With the rapid development of artificial intelligence, computational pathology has been seamlessly integrated into the entire clinical workflow, which encompasses diagnosis, treatment, prognosis, and biomarker discovery. This integration has significantly enhanced clinical accuracy and efficiency while reducing the workload for clinicians. Traditionally, research in this field has depended on the collection and labeling of large datasets for specific tasks, followed by the development of task-specific computational pathology models. However, this approach is labor intensive and does not scale efficiently for open-set identification or rare diseases. Given the diversity of clinical tasks, training individual models from scratch to address the whole spectrum of clinical tasks in the pathology workflow is impractical, which highlights the urgent need to transition from task-specific models to foundation models (FMs). In recent years, pathological FMs have proliferated. These FMs can be classified into three categories, namely, pathology image FMs, pathology image-text FMs, and pathology image-gene FMs, each of which results in distinct functionalities and application scenarios. This review provides an overview of the latest research advancements in pathological FMs, with a particular emphasis on their applications in oncology. The key challenges and opportunities presented by pathological FMs in precision oncology are also explored.
Humans ; Precision Medicine/methods* ; Medical Oncology/methods* ; Artificial Intelligence ; Neoplasms/pathology* ; Computational Biology/methods*

Objective To analyze dynamic contrast-enhanced MRI (DCE-MRI) radiomic features using machine learning algorithms, and to develop and validate a predictive model for HER-2 status in breast cancer. Methods The DCE-MRI images of 272 treatment-naive female patients with breast cancer between 2020 and 2022 were included in this study. Regions of interest (ROIs) were manually segmented using 3d-Slicer software, and radiomic features were extracted. All patients were randomly divided into training sets or validation sets at a ratio of 4∶1. The least absolute shrinkage and selection operator (LASSO) algorithm was used for feature screening on the training set, followed by the development of predictive models using six machine learning algorithms. Internal cross-validation was performed to compare the performance differences between the models. The best-performing model was selected, trained on the training set, and evaluated on the validation set. Evaluation metrics included area under the curve (AUC), sensitivity, specificity, precision, and recall rate. Results The clinical data of patients in the training set and validation set showed no significant differences. Five features were identified by the LASSO algorithm. With these features, six machine learning models were developed on the training set, and their predictive performance was internally cross-validated using the bagging method. XGBoost model had the highest mean AUC (0.696), followed by RF model (0.690); XGBoost model had the highest mean precision (0.756), followed by LR and RF models. Therefore, XGBoost was the optimal model. An HER-2 predictive model was built using the XGBoost algorithm on the training set and applied to the validation set. The AUC, precision, sensitivity, and specificity of the predictive model on the validation set were calculated, and ROC curves, precision-recall curves, calibration curves, and decision-making curves were plotted. Conclusion This study constructed and evaluated different DCE-MRI radiomics-based machine learning models for predicting HER-2 status in breast cancer. Among them, XGBoost algorithm performed the best and has the potential to become a new non-invasive method for preoperative prediction of HER-2 status, providing reliable evidence for personalized clinical diagnosis and treatment.

