Recently, deep learning has achieved impressive results in medical image tasks. However, this method usually requires large-scale annotated data, and medical images are expensive to annotate, so it is a challenge to learn efficiently from the limited annotated data. Currently, the two commonly used methods are transfer learning and self-supervised learning. However, these two methods have been little studied in multimodal medical images, so this study proposes a contrastive learning method for multimodal medical images. The method takes images of different modalities of the same patient as positive samples, which effectively increases the number of positive samples in the training process and helps the model to fully learn the similarities and differences of lesions on images of different modalities, thus improving the model's understanding of medical images and diagnostic accuracy. The commonly used data augmentation methods are not suitable for multimodal images, so this paper proposes a domain adaptive denormalization method to transform the source domain images with the help of statistical information of the target domain. In this study, the method is validated with two different multimodal medical image classification tasks: in the microvascular infiltration recognition task, the method achieves an accuracy of (74.79 ± 0.74)% and an F1 score of (78.37 ± 1.94)%, which are improved as compared with other conventional learning methods; for the brain tumor pathology grading task, the method also achieves significant improvements. The results show that the method achieves good results on multimodal medical images and can provide a reference solution for pre-training multimodal medical images.
Humans ; Algorithms ; Brain/diagnostic imaging* ; Brain Neoplasms/diagnostic imaging* ; Recognition, Psychology

Hepatocellular carcinoma (HCC) is the most common liver malignancy, where HCC segmentation and prediction of the degree of pathological differentiation are two important tasks in surgical treatment and prognosis evaluation. Existing methods usually solve these two problems independently without considering the correlation of the two tasks. In this paper, we propose a multi-task learning model that aims to accomplish the segmentation task and classification task simultaneously. The model consists of a segmentation subnet and a classification subnet. A multi-scale feature fusion method is proposed in the classification subnet to improve the classification accuracy, and a boundary-aware attention is designed in the segmentation subnet to solve the problem of tumor over-segmentation. A dynamic weighted average multi-task loss is used to make the model achieve optimal performance in both tasks simultaneously. The experimental results of this method on 295 HCC patients are superior to other multi-task learning methods, with a Dice similarity coefficient (Dice) of (83.9 ± 0.88)% on the segmentation task, while the average recall is (86.08 ± 0.83)% and an F1 score is (80.05 ± 1.7)% on the classification task. The results show that the multi-task learning method proposed in this paper can perform the classification task and segmentation task well at the same time, which can provide theoretical reference for clinical diagnosis and treatment of HCC patients.
Humans ; Carcinoma, Hepatocellular ; Liver Neoplasms ; Learning

Objectives:This study aimed to assess the efficacy and safety of bendamustine in combination with rituximab (BR regimen) for the treatment of newly diagnosed indolent B-cell non-Hodgkin's lymphoma (B-iNHL) and elderly mantle cell lymphoma (eMCL) .Methods:From December 1, 2020 to September 10, 2022, a multi-center prospective study was conducted across ten Grade A tertiary hospitals in Shandong Province, China. The BR regimen was administered to evaluate its efficacy and safety in newly diagnosed B-iNHL and eMCL patients, and all completed at least four cycles of induction therapy.Results:The 72 enrolled patients with B-iNHL or MCL were aged 24-74 years, with a median age of 55 years. Eastern Cooperative Oncology Group (ECOG) performance status scores of 0-1 were observed in 76.4% of patients, while 23.6% had scores of 2. Disease distribution included follicular lymphoma (FL) (51.4% ), marginal zone lymphoma (MZL) (33.3% ), eMCL (11.1% ), and the unknown subtype (4.2% ). According to the Ann Arbor staging system, 16.7% and 65.3% of patients were diagnosed with stage Ⅲ and stage Ⅳ lymphomas, respectively. Following four cycles of BR induction therapy, the overall response rate was 98.6%, with a complete response (CR) rate of 83.3% and a partial response (PR) rate of 15.3%. Only one eMCL patient experienced disease progression during treatment, and only one FL patient experienced a relapse. Even when evaluated using CT alone, the CR rate was 63.9%, considering the differences between PET/CT and CT assessments. The median follow-up duration was 11 months (range: 4-22), with a PFS rate of 96.8% and an OS rate of 100.0%. The main hematologic adverse reactions included grade 3-4 leukopenia (27.8%, with febrile neutropenia observed in 8.3% of patients), grade 3-4 lymphopenia (23.6% ), grade 3-4 anemia (5.6% ), and grade 3-4 thrombocytopenia (4.2% ). The main non-hematologic adverse reactions such as fatigue, nausea/vomiting, rash, and infections occurred in less than 20.0% of patients.Conclusion:Within the scope of this clinical trial conducted in China, the BR regimen demonstrated efficacy and safety in treating newly diagnosed B-iNHL and eMCL patients.

