Electronic Health Record-Oriented Knowledge Graph System for Collaborative Clinical Decision Support Using Multicenter Fragmented Medical Data: Design and Application Study.

BACKGROUND: The medical knowledge graph provides explainable decision support, helping clinicians with prompt diagnosis and treatment suggestions. However, in real-world clinical practice, patients visit different hospitals seeking various medical services, resulting in fragmented patient data across hospitals. With data security issues, data fragmentation limits the application of knowledge graphs because single-hospital data cannot provide complete evidence for generating precise decision support and comprehensive explanations. It is important to study new methods for knowledge graph systems to integrate into multicenter, information-sensitive medical environments, using fragmented patient records for decision support while maintaining data privacy and security. OBJECTIVE: This study aims to propose an electronic health record (EHR)-oriented knowledge graph system for collaborative reasoning with multicenter fragmented patient medical data, all the while preserving data privacy. METHODS: The study introduced an EHR knowledge graph framework and a novel collaborative reasoning process for utilizing multicenter fragmented information. The system was deployed in each hospital and used a unified semantic structure and Observational Medical Outcomes Partnership (OMOP) vocabulary to standardize the local EHR data set. The system transforms local EHR data into semantic formats and performs semantic reasoning to generate intermediate reasoning findings. The generated intermediate findings used hypernym concepts to isolate original medical data. The intermediate findings and hash-encrypted patient identities were synchronized through a blockchain network. The multicenter intermediate findings were collaborated for final reasoning and clinical decision support without gathering original EHR data. RESULTS: The system underwent evaluation through an application study involving the utilization of multicenter fragmented EHR data to alert non-nephrology clinicians about overlooked patients with chronic kidney disease (CKD). The study covered 1185 patients in nonnephrology departments from 3 hospitals. The patients visited at least two of the hospitals. Of these, 124 patients were identified as meeting CKD diagnosis criteria through collaborative reasoning using multicenter EHR data, whereas the data from individual hospitals alone could not facilitate the identification of CKD in these patients. The assessment by clinicians indicated that 78/91 (86%) patients were CKD positive. CONCLUSIONS: The proposed system was able to effectively utilize multicenter fragmented EHR data for clinical application. The application study showed the clinical benefits of the system with prompt and comprehensive decision support.

Causal knowledge graph construction and evaluation for clinical decision support of diabetic nephropathy.

BACKGROUND: Many important clinical decisions require causal knowledge (CK) to take action. Although many causal knowledge bases for medicine have been constructed, a comprehensive evaluation based on real-world data and methods for handling potential knowledge noise are still lacking. OBJECTIVE: The objectives of our study are threefold: (1) propose a framework for the construction of a large-scale and high-quality causal knowledge graph (CKG); (2) design the methods for knowledge noise reduction to improve the quality of the CKG; (3) evaluate the knowledge completeness and accuracy of the CKG using real-world data. MATERIAL AND METHODS: We extracted causal triples from three knowledge sources (SemMedDB, UpToDate and Churchill's Pocketbook of Differential Diagnosis) based on rule methods and language models, performed ontological encoding, and then designed semantic modeling between electronic health record (EHR) data and the CKG to complete knowledge instantiation. We proposed two graph pruning strategies (co-occurrence ratio and causality ratio) to reduce the potential noise introduced by SemMedDB. Finally, the evaluation was carried out by taking the diagnostic decision support (DDS) of diabetic nephropathy (DN) as a real-world case. The data originated from a Chinese hospital EHR system from October 2010 to October 2020. The knowledge completeness and accuracy of the CKG were evaluated based on three state-of-the-art embedding methods (R-GCN, MHGRN and MedPath), the annotated clinical text and the expert review, respectively. RESULTS: This graph included 153,289 concepts and 1,719,968 causal triples. A total of 1427 inpatient data were used for evaluation. Better results were achieved by combining three knowledge sources than using only SemMedDB (three models: area under the receiver operating characteristic curve (AUC): p < 0.01, F1: p < 0.01), and the graph covered 93.9 % of the causal relations between diseases and diagnostic evidence recorded in clinical text. Causal relations played a vital role in all relations related to disease progression for DDS of DN (three models: AUC: p > 0.05, F1: p > 0.05), and after pruning, the knowledge accuracy of the CKG was significantly improved (three models: AUC: p < 0.01, F1: p < 0.01; expert review: average accuracy: + 5.5 %). CONCLUSIONS: The results demonstrated that our proposed CKG could completely and accurately capture the abstract CK under the concrete EHR data, and the pruning strategies could improve the knowledge accuracy of our CKG. The CKG has the potential to be applied to the DDS of diseases.

Risk Factors of Contrast Agent Intravasation With Sulfur Hexafluoride Microbubble During Hysterosalpingo-Contrast Sonography.