Objective:To explore the prevalence and independent associated factors of vascular calcification (VC) in non-dialysis chronic kidney disease (CKD) patients of stage 3-5.Methods:It was a single-center cross-sectional observational study. Non-dialysis stage 3-5 CKD patients ≥18 years old who were admitted to the Department of Nephrology, the First Affiliated Hospital of Sun Yat-sen University from May 1, 2022 to December 31, 2022 with VC evaluation were enrolled. The patients' general information, laboratory examination and imaging data were collected. Coronary artery calcification (CAC), thoracic aorta calcification (TAC), abdominal aorta calcification (AAC), carotid artery calcification and aortic valve calcification (AVC) were evaluated by cardiac-gated electron-beam CT (EBCT) scans, lateral lumbar x-ray, cervical macrovascular ultrasound and echocardiography, respectively. The differences in clinical data and the prevalence of VC at different sites of patients with different CKD stages were compared, and the prevalence of VC at different sites of patients in different age groups [youth group (18-44 years old), middle-aged group (45-64 years old) and elderly group (≥65 years old)] and patients with or without diabetes were compared. Multivariate logistic regression analysis was used to analyse the independent associated factors of VC for different areas.Results:A total of 206 patients aged (51±14) years were included, including 129 (62.6%) males. There were 44 patients with CKD stage 3 (21.4%), 51 patients with CKD stage 4 (24.8%), and 111 patients with CKD stage 5 (53.9%). CKD was caused by chronic glomerulonephritis [104 cases (50.5%)], diabetic kidney damage [35 cases (17.0%)], hypertensive kidney damage [29 cases (14.1%)] and others [38 cases (18.4%)]. Among 206 patients, 131 (63.6%) exhibited cardiovascular calcification, and the prevalence of CAC, TAC, AAC, carotid artery calcification, and AVC was 37.9%, 43.7%, 37.9%, 35.9% and 9.7%, respectively. The overall prevalence of VC in young, middle-aged and elderly patients was 24.6%, 73.6% and 97.4%, respectively. With the increase of age, the prevalence of VC in each site gradually increased, and the increasing trend was statistically significant (all P<0.001). The overall prevalence of VC in CKD patients with diabetes was 92.5% (62/67), and the prevalence of VC at each site in the patients with diabetes was significantly higher than that in the patients without diabetes (all P<0.001). Multivariate logistic regression analysis revealed that age (every 10 years increase, OR=2.51, 95% CI 1.77-3.56, P<0.001), hypertension ( OR=5.88, 95% CI 1.57-22.10, P=0.009), and diabetes ( OR=4.66, 95% CI 2.10-10.35, P<0.001) were independently correlated with CAC; Age (every 10 years increase, OR=6.43, 95% CI 3.64-11.36, P<0.001) and hypertension ( OR=6.09, 95% CI 1.33-27.84, P=0.020) were independently correlated with TAC; Female ( OR=0.23, 95% CI 0.07-0.72, P=0.011), age (every 10 years increase, OR=3.90, 95% CI 2.42-6.29, P<0.001), diabetes ( OR=5.37, 95% CI 2.19-13.19, P<0.001) and serum magnesium ( OR=0.01,95% CI 0-0.35, P=0.014) were independently correlated with AAC. Moreover, age and diabetes were independently correlated with carotid artery calcification, AVC and overall VC Conclusions:The prevalence of VC in non-dialysis CKD patients of stage 3-5 is 63.59%, of which CAC reaches 37.9%, TAC is the most common one (43.7%), while AVC is the least one (9.7%). Age and diabetes are the independent associated factors for VC of all sites except TAC, while hypertension is an independent associated factor for both CAC and TAC.

Objective:To describe the current situation of mental health service utilization of community pa-tients with five mental disorders in Inner Mongolia Autonomous Region and provide reference for health education and formulating relevant policies.Methods:The multistage stratified sampling method with unequal probability was used to select a total of 12 315 community residents aged 18 and over in Inner Mongolia Autonomous Region.Using Composite International Diagnostic Interview,mood disorders,anxiety disorders,substance use disorders,intermit-tent explosive disorders,and eating disorders,and health service utilization were investigated.Descriptive statistics was completed by single factor analysis method.Results:The lifetime rates of consultation and treatment of any mental disorder were 18.7％and 10.2％,respectively.The highest proportion of patients received treatment by non-medical professionals was 31.4％,followed by psychiatrists in psychiatric hospital or psychologists in general hospitals.Among the patients,1.1％of them received medication,and 2.5％received psychotherapy.Conclusion:The utilization rate of mental health services in community patients with five mental disorders is relatively low.It is necessary to conduct health education for medical help seeking properly.

Objective:To explore the changes of large-scale functional brain networks and network topological properties in patients with major depressive disorder (MDD) whose diagnosis had not changed after 5 years of follow-up.Methods:Totally 521 cases of hospitalized MDD patients were recruited from January 2012 to August 2018, and another 204 cases of gender- and age-matched healthy controls were recruited. All participants completed resting-state functional magnetic resonance scanning and clinical assessment. Their diagnosis were reviewed 5 years after discharge.A total of 258 participants whose diagnosis had not changed were counted into the MDD group for analysis. The differences in large-scale brain network connectivity between the two groups were analyzed by constructing a whole-brain functional network, on the basis of which the altered topological properties of the sensorimotor network (SMN), visual network (VN) and default mode network (DMN) were further analyzed between the two groups.The SPSS 24.0 software was used for data analysis and the independent sample t-test and χ2 test were used for the data comparison of the two groups. Results:Compared with the healthy controls, the MDD group had significantly decreased network clustering, mainly involving the SMN, VN and DMN (edge P<0.001, cluster P<0.05). The MDD group had decreased functional connectivity(FC) strength within the SMN, VN and DMN networks, the FC strength between the SMN and VN networks, between the frontoparietal network (FPN) and the DAN networks were decreased(all P<0.05, FDR corrected). Graph-theory analysis showed that local efficiency, clustering coefficient, and normalized shortest path length were decreased in the MDD group, node efficiency was decreased in the left ventral medial prefrontal cortex and the middle of the bilateral insula, node centrality was decreased in the middle of the bilateral insula and occipital lobe, and the betweenness was decreased in the middle of the right insula (all P<0.05, FDR corrected). Conclusion:MDD exhibits abnormal network functional connectivity, disruption of network topological properties, diminished optimal information processing, and to some extent reflects the severity of depressive symptoms. The decreased ability of information transfer flow in the insula plays an important role for the functional abnormality of the network.

