲ and class Ⅳ bones can tolerate is up to 1 times due to excessive strain of cancellous bone;but the crown-root ratio that cortical bone in class Ⅲ bone can tolerate is up to 2 times.Whether class Ⅲ bone can tolerate implant restoration with a high crown-root ratio and whether cancellous bone can tolerate higher strains needs further study.

ObjectiveThe essence of syndrome manifestation recognition in traditional Chinese medicine (TCM) is to infer the body’s latent pathogenesis state from clinical observational information, rather than to perform simple label matching. However, previous studies have largely modeled this task as syndrome pattern classification within a fixed label space, which does not adequately reflect the cognition process of TCM syndrome differentiation centered on pathogenesis reasoning, and is also insufficient to capture the openness, semantic variability, and cross-disease reusability of syndrome manifestation expression. This study aimed to investigate whether introducing pathogenesis reasoning chain-of-thought (PR-CoT) supervision into large language models (LLMs) could improve the quality and cognitive consistency of syndrome manifestation recognition and support cross-disease transfer. MethodsSyndrome manifestation recognition was formulated as a conditional generation task under the framework of clinical observational information (X)→pathogenesis structure (Z)→syndrome pattern output (Y), where Z serves as an explicit intermediate structural variable linking the clinical evidence and syndrome judgment. Within this framework, a PR-CoT-supervised dataset for syndrome manifestation recognition was constructed based on medical case records of spleen-stomach disorders. After preprocessing, information extraction, manual proofreading, and data cleaning, the dataset comprised 4 800 training cases, 400 development cases, and 400 test cases. Each sample was annotated with a structured PR-CoT consisting of three progressive levels: clinical information summarization, comprehensive pathogenesis analysis, and syndrome pattern output. Supervised fine-tuning was conducted on open-source LLMs, with an end-to-end model serving as the baseline. Qwen3-32B was used as the primary experimental model, and Qwen3-14B as the scale comparison model. A progressive multidimensional evaluation framework was further established, comprising a structural parsing level, a semantic similarity level, and an expert blind review level. At the structural parsing level, syndrome pattern expressions were decomposed into structural elements and evaluated using Precision, Recall, F1 score, and Jaccard similarity. At the semantic similarity level, independent LLMs scored the theoretical proximity between predicted and reference syndrome patterns. At the expert blind review level, three TCM experts independently evaluated model outputs on two dimensions: syndrome differentiation consistency and terminology standardization of syndrome patterns. In addition, zero-shot cross-disease transfer evaluation was conducted on gynecological and heart-system disorder test sets. ResultsAt the structural parsing level, PR-CoT supervision did not lead to a stable improvement in the element-wise overlap of syndrome pattern structural components. Compared with the corresponding baselines, neither Qwen3-32B nor Qwen3-14B showed consistent advantages in structural matching metrics after the introduction of PR-CoT supervision. In contrast, at the semantic similarity level, PR-CoT supervision produced stable positive gains across different model scales and evaluation systems. The average semantic score of Qwen3-32B increased from 6.425 8 in the baseline model to 6.585 0 after PR-CoT supervision, and that of Qwen3-14B increased from 5.870 0 to 5.964 2. At the expert blind review level, the overall score of Qwen3-32B (PR-CoT) was 7.026 0±0.107 7, higher than 6.416 3±0.288 9 for its baseline. In zero-shot cross-disease testing, the PR-CoT model still showed advantages in semantic evaluation and expert evaluation on both gynecological and heart-system disorder test sets, indicating a certain degree of transferability. ConclusionThe benefits of PR-CoT supervision are mainly reflected in TCM semantic consistency and clinical plausibility, rather than in improved hard matching of structural elements. These findings support understanding syndrome manifestation recognition as a process of generating and expressing latent pathogenesis structures, rather than as a classification task within a traditional fixed label space. By introducing pathogenesis reasoning as an explicit intermediate structure into the modeling process and combining it with a progressive multidimensional evaluation framework, this study provides a methodological pathway for intelligent TCM syndrome differentiation that integrates theoretical alignment, interpretability, and multi-level evaluation.

