Traditional Chinese medicine (TCM) represents a paradigmatic approach to personalized medicine, developed through the systematic accumulation and refinement of clinical empirical data over more than 2000 years, and now encompasses large-scale electronic medical records (EMR) and experimental molecular data. Artificial intelligence (AI) has demonstrated its utility in medicine through the development of various expert systems (e.g., MYCIN) since the 1970s. With the emergence of deep learning and large language models (LLMs), AI's potential in medicine shows considerable promise. Consequently, the integration of AI and TCM from both clinical and scientific perspectives presents a fundamental and promising research direction. This survey provides an insightful overview of TCM AI research, summarizing related research tasks from three perspectives: systems-level biological mechanism elucidation, real-world clinical evidence inference, and personalized clinical decision support. The review highlights representative AI methodologies alongside their applications in both TCM scientific inquiry and clinical practice. To critically assess the current state of the field, this work identifies major challenges and opportunities that constrain the development of robust research capabilities-particularly in the mechanistic understanding of TCM syndromes and herbal formulations, novel drug discovery, and the delivery of high-quality, patient-centered clinical care. The findings underscore that future advancements in AI-driven TCM research will rely on the development of high-quality, large-scale data repositories; the construction of comprehensive and domain-specific knowledge graphs (KGs); deeper insights into the biological mechanisms underpinning clinical efficacy; rigorous causal inference frameworks; and intelligent, personalized decision support systems.
Medicine, Chinese Traditional/methods* ; Artificial Intelligence ; Humans ; Precision Medicine ; Decision Support Systems, Clinical

Objective To compare the clinical effect of left atrial appendage(LAA)plus atrial septal defect(ASD)closure therapy and ASD closure therapy in treating ASD associated with atrial fibrillation(AF).Methods A total of 102 patients with ASD complicated by non-valvular AF,who were admitted to the General Hospital of Northern Theater Command of China from January 2016 to December 2023,were enrolled in this study.Of the 102 patients,simultaneous one-stop interventional transcatheter LAA plus ASD closure was performed in 52(LAA+ASD closure group)and ASD closure was performed in 50.(ASD closure group).The perioperative and postoperative 30 d,90 d,180 d clinical safety and efficacy were compared between the two groups.Telephone follow-up was conducted,the complications such as embolization and bleeding were recorded,and the medium-to-long-term follow-up results were compared between the two groups.Results The immediate surgical success rate in both groups was 100％.The immediate postoperative monitoring showed that the occlusion effect was satisfactory.In LAA plus ASD closure group,LACBES LAA occluder was used in 27 patients and LAmbre LAA occluder was adopted in 25.There were no statistically significant differences in the patients' baseline characteristics between the two groups(all P＞0.05).In the LAA+ASD closure group,3 patients developed cardiac tamponade,among them 2 patients were cured after pericardiocentesis drainage and one patient was referred to the surgery department to receive occluder removal and intracardiac repair.Medium-to-long-term follow-up was conducted in 101 patients with a median follow-up period of 37.6 months.The incidence of embolic events in the LAA+ASD closure group was lower than that in the ASD closure group(3.9％vs.18.0％,P=0.028).The incidence of bleeding events in the ASD closure group was higher than that in the LAA+ASD closure group(16.00％vs.1.96％,P=0.016).Kaplan-Meier analysis indicated that the risk of occurring embolic events and bleeding events in the LAA+ASD closure group was strikingly lower than that in the ASD closure group(HR=4.295 and 7.888 respectively,95％CI:1.317-14.010 and 2.135-29.140 respectively,P=0.040 9 and P=0.020 8 respectively).Conclusion Simultaneous interventional transcatheter LAA plus ASD closure can effectively prevent embolic events such as stroke,etc.in patients with ASD complicated by AF,and its bleeding risk is lower than simple ASD closure.