Objective:To investigate the effects of glycopyrrolate on intestinal spasm and hemodynamics in painless colonoscopy.Methods:A total of 100 patients who were scheduled to undergo painless colonoscopy were selected as the study subjects and randomly divided into two groups by a computerized number method. Ten patients in both groups dropped out because of disruption of the study protocol, and 45 patients from each group were included in the final analysis. Before anesthesia induction, patients in group glycopyrrolate (group G) were injected with 0.2 mg glycopyrrolate, while those in congtrol group (group C) were injected with an equal amount of saline. The heart rate, systolic blood pressure, and diastolic blood pressure were recorded at T 0 (baseline period), T 1 (after anesthesia induction), T 2 (colonoscopy over sigmoid colon), T 3 (colonoscopy over the liver region), T 4 (after the end of examination), and T 5 (at the awakening phase), and the degree of intestinal spasm was assessed intraoperatively using the Likert′s four-point scale. The numerical rating scale (NRS) was used to assess preoperative and postoperative pain. The incidence of adverse events was recorded. Results:The general data at baseline were not statistically different between the two groups ( P>0.05). During the procedure, patients in group G had lower intraoperative intestinal spasm scores than those in group C ( P=0.028). Intraoperative hypotension and bradycardia occurrence were lower in group G than in group C ( P<0.05), and intraoperative norepinephrine use was also lower than in the group C ( P=0.034). Postoperative visual analog scale pain scores were lower in group G ( P=0.047), but patients who used glycopyrrolate had a higher proportion of dry mouth ( P=0.035). Conclusion:During painless colonoscopy, preoperative administration of glycopyrrolate significantly improved intraoperative hemodynamic fluctuations, reduced the incidence of hypotension and bradycardia, and relieved postoperative pain. However, glycopyrrolate use resulted in the risk of dry mouth.

Many studies have found a correlation between suicidal behavior (SB) and anhedonia, the main symptom of depression, in terms of both psychological and neurophysiological findings. The purpose of this review is to find the relationship between the two neuroimaging mechanisms, and to provide help for the future study of how the brain imaging changes can promote the mechanism of SB in depression patients with anhedonia symptoms. This review also emphasizes the necessity of intervention for the symptoms of anhedonia when preventing depression from committing suicide. The latest research results were reviewed about anhedonia in depression and magnetic resonance imaging of SB.The results showed that the default network, insula, lateral orbitofrontal gyrus, anterior cingulate gyrus, ventral striatum gyrus, ventral lateral and dorsolateral prefrontal gyrus, thalamus and habenula nucleus were dysfunction in depression with state anhedonia symptoms, affecting SB in terms of mood, execution, reward and aversion processing, especially the low lethal SB.

Objective:To analyze the compliance with enhanced recovery after surgery (ERAS) protocol in geriatric patients with fresh fracture.Methods:A retrospective study was conducted on the data of the patients with fresh extremity fracture which had been included in the ERAS perioperative protocol database during May 2019 and January 2022 at Department of Orthopaedic Trauma, Beijing Jishuitan Hospital. The patients ≥65 years were selected as a study group which was matched by a control group of the patients < 65 years in sex, fracture type and date frame of hospitalization at a ratio of 1∶1. The 2 groups were compared in the compliance with the 14 ERAS core perioperative elements.Results:The study group and the control group each included 66 patients who were matched in sex and fracture type. 62.1% (41/66) of the patients in the study group had combined diseases, significantly more than that [16.7% (11/66)] in the control group( P<0.001). Altogether, the compliance with the 14 ERAS core perioperative elements was 78.6 (71.4, 85.7) % in both groups, showing no significant difference between them ( P>0.05). Respectively, the compliance with the postoperative oral intake in the study group (80.3%, 53/66) was significantly lower than that in the control group (92.4%, 61/66) ( P<0.05); the compliance with the other 13 elements showed no statistically significant difference between the 2 groups ( P>0.05). Conclusion:The ERAS perioperative protocol can be carried out smoothly in geriatric patients with fresh fracture whose compliance may be comparable to that of the none-elderly patients.