Objective:To evaluate the clinical efficacy and safety of fecal microbiota transplantation(FMT)for the treatment of chronic functional constipation in the elderly.Methods:A total of 33 elderly patients with chronic functional constipation were included and given three sessions of FMT.Changes in fecal characteristics, constipation, mood and quality of life in these patients were evaluated using the Bristol stool form scale(BSFS), the constipation assessment scale(CAS), patient assessment of constipation symptoms(PAC-SYM), the Zung self-rating anxiety scale(SAS), the Zung self-rating depression scale(SDS), and the patient assessment of constipation quality of life(PAC-QOL)before and 12 weeks after treatment.The clinical efficacy was based on comparison between pre-and post-treatment results for each patient.Results:Clear improvement was achieved in 33 patients 12 weeks after treatment, compared with before transplantation.Post-treatment scores of the constipation assessment scale and symptom self-assessment questionnaire for patients with constipation were(8.9±1.2)scores and(26.5±2.4)scores, respectively, significantly lower than pre-transplantation scores of(12.2±1.1)scores and(32.4±2.4)scores( t=15.034, 13.904, both P<0.001). Similarly, post-treatment scores were also lower than pre-transplantation levels for the self-rating anxiety scale[(50.4±8.4)scores vs.(57.5±9.0)scores, t=10.333, P<0.001], the self-rating depression scale[(50.6±8.4)% vs.(55.0±10.5)%, t=5.301, P<0.001], and self-assessment questionnaire for quality of life[(88.2±7.3)scores vs.(103.7±7.3)scores, t=23.300, P<0.001]. Conclusions:FMT can improve fecal characteristics and constipation symptoms, relieve anxiety and depression, improve the quality of life, and provide a new option for the treatment for elderly patients with chronic functional constipation.

According to the pathological characteristics of symptomatic chronic thoracic and lumbar osteoporotic vertebral fracture (SCOVF), the different clinical treatment methods are selected, including vertebral augmentation, anterior-posterior fixation and fusion, posterior decompression fixation and fusion, and posterior correction osteotomy. However, there is still a lack of a unified understanding on how to choose appropriate treatment method for SCOVF. In order to reflect the new treatment concept and the evidence-based medicine progress of SCOVF in a timely manner and standardize its treatment, the clinical guideline for surgical treatment of SCOVF is formulated in compliance with the principle of scientificity, practicability and advancement and based on the level of evidence-based medicine.

Objective To investigate the clinical effects of levosimmentan combined with freeze-dried recombinant human brain natriuretic peptide in elderly patients with heart failure. Methods Eighty patients with heart failure diagnosed in our hospital from January 2015 to January 2018 were selected as subjects. According to the random number table prepared in Excel 2007,40 patients in the experimental group and the control group were given. For basic treatments such as diuresis,the experimental group was treated with levosimtan combined with lyophilized recombinant human brain natriuretic peptide,and the control group was treated with levosimtan and dobutamine. The left ventricle of the two groups before and after treatment was compared. End-diastolic period (LVEDD),left ventricular ejection fraction(LVEF),cardiac index(CI),stroke volume(SV),serum cystatin C (Cys-C),amino terminal brain natriuretic peptide precursor(NT-proBNP),mean arterial pressure(MAP), pulmonary capillary pressure(PCWP). Results Before treatment,the differences of LVEDD,LVEF,CI and SV levels between the experimental group and the control group were not statistically significant(P > 0.05). After treatment,the LVEDD of the experimental group was significantly lower than that of the control group(P < 0.05).The LVEF,CI and SV values in the experimental group were significantly higher than those in the control group (P < 0.05). There was no significant difference in serum Cys-C and NT-proBNP levels between the experimental group and the control group before treatment(P > 0.05). After treatment,the serum levels of Cys-C and NT-proB-NP in the experimental group were significantly lower than those in the control group(P < 0.05). Before treatment,the difference of MAP and PCWP levels between the experimental group and the control group was not statistically significant(P > 0.05). After treatment,the MAP and PCWP levels in the experimental group were significantly lower than those in the control group(P < 0.05). The incidence of adverse reactions in the experimental group was 10.00％ and the difference between the control group and the control group was not statistically significant(P > 0.05). Conclusions Levosimmentan combined with freeze-dried recombinant human brain natriuretic peptide in patients with senile heart failure can significantly improve cardiac function,reduce Cys-C,NT-proBNP levels,and improve hemodynamic parameters.