Objective To identify the pathogenic bacteria such as Staphylococcus aureus and Enterococcus faecalis in bloodstream infec-tions with high confidence based on three deep learning models such as GoogleNet,ResNet101,and Vgg19,compare the performance and classification ability of these models,and explore the feasibility of applying the deep learning models for the rapid identification of pathogenic bacteria in bloodstream infections.Methods The preprocessed Gram-stained bacterial images,including 1 682 images for Staphylococcus aureus and 1 723 for Enterococcus faecalis,and 688 blank control microscopic images were input into three models for training and validation,respectively.Among them,1 344 images for Staphylococcus aureus,1 376 for Enterococcus faecalis,and 544 blank control images were used for training,and the remaining images were used for validation.The model with the best performance was identified according to the classification parameters between the models.Results The ResNet101 model had the lowest cross-en-tropy loss value(0.008 710 3),the largest Epoch value(93),and the highest accuracy rate(99％)for identifying the three types of validation set images.The cross-entropy loss value,Epoch value,and accuracy rate of the GoogleNet model were 0.063 89,86 and 98.6％,respectively,for identifying the three types of validation set images.Those of the Vgg19 model were 0.035 682,86 and 97.7％,respectively.Conclusion The ResNet101 model has the best performance in the classification of three kinds of images.The deep learning model may accurately,reliably and rapidly identify the Gram-stained images of pathogenic bacteria such as Staphylococcus aureus and Enterococcus faecalis in bloodstream infections.

Objective:To investigate the effectiveness of the modified Bikini approach in the fixation of acetabular fracture with an integrated wing-shaped anatomical locking plate.Methods:A retrospective study was conducted to analyze the data of 54 patients with acetabular fracture who had been treated at Department of Trauma Orthopedics, Orthopedic Medical Center, The Third Hospital Affiliated to Southern Medical University from May 2017 to June 2021. The patients were divided into 2 groups based on different surgical approaches: an observation group [26 cases, 6 males, 20 females; aged 40.0 (29.8, 46.8) years] treated with fixation with an integrated wing-shaped anatomical locking plate through the modified Bikini approach, and a control group [28 cases, 10 males, 18 females; aged 34.5 (24.0, 43.5) years] treated with fixation with an integrated wing-shaped anatomical locking plate through the lateral-rectus approach. The incision length, operation time, intraoperative bleeding, length of hospital stay, quality of postoperative fracture reduction, visual analog scale (VAS) for pain, hip function, Vancouver Scar Scale (VSS) and Patient and Observer Scar Assessment Scale (POSAS) were compared between the 2 groups.Results:The differences in the preoperative general data between the 2 groups were not statistically significant, indicating comparability ( P>0.05). There were no statistically significant differences between the 2 groups either in terms of incision length, operation time, intraoperative bleeding, or hospital stay ( P>0.05). The 2 groups were not significantly different in the excellent/good rate of fracture reduction [100.0% (26/26) versus 92.9% (26/28)], VAS at 1 month postoperation [2.0(1.0, 3.0) versus 2.0(1.0, 3.0)], or the modified Merle d'Aubigné and Postel hip score at 12 months postoperation [13.5(12.3, 14.8) versus. 14.0(13.0, 15.0)] ( P>0.05). However, the VSS [4.50(4.00, 6.00)] and POSAS (29.85±10.05) at 12 months postoperation in the observation group were significantly lower than those in the control group [6.50(5.00, 8.25) and 37.11±11.75] ( P<0.05). Conclusion:In the fixation of acetabular fracture with an integrated wing-shaped anatomical locking plate, the modified Bikini approach can not only achieve as fine early clinical efficacy as the lateral-rectus approach, but also demonstrate the aesthetic advantages of smaller incision scar and more hidden incision.