ObjectiveThe essence of syndrome manifestation recognition in traditional Chinese medicine (TCM) is to infer the body’s latent pathogenesis state from clinical observational information, rather than to perform simple label matching. However, previous studies have largely modeled this task as syndrome pattern classification within a fixed label space, which does not adequately reflect the cognition process of TCM syndrome differentiation centered on pathogenesis reasoning, and is also insufficient to capture the openness, semantic variability, and cross-disease reusability of syndrome manifestation expression. This study aimed to investigate whether introducing pathogenesis reasoning chain-of-thought (PR-CoT) supervision into large language models (LLMs) could improve the quality and cognitive consistency of syndrome manifestation recognition and support cross-disease transfer. MethodsSyndrome manifestation recognition was formulated as a conditional generation task under the framework of clinical observational information (X)→pathogenesis structure (Z)→syndrome pattern output (Y), where Z serves as an explicit intermediate structural variable linking the clinical evidence and syndrome judgment. Within this framework, a PR-CoT-supervised dataset for syndrome manifestation recognition was constructed based on medical case records of spleen-stomach disorders. After preprocessing, information extraction, manual proofreading, and data cleaning, the dataset comprised 4 800 training cases, 400 development cases, and 400 test cases. Each sample was annotated with a structured PR-CoT consisting of three progressive levels: clinical information summarization, comprehensive pathogenesis analysis, and syndrome pattern output. Supervised fine-tuning was conducted on open-source LLMs, with an end-to-end model serving as the baseline. Qwen3-32B was used as the primary experimental model, and Qwen3-14B as the scale comparison model. A progressive multidimensional evaluation framework was further established, comprising a structural parsing level, a semantic similarity level, and an expert blind review level. At the structural parsing level, syndrome pattern expressions were decomposed into structural elements and evaluated using Precision, Recall, F1 score, and Jaccard similarity. At the semantic similarity level, independent LLMs scored the theoretical proximity between predicted and reference syndrome patterns. At the expert blind review level, three TCM experts independently evaluated model outputs on two dimensions: syndrome differentiation consistency and terminology standardization of syndrome patterns. In addition, zero-shot cross-disease transfer evaluation was conducted on gynecological and heart-system disorder test sets. ResultsAt the structural parsing level, PR-CoT supervision did not lead to a stable improvement in the element-wise overlap of syndrome pattern structural components. Compared with the corresponding baselines, neither Qwen3-32B nor Qwen3-14B showed consistent advantages in structural matching metrics after the introduction of PR-CoT supervision. In contrast, at the semantic similarity level, PR-CoT supervision produced stable positive gains across different model scales and evaluation systems. The average semantic score of Qwen3-32B increased from 6.425 8 in the baseline model to 6.585 0 after PR-CoT supervision, and that of Qwen3-14B increased from 5.870 0 to 5.964 2. At the expert blind review level, the overall score of Qwen3-32B (PR-CoT) was 7.026 0±0.107 7, higher than 6.416 3±0.288 9 for its baseline. In zero-shot cross-disease testing, the PR-CoT model still showed advantages in semantic evaluation and expert evaluation on both gynecological and heart-system disorder test sets, indicating a certain degree of transferability. ConclusionThe benefits of PR-CoT supervision are mainly reflected in TCM semantic consistency and clinical plausibility, rather than in improved hard matching of structural elements. These findings support understanding syndrome manifestation recognition as a process of generating and expressing latent pathogenesis structures, rather than as a classification task within a traditional fixed label space. By introducing pathogenesis reasoning as an explicit intermediate structure into the modeling process and combining it with a progressive multidimensional evaluation framework, this study provides a methodological pathway for intelligent TCM syndrome differentiation that integrates theoretical alignment, interpretability, and multi-level evaluation.

Background:To compare the efficacy and safety of monthly administrations of gonadotropin releasing hormone (GnRH) agonists LY01005 and Zoladex ? in Chinese patients with premenopausal breast cancer. Methods:From October 2020 to November 2021, 188 premenopausal breast cancer patients were enrolled in 34 hospitals and randomized 1:1 to receive either LY01005 or Zoladex ? every 28 days for a total of three injections. All patients concomitantly received oral tamoxifen (TAM). The primary efficacy endpoint was cumulative probability of maintaining menopausal level [oestradiol (E2) ≤30 pg/ml] from day 29 to day 85. The second efficacy endpoint included changes in E2, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) compared with the baseline. Pharmacokinetics (PK), pharmacodynamics (PD), and safety were analyzed. The study also evaluated the pharmacokinetic and pharmacodynamic characteristics of LY01005. Results:A total of 188 patients were randomised and 187 patients received either LY01005 or Zoladex ?. Cumulative probabilities of maintaining menopausal level (E2≤30 pg/ml) from day 29 to day 85 were 93.1% for LY01005 and 86.3% for Zoladex ?. The between-group difference was 6.8% (95% CI: -2.3%, 15.9%) and primary efficacy in the LY01005 group was not inferior to that in the Zoladex ? group. Changes in E2, LH, and FSH levels compared with the baseline were equivalent between the two groups (E2: 89.34% to 90.23% vs. 82.11% to 85.02%; LH: 88.89% to 95.52% vs. 89.70% to 97.02%; FSH: 75.36% to 80.85% vs.73.07% to 80.24%, respectively). After three consecutive doses of LY01005, the LH and FSH levels of the subjects showed a transient increase after the first dose, reached a peak on the second day and then started to decrease. The LH and FSH reached a lower level and remained at or below that level until the 85th day. Both treatments were well-tolerated. Conclusion:LY01005 is as effective as Zoladex ? in suppressing E2 to menopausal levels in Chinese patients with premenopausal breast cancer, with a similar safety profile.