Objective:To investigate the molecular epidemiological characteristics of HIV-1 infection among men who have sex with men (MSM) in Jiaxing city, and provide methods and ideas for the prevention and control of HIV-1 infection.Methods:This study retrospectively collected the blood samples from all newly reported cases of HIV-1 infection among MSM without antiviral treatment from 2020 to 2022. HIV-1 pol genes in the blood samples were amplified and sequenced. MEGA v6.0 software was used to analyze nucleic acid sequences. A phylogenetic tree was constructed to analyze HIV-1 subtypes. The calibrated population resistance program (CPR) was used to detect drug-resistant mutations. After calculating the genetic distance between gene sequences, molecular transmission networks were constructed using Cytoscape v3.6.0 software. Results:A total of nine genetic subtypes were identified, with CRF07_BC (43.3%) and CRF01_AE (36.9%) accounting for the most. The recombinant forms that were not clustered with the reference subtype accounted for 5.0%. Drug-resistant mutations were identified in 21 cases (7.0%), and the mutation rates among strains of CRF07_BC and CRF01_AE subtypes were 8.2% and 7.8%, respectively. The detection rates of drug-resistant mutations to protease inhibitors, nucleoside reverse transcriptase inhibitors and non-nucleoside reverse transcriptase inhibitors were 1.7%, 0.7% and 4.7%, respectively. The optimal genetic distance threshold of the molecular network was 0.018. At this genetic distance, the molecular network access rate was 43.3%, and 38 molecular clusters (ranging from 2 to 26 nodes per cluster) were included. Cases with ≥6 homosexual partners and a degree value of ≥4 were more likely to access the active molecular cluster. The population of high-risk transmission cases with TNS>0.75 were mainly nonlocal residents and commercial service providers. The recombinant forms showed high similarity in sequences with the strains from other provinces.Conclusions:The distribution of HIV-1 subtypes in MSM in Jiaxing city was complex. Given the increasing detection rate of recombinant forms, the high transmission rate of drug-resistant mutations, and the active molecular clusters and high-risk transmission clusters mainly detected in nonlocal residents and commercial service providers, strengthened surveillance and intervention are needed.

Objective:To observe the effect of scleral buckling surgery (SB) in the treatment of rhegmatogenous retinal detachment (RRD) with subretinal hyperplasia (SRP).Methods:A retrospective case study. From January 2016 to December 2018, 31 patients with old RRD with SRP who were treated with SB in Department of Ophthalmology, Central Theater Command General Hospital were included in the study. There were 18 males with 20 eyes and 13 females with 15 eyes. Age was (26.5±8.7) years. The course of disease was (12.6±10.3) months. The best corrected visual acuity (BCVA) test was performed using the international standard visual acuity chart, which was converted to logarithm of the minimum angle of resolution (logMAR) visual acuity at the time of recording. Retinal detachment ranges ≤2, >2-<3, ≥3 quadrants were 10 (28.6%, 10/35), 20 (57.1%, 20/35), and 5 (14.3%, 5/35) eyes, respectively. All affected eyes were treated with SB. Among them, 22 eyes (63.0%, 22/35) underwent local Scleral buckling, 11 eyes (31.4%, 11/35) underwent combined encircling buckle, and 2 eyes (5.7%, 2/35) underwent encircling buckle alone. Subretinal fluid drainage was performed in 33 eyes (94.3%, 33/35). The mean follow-up time was 18.2 months. Relevant examinations were performed with the same equipment and methods before operation to observe BCVA and retinal reattachment. Paired sample t test was used to compare logMAR BCVA before and after operation. Results:At the last follow-up, retinal reattachment occurred in 32 eyes (91.4%, 32/35) of 35 eyes. The retina did not reset in 3 eyes (8.6%, 3/35). logMAR BCVA of affected eye was 0.67±0.29 (finger counting-1.0). The difference of logMAR BCVA before and after operation was statistically significant ( t=5.133, P=0.036). In 35 eyes, visual acuity improved, stabilized and decreased in 19 (54.3%, 19/35), 13 (37.1%, 13/35) and 3 (8.6%, 3/35) eyes, respectively. Ten months after surgery, the silicone tape was exposed and infected 1 eye. After the silicone tape was removed, the infection subsided and the retina was in place. There were no intraocular hemorrhage, vitreoretinal impaction, endophthalmitis and other complications during and after operation. Conclusion:SB treatment of RRD with SRP can achieve good retinal reposition and improve visual acuity to some extent.