Objective:To evaluate our self-designed pubic symphysis orthotic compression anatomic plate (PSOCAP) in the treatment of fractures and dislocations around the pubic symphysis.Methods:A retrospective study was conducted to analyze the 16 patients with fracture or dislocation around the pubic symphysis who had been treated by our self-designed PSOCAP at Department of Traumatic Surgery, Center for Orthopaedic Surgery, The Third Hospital Affiliated to Southern Medical University from January 2021 to June 2022. There were 8 males and 8 females with an age of (41±19) years. According to the Tile classification for pelvic fractures, there were 3 cases of type B1, 2 cases of type B3, 1 case of type C1.2, 4 cases of type C1.3, 3 cases of type C2, and 3 cases of type C3. There were 8 cases of pubic symphysis separation and 8 fractures of the pubic ramus (2 ones at Nakatani zone Ⅰ and 6 ones at Nakatani zone Ⅱ). Time from injury to surgery was 16 (11, 53) days, ranging from 4 to 348 days. The fractures or dislocations around the pubic symphysis were exposed by the modified Stoppa approach, reduced with the assistance of PSOCAP and fixated with PSOCAP; the posterior pelvic ring was reduced and fixated by corresponding surgical methods. Recorded were the surgical time, intraoperative bleeding, postoperative quality of fracture reduction, surgical complications, and functional recovery at the last follow-up concerning the pelvic anterior ring.Results:Surgery went on successfully in the 16 patients. Their surgical time was (58±15) min, ranging from 40 to 90 min, and their intraoperative bleeding 85 (63, 150) mL, ranging from 50 to 250 mL. According to the Matta scoring, the fracture reduction was evaluated as excellent in 10 cases, as good in 3 cases and as fair in 3 cases. The (10±3)-month follow-up for the 16 patients revealed complete fracture union for all after (12±2) weeks. According to the Majeed scoring at the last follow-up, the pelvic function was evaluated as excellent in 5 cases, as good in 7 cases, and as fair in 4 cases. No such postoperative complications as fracture displacement or internal fixation failure occurred.Conclusion:Owing to the biplane and integrated structure, our self-designed PSOCAP can help reduce the fractures or dislocations around the pubic symphysis to achieve anatomical reduction and strong internal fixation, leading to good clinical efficacy.

Objective:To explore the surgical methods and treatment effects of adult anterior dislocation of the sacroiliac joint (AADSJ).Methods:A multi-center retrospective case series study was conducted to analyze the clinical data of 25 cases admitted in 5 clinical centers (affiliations of authors in this article) from January 2016 to January 2021. There were 18 males and 7 females, aged 38.8±15.5 years (range, 18-83 years). The AADSJ clinical classification system was formulated based on the radiographic morphology of anterior dislocation of the sacroiliac joint, which includes two types. Type I: complete anterior dislocation of the sacroiliac joint, and displacement of the entire iliac auricular surface to the front of the sacrum. Type II: fracture of the sacroiliac joint combined with anterior dislocation, subdivided into 3 subtypes. Type IIa: iliac fracture involves the anterior 1/3 of the sacroiliac joint, and dislocation of the ilium anterior to the sacrum. Type IIb: iliac fracture involves the posterior 2/3 of the sacroiliac joint, and dislocation of the ilium anterior to the sacrum. Type IIc: iliac fracture involves the posterior 2/3 of the sacroiliac joint, and dislocation of the ilium anteromedial to the sacrum. The reliability and repeatability of the clinical classification, Tile classification and Young-Burgess classification were performed based on the results of two-phase assessments in four observers. The operations were performed by the lateral-rectus approach and the ilioinguinal approach. The operation time and intraoperative bleeding were recorded. Pelvic X-ray and CT scan were rechecked after the operation. The quality of fracture reduction was evaluated according to Matta score. The postoperative functional rehabilitation was evaluated according to the Majeed rehabilitation standard at one-year follow-up.Results:Among 25 cases in this study, there were 3 cases of Type I, 5 cases of Type IIa, 9 cases of Type IIb and 8 cases of Type IIc according to the clinical classification system. The Kappa values of reliability tests for the clinical classification, Tile classification and Young-Burgess classification were 0.681, 0.328 and 0.383, respectively. The Kappa values of repeatability tests for the clinical classification, Tile classification and Young-Burgess classification were 0.690, 0.221 and 0.395, respectively. The reliability and repeatability of the AADSJ clinical classification were significantly better than other classifications. There were 14 cases underwent lateral rectus abdominis approach and 11 cases underwent ilioinguinal approach. The operative time for managing anterior dislocation of the sacroiliac joint was 122.0±50.7 min (range, 65-148 min) through the lateral rectus abdominis approach, and through the ilioinguinal approach was 178.0±49.9 min (range, 110-270 min), with a significant difference ( t=2.76, P=0.011). The amount of intraoperative blood loss through the lateral rectus approach was 680±330 ml (range, 350-2,120 ml), which was significantly less than that through the ilioinguinal approach (1,660±968 ml, 680-3,300 ml), with a significant difference ( t=3.55, P=0.002). The follow-up period was 1-3 years. At one week after surgery, the quality of fracture reduction evaluated by Matta score showed that the excellent and good reduction rate of the lateral-rectus approach was 79% (11/14), and that of the ilioinguinal approach was 73% (11/14), with no statistically significant difference ( P=1.000). At a one-year follow-up, according to Majeed's criteria, the overall excellent and good rate of the lateral-rectus approach was 64% (9/14), which is similar to 64% (7/11) of that of the ilioinguinal approach. No fracture reduction loss or internal fixation loosening failure occurred. Conclusion:The AADSJ clinical classification system can accurately describe the imaging features and clinical manifestations of AADSJ, with high reliability and repeatability. The AADSJ can be treated by the lateral-rectus approach or the ilioinguinal approach, with similar therapeutic effects but the former having less trauma.