This study aimed to identify mutations in the human insulin receptor gene (INSR) and investigate their role in the pathogenesis of severe insulin resistance syndrome. Sanger sequencing of the INSR gene was performed on a patient clinically suspected of having type A insulin resistance syndrome admitted to the Department of Endocrinology, the First Medical Center of Chinese PLA General Hospital. Upon identifying mutations, relevant exons were sequenced in her first-degree relatives. Additionally, control groups consisting of individuals with type 2 diabetes and those with normal glucose tolerance were screened for the mutation detected in the patient. Functional predictions of the INSR protein were made using MutationTaster, SIFT, and PolyPhen2 software. A previously unreported heterozygous missense mutation, c.3652G/A (Asp1218Asn), in exon 20 was identified in both the proband and her father. This mutation was not present in any of the control individuals. Multiple prediction tools indicate that this mutation likely disrupts gene/protein structure or function. The c.3652G/A (Asp1218Asn) heterozygous mutation in INSR is a novel variant that plays a significant role in the pathogenesis of severe insulin resistance in this Chinese family.

Objective:To apply artificial intelligence（AI）in classifying ovarian tumors on ultrasound images，and compare the diagnostic results of several sonographers with varying seniority levels.Methods:A total of 645 patients diagnosed with adnexal masses via gynecological ultrasound examination at Qilu Hospital of Shandong University from January 2021 to December 2024 were enrolled. Three deep learning architectures，i.e.，Alexnet，Densenet121，and Resnet50 were developed and used to internally test the classification effectiveness of ovarian tumors，while the optimal model was selected for external testing. Two junior sonographers and two senior sonographers were recruited to independently diagnose ovarian tumors in the external test dataset. Subsequently，the benign and malignant results of the model's predictions were disclosed to each sonographer，and their revised diagnoses on the same external test data in combination with the best AI model were recorded.Results:The optimal model achieved an accuracy of 0.941，sensitivity of 0.936，and specificity of 0.944 on the internal test dataset，and maintained robust performance on the external test dataset with accuracy of 0.891，sensitivity of 0.880，and specificity of 0.907. Compared to junior sonographers，the optimal model demonstrated significantly higher sensitivity in discriminating benign from malignant ovarian tumors（0.880 vs. 0.723，0.602；all P<0.05）. No statistically significant difference was observed in diagnostic accuracy between the optimal model and senior sonographer 1（ P=0.05）. With assistance from the optimal model，junior sonographers achieved significant improvements in both sensitivity and specificity（sensitivity：0.723 vs. 0.843，0.602 vs. 0.819；specificity：0.778 vs. 0.833，0.685 vs. 0.741；all P<0.05）. Conclusions:The optimal model achieves comparable performance to that of senior sonographers in ovarian tumor classification. With model assistance，the diagnostic performance of junior sonographers is significantly improved.