OBJECTIVES: To analyze the risk factors of sulfur hexafluoride microbubble contrast agent intravasation during hysterosalpingo-contrast sonography (HyCoSy), and to explore a simple prediction model by the obvious clinical history. METHODS: This was a retrospective study included 299 infertility women who had undergone HyCoSy examination from July 1, 2018 to June 31, 2019. The factors were recorded, including age, endometrial thickness, balloon length, infertility type, history of intrauterine surgery, history of pelvic surgery, and tubal patency. The method of multivariate logistic regression analysis was adopted to analyze the risk factors affecting the contrast agent intravasation, and the receiver operating characteristic curves were plotted to test their efficacy. RESULTS: Secondary infertility, a history of intrauterine surgery, thin endometrial thickness, and tubal obstruction were all risk factors of the occurrence of intravasation (P < .05). And the area under the receiver operating characteristic curves of the multifactor-combined prediction model of the intravasation was significantly larger than that of single-factor. CONCLUSIONS: Sonographers and gynecologists should be familiar with the risk factors of intravasation and select the appropriate timing of HyCoSy toward reducing the occurrence of intravasation and other complications after thoroughly explaining and communicating with the patients.

Contrast-enhanced ultrasound for the preoperative assessment of laryngeal carcinoma: a preliminary study.

BACKGROUND: Contrast-enhanced ultrasound (CEUS) is considered an attractive imaging technique to evaluate tumor microcirculation. However, the validity of CEUS for assessing laryngeal carcinoma is unclear. PURPOSE: To compare the performance of CEUS with conventional US and contrast-enhanced computed tomography (CECT) in the diagnosis and preoperative T-staging of laryngeal carcinoma. MATERIAL AND METHODS: Forty-one consecutive patients with laryngeal carcinoma underwent conventional high-frequency US, CEUS, and CECT before surgery. The CEUS characteristics of laryngeal carcinoma were recorded. The imaging findings of CEUS and conventional US were compared with CECT findings and the postoperative pathological examination. RESULTS: CEUS showed hyperenhancement in 38 cases and isoenhancement in three cases. Homogeneous distribution of contrast agent was found in 20 cases and heterogeneous distribution in 21 cases, of which 16 cases showed local perfusion defects. In the enhanced phase, rapid entry was observed in 37 cases, synchronous entry was observed in two cases, and slow entry was observed in two cases. Rapid exit was observed in 25 cases and slow exit was observed in 16 cases. The pretherapeutic T-staging accuracy was not significantly different between conventional US, CEUS, and CECT (P ≥ 0.500). A high sensitivity and specificity were achieved by CEUS in the evaluation of involvement of thyroid cartilage. CONCLUSION: Compared with conventional US and CECT, CEUS has a reliable initial T-staging accuracy and diagnostic properties for detecting laryngeal cartilage invasion.

Modular Dual-Tasked C-H Methylation via the Catellani Strategy.

We report a dual-tasked methylation that is based on cooperative palladium/norbornene catalysis. Readily available (hetero)aryl halides (39 iodides and 4 bromides) and inexpensive MeOTs or trimethylphosphate are utilized as the substrates and methylating reagent, respectively. Six types of "ipso" terminations can modularly couple with this "ortho" C-H methylation to constitute a versatile methylation toolbox for preparing diversified methylated arenes. This toolbox features inexpensive methyl sources, excellent functional-group tolerance, simple reaction procedures, and scalability. Importantly, it can be uneventfully extended to isotope-labeled methylation by switching to the corresponding reagents CD3OTs or 13CH3OTs. Moreover, this toolbox can be applied to late-stage modification of biorelevant substrates with complete stereoretention. We believe these salient and practical features of our dual-tasked methylation toolbox will be welcomed by academic and industrial researchers.

Analysis of the depression and anxiety status and related risk factors in patients with lumbar disc herniation.

OBJECTIVES: To evaluate the depression and anxiety status and related risk factors in patients with lumbar disc herniation, and help spine surgeons better identify those patients who need psychological care. METHODS: A cross-sectional study was performed on patients with lumbar disc herniation treated in our hospital between October 2015 and August 2018. Visual analog scale and Oswestry disability index were used to assess pain intensity and lumbar function, and Zung self-rating depression and anxiety scale were employed to evaluate the depression and anxiety status of the patients, and the demographic and clinical data including age, gender, marital status, occupation type, employment status, education level, surgery history, herniation type, disease duration, and insurance status were collected for analysis. RESULTS: In the current study, 165 patients were enrolled based on the inclusion and exclusion criteria. In multivariate logistic regression analysis, gender (p=0.03), pain intensity (p=0.01), self-rating anxiety scale (SAS) (p=0.00), and disease duration (p=0.001) were identified as independent risk factors for depression status, and pain intensity (p=0.02), disease duration (p=0.002) and SDS (Zung self-rating depression scale) (p=0.003) were independent risk factors for anxiety status in patients with lumbar disc herniation. There was a significant correlation between Zung self-rating depression and anxiety scale in patients with lumbar disc herniation (p<0.05). CONCLUSION: Psychological intervention is critical for patients with lumbar disc herniation, especially for those female patients with severe pain and longer disease duration.