Objective:To investigate the surgical method of LC-II screws for fragility fractures of the pelvis (FFP) in the elderly and evaluate its clinical efficacy.Methods:A retrospective analysis was performed on 45 patients with FFPs operated in our department from January 2011 to January 2022. The clinical information was as follows. The FFP classification of pelvic fracture was IIIA in 26 cases and IIIB in 19 cases. Among them, 22 cases were fixed with closed reduction and modified LC-II screws (experimental group), and 23 cases were fixed with open reduction and reconstruction plates (control group). In the experimental group, there were 6 males and 16 females. The age range was 62-90 years, with an average of 73.2±9.2 years. The FFP classification of pelvic fracture was IIIA in 12 cases and IIIB in 10 cases. In the control group, there were 8 males and 15 females. The age range was 60-87 years, with an average of 72.8±6.6 years. FFP classification of pelvic fracture was IIIA in 14 cases and IIIB in 9 cases. After admission, pelvic X-ray and CT scan were performed, and the surgery was prepared. In the experimental group, after closed reduction of the posterior ring, the modified LC-II screw was inserted below the anterior inferior iliac spine (AIIS) toward the sacroiliac joint and penetrated the sacroiliac joint. For combined anterior ring fractures, the INFIX was used for anterior ring fixation. In the control group, the posterior ring was fixed with a reconstruction plate and/or sacroiliac screw after open reduction through the lateral rectus approach (LRA). The clinical efficacy was evaluated between the experimental group and the control group.Results:All 45 patients were successfully operated and followed up for 6 months to 3 years. All the pelvic fractures healed. In the experimental group of 22 cases, the time from injury to operation was 3-9 days, with an average of 5.8±1.9 days; the operation time was 25-70 min, with an average of 42.0±12.9 min. The intraoperative bleeding was 20-40 ml, with an average of 29.1±6.7 ml. According to the X-ray reduction evaluation criteria of Matta, 7 cases were excellent, 11 cases were good and 4 cases were medium, with an excellent and good rate of 81.8%. According to rehabilitation criteria of Majeed, 10 cases were excellent, 6 cases were good and 6 cases were fair, with a total excellent and good rate of 72.7%. At the last follow-up, sacroiliac joint pain was evaluated by VAS score: 0 in 10 cases, <3 in 7 cases, and 4-6 in 5 cases. No internal fixation loosening occurred. In the control group, the time from injury to operation was 5-20 days, with an average of 9.9±3.8 days; the operation time was 50-150 min, with an average of 89.1±29.5 min; the intraoperative bleeding was 220-1 000 ml, with an average of 509.2±214.3 ml. According to the X-ray reduction evaluation criteria of Matta, 16 cases were excellent, 4 cases were good and 3 cases was medium, with an excellent and good rate of 87.0%. According to rehabilitation criteria of Majeed, 12 cases were excellent, 6 cases were good and 7 cases were fair, with a total excellent and good rate of 78.3%. At the last follow-up, sacroiliac joint pain was evaluated by VAS score: 0 in 14 cases, <3 in 6 cases, and 4-6 in 3 cases. In the control group, posterior ring plate loosening was found in 2 cases and anterior ring pubic ramus plate and screw loosening was found in 4 cases, but there was no reduction loss.Conclusion:The modified LC-II screw is theoretically feasible in the treatment of FFP. Preliminary clinical results show good safety and efficacy, providing a new idea for minimally invasive treatment of FFP.