Objective To investigate the prognostic value of the platelet-to-lymphocyte ratio(PLR)at different early time points in adult patients undergoing veno-arterial extracorporeal membrane oxygenation(VA-ECMO).Methods A retrospective study was conducted,selecting 55 adult patients who underwent VA-ECMO treatment at the First Hospital of Jiaxing from June 2020 to October 2022 as the study subjects.Then,the patients'gender,age,past history[including hypertension,diabetes,heart disease,chronic obstructive pulmonary disease(COPD)],and the reason for extracorporeal membrane pulmonary oxygenation(ECMO)adjuvant therapy[including severe myocarditis,acute myocardia infarction,in-hospital and out-of-hospital cardiac arrest,severe closed craniocerebral injury,severe pneumonia,pelvic fracture,other(pulmonary embolism,electrocution,traumatic hepatic rupture,post-partum hemorrhage,severe acute pancreatitis,crush syndrome)],acute physiology and chronic health evaluationⅡ(APACHEⅡ),sequential organ failure assessment(SOFA)at the time of admission,and ECMO peripheral blood tests[creatinine,alanine aminotransferase(ALT),aspartate aminotransferase(AST),blood lactate acid(Lac),white blood cell count(WBC),neutrophil count(NEU),lymphocyte count(LYM),hemoglobin(Hb),and platelet count(PLT)]and the last time prior to ECMO assistance,24 hours prior to the occurrence of acute kidney injury(AKI),and 24 hours after the occurrence of AKI.PLR levels at 24 hours ECMO,and the proportion of continuous renal replacement therapy(CRRT).The patients were divided into a death group and a survival group based on their 30-day prognosis and further categorized into a CRRT group and a non-CRRT group based on whether CRRT was administered.Clinical indicators of patients with different prognosis and the differences in PLR levels between CRRT and non-CRRT groups were compared.Logistic regression analysis was used to identify independent risk factors affecting the 30-day prognosis of VA-ECMO patients.The receiver operator characteristic(ROC curves)were plotted to evaluate the prognostic predictive value of each risk factor.Results Compared to the survival group,the death group had significantly higher APACHEⅡscores,SOFA scores,LYM and proportion receiving CRRT[APACHEⅡscore:34.00(28.50,36.00)vs.25.00(14.75,34.00),SOFA score:5.00(4.00,6.50)vs.3.00(2.00,5.25),LYM(×109/L):3.40±1.97 vs.2.24±2.11,proportion receiving CRRT:91.30%(21/23)vs.62.50%(20/32)],and a significantly lower level of the last PLR prior to ECMO adjuvant[30.00(21.06,48.17)vs.58.82(41.80,145.72)],and the differences were statistically significant(all P<0.05).Logistic regression analysis showed that the levels of the last PLR before ECMO assistance[odds ratio(OR)=0.965,95%confidence interval(95%CI)was 0.938-0.993,P=0.013],APACHEⅡscore at the time of admission(OR=1.121,95%CI was 1.018-1.234,P=0.020),and CRRT(OR=7.734,95%CI was 1.042-57.401,P=0.045)were independent risk factors affecting the prognosis of the VA-ECMO patients at 30 days after adjuvant;the ROC curve analysis showed that APACHEⅡscore,CRRT and the last PLR level before ECMO assistance had a predictive value for the prognosis of VA-ECMO patients 30 days after assistance,in which the APACHEⅡscore+the last PLR level before ECMO assistance had the greatest predictive value in predicting the prognosis of the patients,with area under the curve(AUC)of 0.846,with a sensitivity of 62.5%and a specificity of 95.7%.Higher early PLR levels were associated with better prognosis.In the CRRT group,PLR levels at 24 hours before ECMO initiation,24 hours before AKI onset,and 24 hours after AKI onset were significantly lower than those in the non-CRRT group(all P<0.05).Conclusion Early PLR levels and CRRT administration have significant predictive value for the prognosis of patients undergoing VA-ECMO therapy.

Objective:To investigate the clinical effect of indocyanine green angiography (ICGA)-assisted design and harvest of expanded flaps for scar reconstruction.Methods:This study was a retrospective observational study. From April 2019 to August 2023, 19 patients with scars (8 males, 11 females; aged 3-38 years) treated at the Plastic Surgery Hospital of Peking Union Medical College and Chinese Academy of Medical Sciences met the inclusion criteria. The scars were distributed on the head, face, trunk, and extremities. In stage Ⅰ surgery, skin soft tissue expanders were implanted in suitable areas around the scars for skin soft tissue expansion. In stage Ⅱ surgery, the scar tissue was excised, resulting in wound areas ranging from 100 to 210 cm 2, and expanded flaps were designed. ICGA was used to identify target perforators and their accompanying veins, and the flap design was adjusted to ensure the inclusion of complete arterial and venous axes. The expanded flap with an area of 120 to 240 cm2 was harvested using unilateral back-cut technique and transferred to the recipient site, and the donor site wound was sutured directly. The durations of the arterial and venous phases of ICGA during flap design were recorded. The length-to-width ratios of the back-cut flaps were calculated for different regions. After stage Ⅱ surgery, the blood perfusion and survival of the flap, the wound healing at the donor site, and the occurrence of complications were observed. During follow-up, the appearance, color, and texture of the patient's flap were observed. Results:The arterial phase of ICGA lasted 10-27 (18±5) s, and the venous phase lasted 78-116 (100±10) s. The length-to-width ratios of the back-cut flaps were 1.22±0.32, 1.63±0.12, and 1.15±0.21 for the head and neck, trunk, and limb regions, respectively. After stage Ⅱ surgery, one patient had a large area of insufficient blood perfusion in the flap. By comparing ICGA images before and after flap transfer, the sutures at the oral commissure were loosened, the blood flow of the flap was restored. The blood perfusion of the flaps in other patients was good. All flaps survived completely, with well-healed donor site wounds and no complications. During 0.5-14.0 months of follow-up, all flaps of patients demonstrated excellent appearance, with color and texture matching the surrounding skin.Conclusions:As a means of superficial blood flow visualization, ICGA can not only clearly show the microvascular distribution of the expanded flap before operation, assist in optimizing the design of the flap, but also evaluate the blood perfusion of the flap after operation, reduce the occurrence of complications, and provide a full-process navigation for the harvesting of expanded flaps, thereby improving the safety of flap transfer for scar reconstruction.