POPCORN: A web service for individual PrognOsis prediction based on multi-center clinical data CollabORatioN without patient-level data sharing.

BACKGROUND AND OBJECTIVE: Clinical prognosis prediction plays an important role in clinical research and practice. The construction of prediction models based on electronic health record data has recently become a research focus. Due to the lack of external validation, prediction models based on single-center, hospital-specific datasets may not perform well with datasets from other medical institutions. Therefore, research investigating prognosis prediction model construction based on a collaborative analysis of multi-center electronic health record data could increase the number and coverage of patients used for model training, enrich patient prognostic features and ultimately improve the accuracy and generalization of prognosis prediction. MATERIALS AND METHODS: A web service for individual prognosis prediction based on multi-center clinical data collaboration without patient-level data sharing (POPCORN) was proposed. POPCORN focuses on solving key issues in multi-center collaborative research based on electronic health record systems; these issues include the standardization of clinical data expression, the preservation of patient privacy during model training and the effect of case mix variance on the prediction model construction and application. POPCORN is based on a multivariable meta-analysis and a Bayesian framework and can construct suitable prediction models for multiple clinical scenarios that can effectively adapt to complex clinical application environments. RESULTS: POPCORN was validated using a joint, multi-center collaborative research network between China and the United States with patients diagnosed with colorectal cancer. The performance of the models based on POPCORN was comparable to that of the standard prognosis prediction model; however, POPCORN did not expose raw patient data. The prediction models had similar AUC, but the BMA model had the lowest ECI across all prediction models, indicating that this model had better calibration performance than the other models, especially for patients in Chinese hospitals. CONCLUSIONS: The POPCORN system can build prediction models that perform well in complex clinical application scenarios and can provide effective decision support for individual patient prognostic predictions.

Comparison of two types of exercises in the treatment of lumbar spinal stenosis.

OBJECTIVE: To evaluate the efficacy of core stability exercise versus conventional exercise in the treatment of lumbar spinal stenosis. METHODS: Between January 2014 and May 2017, patients with lumbar spinal stenosis were recruited and divided into group of core stability exercise or conventional exercise randomly. All the patients were treated using middle frequency electrotherapy, in addition to that, the patients in group of core stability exercise were treated using core stability exercise. The patients in group of conventional exercise were treated using conventional exercise. The outcome was evaluated using Japanese Orthopedic Association (JOA) score, self-reported walking capacity and lumbar lordosis angle at baseline and after treatment. RESULTS: In the current study, sixty-two patients with lumbar spinal stenosis met the inclusion and exclusion criteria, in which 33 patients were included in group of core stability exercise and 29 in group of conventional exercise. After treatment, both Japanese Orthopedic Association scores (p<0.05) and self-reported walking capacity (p<0.05) increased significantly in each group when compared with baseline. The self-reported walking capacity and JOA scores in the group of core stability exercise were significantly higher than those in the conventional exercise group (p<0.05). However, both the intragroup and intergroup comparison of lumbar lordosis presented with no significance (p>0.05). CONCLUSION: Core stability exercise presents with better efficacy than conventional exercise in the treatment of lumbar spinal stenosis.

Degradation of the Separase-cleaved Rec8, a Meiotic Cohesin Subunit, by the N-end Rule Pathway.

The Ate1 arginyltransferase (R-transferase) is a component of the N-end rule pathway, which recognizes proteins containing N-terminal degradation signals called N-degrons, polyubiquitylates these proteins, and thereby causes their degradation by the proteasome. Ate1 arginylates N-terminal Asp, Glu, or (oxidized) Cys. The resulting N-terminal Arg is recognized by ubiquitin ligases of the N-end rule pathway. In the yeastSaccharomyces cerevisiae, the separase-mediated cleavage of the Scc1/Rad21/Mcd1 cohesin subunit generates a C-terminal fragment that bears N-terminal Arg and is destroyed by the N-end rule pathway without a requirement for arginylation. In contrast, the separase-mediated cleavage of Rec8, the mammalian meiotic cohesin subunit, yields a fragment bearing N-terminal Glu, a substrate of the Ate1 R-transferase. Here we constructed and used a germ cell-confinedAte1(-/-)mouse strain to analyze the separase-generated C-terminal fragment of Rec8. We show that this fragment is a short-lived N-end rule substrate, that its degradation requires N-terminal arginylation, and that maleAte1(-/-)mice are nearly infertile, due to massive apoptotic death ofAte1(-/-)spermatocytes during the metaphase of meiosis I. These effects ofAte1ablation are inferred to be caused, at least in part, by the failure to destroy the C-terminal fragment of Rec8 in the absence of N-terminal arginylation.