Objective:To investigate the efficacy of intelligent visualization system (HoloSight Intelligent Visualization System) assisted reduction and screw fixation in the treatment of elderly pelvic fragility fracture.Methods:From January 2016 to December 2022, clinical data of 18 elderly patients aged over 75 years with pelvic fragility fractures surgically treated by our team were retrospectively analyzed. Among them, 11 cases were treated with closed reduction and hand-inserted percutaneous screw fixation to fix the posterior pelvic ring (control group), and 7 cases were treated with the assistance of the HoloSight intelligent visualization system for reduction and minimally invasive fixation (experimental group). In the control group, there were 3 males and 8 females with an age range of 75-94 years (mean age, 82±12 years). The pelvic fractures were classified as FFP type IIIa in 5 cases, IIIb in 2 cases, IIIc in 3 cases, and IV in 1 case. In the experimental group, there were 1 male and 6 females with an age range of 76-100 years (mean age. 83±14 years). The pelvic fractures were classified as FFP type IIIa in 4 cases, IIIc in 2 cases, and IV in 1 case. The surgical time, blood loss, effective intraoperative fluoroscopy times, fracture reduction quality (according to Matta standards), visual analogue scale (VAS), limb function rehabilitation (Majeed score) and postoperative complications were recorded and evaluated in both groups.Results:All patients underwent surgery successfully and were followed up for 6 months to 3 years(12 months on average), all pelvic fractures healed. Among the 7 cases of experimental group, a total of 13 screws were placed, and the adjustment times of guide pin for each screw were 3±1 times (range, 1-5 times), while 18 screws of the control group with the adjustment times of 7±2 times (range, 4-10 times), statistical difference was present ( t=6.99, P<0.001). The surgical time in experimental group (63±12 min) was shorter than that in control group 88±23 min, while effective intraoperative fluoroscopy times in experimental group (9±3 times) was less than that in control group (35±7 times), the difference were both statistically significant ( t=2.69, P=0.016; t=9.22, P<0.001). The intraoperative blood loss was 38±12 ml in the experimental group and 55±26 ml in control group, with no significant difference ( t=1.61, P=0.127). According to Matta's reduction criteria after surgery, the results of experimental group were excellent in 4 cases, good in 2, and fair in 1, while the result of control group were excellent in 5 cases, good in 3, fair in 2, and poor in 1. At the last follow-up, among the patients in experimental group, the Majeed score was 84±11, excellent in 3 cases, good in 2, and fair in 2. In control group, the score was 79±17, with excellent in 5 cases, good in 3, fair in 3. No statistical difference was observed in two groups ( t=0.69, P=0.501). The VAS was 6.4±2.6 preoperatively and 2.4±0.8 postoperatively in the experimental group, while in the control group was 6.9±3.1 preoperatively and 2.7±1.3 postoperatively, the data suggested an improvement in both groups ( t=3.89, P=0.002; t=4.14, P<0.001), while no statistical significant was observed in two groups at last follow-up ( t=0.55, P=0.593). Two cases of experimental group had loosening or withdrawal of the fixation screws in 9 and 12 months, but the fractures healed. Four cases of control group loose fixation screws in 4, 6, 9, and 12 months, two cases underwent revision surgery and other two cases healed. Conclusion:Intelligent visualization system assisted reduction and screw fixation in the treatment of elderly pelvic fragility fractures have the advantages of good reduction, accurate nail placement, less bleeding, less effective fluoroscopy, which improves the safety of surgery.