Objective To evaluate the effect of online teaching platform in clinical neurology prac-tice. Methods The participants were medical undergraduate students in the process of clinical practice in Department of Neurology of XiJing hospital, who were divided into the traditional practice teaching group (n=77) and online platform auxiliary practice teaching group (n=90), respectively carry-ing out the corre-sponding rotation practice teaching . Unified objective structured clinical examination was given to two groups of students to implement departmental rotation examination , and investigation was made among teachers and students on the evaluation of the effect of practice. SPSS 19.0 was used to do t test to the two sets of corresponding evaluation data. Results Total score of departmental rotation examination in online platform auxiliary practice teaching group (69.33 ±2.74) was significantly higher than traditional teaching group (67.23±2.50) (P=0.000). Scores of basic theories, skill and operation, clinical case analysis in online platform auxiliary practice teaching group were significantly higher than traditional teaching group (P<0.05). However, medical record writing scores in online platform auxiliary practice teaching group (7.39± 1.09) were significantly lower than the traditional teaching group (8.03±1.03) (P=0.000). Teachers' eval-uation of practice effect was obviously higher in online platform auxiliary practice teaching group (12.33± 0.52) than that in traditional teaching group (10.67 ±1.03) (P=0.005). Students' evaluation of interest in learning, content arrangement, teachers' sense of responsibility, evaluation of teaching ability, practice effect were obviously higher than that of traditional teaching group (P<0.05). Conclusion This study preliminary has achieved good practical effect by using online auxiliary teaching platform, and got high evaluation by both teachers and students. Inter-active communication and sharing of resources is easy in the online plat-form of practice teaching . On-line platform will help students develop with clinical neurology diagnosis thinking.

Objective To evaluate effect of intensive insulin treatment(IIT)on healthcare-associated infection (HAI)rate in patients with acute stroke and stress hyperglycemia.Methods Databases,including PubMed,Em-base,Cochrane Library,WanFang,and China National Knowledge Infrastructure(CNKI)Data,were electronically searched,relevant journals and references of the included literatures were also searched manually,literatures were selected according to the uniform inclusion and exclusion criteria,incidence of HAI and mean blood glucose in patients who received IIT for acute stroke were assessed systematically.Results A total of 13 randomized controlled trials (RCT)involving 1 032 patients were included in this systematic review.Meta-analysis results showed that 10 studies involving 832 patients were finally enrolled for comparing HAI rate, HAI rates in IIT group and conventional insulin treatment group were 28.3% and 56.1 %,respectively(Z =4.50 ),difference between two groups was statistically significant (RR=0.53 [95 %CI :0.40 to 0.70],P <0.001 );A total of 328 patients in 5 studies were finally included in the comparison of blood glucose,difference in mean blood glucose between two groups was statistically significant(MD =-2.52 [95% CI :-4.30 to -0.74],P =0.006).Funnel plot of HAI rate revealed that there was publication bias.Conclusion IIT is used for the regulation of stress hyperglycemia in acute stroke,it can reduce the incidence of HAI and blood glucose in patients.

Objective To establish HPLC method for determination of (R,S)-epigoitrin in Banlangen granules and discuss the content limitation. Methods The samples were separated on SHIMADZU VP-ODS (150 mm×4. 6 mm, 5 μm) and the mobile phase consisted of methol-0. 02% phosphoric acid solution (7:93) at a flow rate of 1. 0 mL. min-1 . The detection wavelength was set at 245 nm and the injection volume (in automatic sampler) was 10 μL. The content limitation was assessed according to the transfer rate and statistical data of the results. Results (R,S)-epigoitrin showed a good linear relationship at concentration of 0. 058 7 – 150. 349 5 μg·mL-1(r = 0. 999 7, n = 7). The average recovery rate was 98. 72% , 98. 40% and 98. 60% , respectively; RSD was 1. 84% , 0. 50% and 1. 82% , respectively. The content limitation of (R,S)-epigoitrin was unreasonable according to the transfer rate and the statistical data of the results. Conclusion The method is easy and simple to operate, accurate and stable in results, and highly specific, thus it is applicable for the quality control of Banlangen granules. The content limitation should be determined on the basis of pharmacokinetic and pharmacodynamic data.

Objective To establish the quality standard of common selfheal fruit-spike granules. Mehtods Rosmarinicacid in the granules were identified by TLC. The extractives were tested by the method for determination of water-soluble extractives. The contents of rosmarinicacid were determined by HPLC. Results TLC had obvious characteristics with the spots clear and well-separated. The extraction limit was no less than 10.0%. Rosmarinicacid showed a good linear relationship at 0.062 5~2.008 0 mg/ml (r=0.999 8,n=6) with average recovery of 99.50% (RSD=1.21%,n=9). The content limitation of rosmarinicacid was discussed as 10.0 mg per pack according to the experiment data.Conclusion The method is easy-operated and accurate which has a good specificity for the quality control of common selfheal fruit-spike granules.