Objective:To investigate the safety and clinical efficacy of the lateral-rectus approach combined with the Pfannenstiel approach in the treatment of pelvic fractures complicated with urethral rupture.Methods:From January 2013 to June 2021, 20 patients with pelvic fracture complicated with urethral rupture were surgically managed through the lateral-rectus approach and the Pfannenstiel approach at Department of Traumatic Surgery, Center for Orthopaedic Surgery, the Third Hospital Affiliated to Southern Medical University. They were 15 males and 5 females, with an average age of 42 years (from 18 to 55 years). By the Tile classification, there were 11 cases of type B and 9 cases of type C. The first-stage urethral realignment was performed via the Pfannenstiel approach in the supine position after general anesthesia in conjunction with an urologist; at the second-stage, the lateral-rectus approach was used to reduce and fixate the acetabular or pelvic fractures. The operation time, intraoperative blood loss, fracture reduction quality, pelvic functional recovery and complications were documented.Results:In this cohort, the operation time ranged from 80 to 240 min, averaging 140.5 min; the time for simple urethral convergence ranged from 20 to 30 min; the intraoperative blood loss ranged from 400 mL to 2,000 mL, averaging 730 mL. According to the Mears andVelyvis evaluation for fracture reduction quality, anatomical reduction was achieved in 13 cases, satisfactory reduction in 6 cases, and unsatisfactory reduction in one. The 20 patients were followed up for 12 to 68 months (mean, 37 months) after surgery. One fracture got nonunited but the other fractures got united after 3.0 to 4.5 months (mean, 3.5 months). According to the Majeed scoring system, the pelvic function at 12 months after surgery was excellent in 12 cases, good in 6 and fair in 2, giving an excellent and good rate of 90% (18/20). Screw loosening was found in one patient, traction injury to the lumbosacral trunk nerve in another patient, varying degrees of dysuria which responded to periodic urethral dilation in 8 patients, urethral stricture in 3 patients and erectile dysfunction in 5 patients. No abdominal hernia or pelvic infection was observed.Conclusions:The lateral-rectus approach combined with the Pfannenstiel approach can be used effectively to reduce and fixate the pelvic and acetabular fractures, and to repair the urethral rupture in one stage as well. They are also safe due to a low incidence of such complications as abdominal wall hernia and pelvic infection.

Objective:To investigate the application of three-dimensional (3D) image fusion technique of pelvic enhanced CT and magnetic resonance neurography (MRN) on the patients of pelvic fractures with lumbosacral plexus injury.Methods:From January 2019 to December 2020, 15 patients (11 males, 4 females, mean age 26.67±11.34 of pelvic fracture (AO classification C1.3 of 11 cases, C2 of 2 cases, C3 of 2 cases) with lumbosacral plexus injury underwent 3D image fusion of pelvic enhanced CT and MRN. All patients exhibited lower limb nerve dysfunction after injury. The preoperative muscle strength were evaluated by British Medical Research Council (BMRC) criteria: grade 0 in 2 cases, grade 1 in 5, grade 2 in 7 and grade 3 in 1. Physical examination and enhanced CT combined with MRN 3D image fusion technology were used to accurately determine the injury site of lumbosacral plexus nerve, fracture displacement, and the direction of blood vessels and ureters, and develop corresponding surgical plans. The neurological functions were measured in post-operation follow-up.Results:All 15 operations were successfully completed and the site of lumbosacral plexus injury during operation was consistent with preoperative fusion image, and inconsistency of injury characteristic in 4 patients (4/15, 27%). Among the 6 cases of laparoscopic surgery, 2 cases were changed to open surgery because of the difficulty of nerve decompression. The average time of 4 cases of laparoscopic surgery was 116.27±26.46 min and intraoperative blood loss was 102.50±79.32 ml. The average time of 11 cases of open operation was 123.64±38.28 min, and intraoperative blood loss was 713.64±393.12 ml. For the opening operations, 9 cases were successfully decompressed, and nerve disruption was observed in 2 patients. All 15 patients were followed up. The average follow-up time was 9.33±2.19 months. The radial bony healing was observed in each case at the end of follow-up, mean healing time was 5.03±1.04 months (range, 3-6 months). In addition, patients' muscle strength recovered, 8 cases of grade 5, 3 cases of grade 4, 1 case of grade 3, 1 case of grade 2, and 2 cases of grade 0 after surgery. The difference was statistically significant ( Z=3.27, P<0.001). On the other hand, in sensory function assessment, there were 8 cases of grade S5, 1 case of grade S4, 3 cases of grade S3, 1 case of grade S2 and 2 cases of grade S0 after surgery. Conclusion:3D image fusion of enhanced CT and MRN can accurately locate and characterize pelvic fractures combined with lumbosacral plexus injury before surgery, accurately guide the formulation of surgical plan, and reduce surgical trauma and related complications.