Effect of biodegradable polymer drug-eluting stents versus biocompatible polymer everolimus-eluting stents: a meta-analysis.

OBJECTIVE: Biocompatible polymer everolimus-eluting stents (EES) are associated with risk of stent thrombosis (ST); biodegradable polymer drug-eluting stents (BP-DES) were designed to reduce these risks. However, the long-term benefits are not completely clear. METHOD: We undertook a meta-analysis of randomized studies identified in systematic searches of MEDLINE, EMBASE, and the Cochrane Database. Primary outcome was the risk of ST. RESULTS: Twelve studies (11,692 patients) were included. Overall, compared with EES, BP-DES were associated with a broadly equivalent risk of definite and probable ST (OR, 0.91; 95% CI, 0.55 to 1.50; P = 0.71; I2 = 0.0%), early ST (OR, 2.25; 95% CI, 0.78 to 6.47; P = 0.13; I2 = 0.0%), late ST (OR, 3.57; 95% CI, 0.42 to 30.58; P = 0.25; I2 = 0.0%) and very late ST (OR, 0.50; 95% CI, 0.05 to 5.52; P = 0.57). Meanwhile, there was no significant difference in all-cause mortality (OR, 1.07; 95% CI, 0.86 to 1.32; P = 0.54; I2 = 0.0%), myocardial infarction (OR, 1.07; 95% CI, 0.88 to 1.30; P = 0.47; I2 = 0.0%), target vessel revascularization (OR, 1.02; 95% CI, 0.86 to 1.21; P = 0.80; I2 = 12.0%), and major adverse cardiac events (OR, 1.04; 95% CI, 0.93 to 1.16; P = 0.53; I2 = 0.0%). Furthermore, angiographic data showed that in-stent and in-segment late luminal loss were similar between the two groups. CONCLUSIONS: Compared with biocompatible polymer EES, biodegradable polymer stents appear to have equivalent clinical benefits.

Wnt/ß-Catenin Signaling Exacerbates Keloid Cell Proliferation by Regulating Telomerase.

OBJECTIVES: Our goal was to investigate the relationship between keloid and telomerase as well as clarifying the influence of Wnt/ß-catenin signaling on keloid cell proliferation. METHODS: Tissues from 18 keloid patients were collected for further study. Keloid progenitor cells (KPC) and skin progenitor cells (SKP) were both included in this study. Lenti-virus transfection was used to divide cells into different groups in which cells were treated with different substances: negative control (NC) group, wnt10a siRNA group, ß-catenin siRNA group and TERT siRNA group. KPC cells were injected into 20 male BALB/c nude mice in order to build tumor models. Several experiments including immunohistochemistry, western blot and RT-PCR were conducted in order to detect the corresponding protein expressions and relative mRNA levels. MTT assay and flow cytometry were also conducted for assessing cell proliferation and apoptosis status. RESULTS: ß-catenin and telomerase expression levels in keloid tissues were elevated compared to normal tissues (all P < 0.05). KPC cells in keloid exhibited more dynamic telomerase activity than SKP cells (P < 0.05). Luciferase activity assay confirmed that ß-catenin could directly interact with telomerase. After wnt10a/ß-catenin signaling pathway was inhibited, the proliferation of KPC cells was significantly suppressed and the apoptosis rate was remarkably increased (all P < 0.05). Results from tumor models also validated that wnt10a/ß-catenin signaling pathway influenced the activity and length of telomerase. CONCLUSIONS: Wnt/ß-catenin signaling pathway is able to exacerbate keloid cell proliferation and inhibit the apoptosis of keloid cells through its interaction with telomerase.

Scarless Epicanthoplasty and Concomitant Double Eyelidplasty in Chinese Eyelids.

BACKGROUND: The presence of mild or moderate medial epicanthus is typical in Asian patients. Numerous epicanthoplasty techniques have been described previously. However, these methods usually leave obvious scars in the medial canthal area. The aim of this report is to introduce a novel epicanthoplasty technique and a concomitant double eyelidplasty which avoid leaving scars in the medial canthal region. METHODS: From July 2013 to July 2015, 252 patients received epicanthoplasty and concomitant double eyelidplasty with this new technique. The medial epicanthus was corrected through the medial end of the eyelid crease incision. One hundred eighteen of these patients were followed up for 3-24 months (8 months in average). The preoperative and postoperative interepicanthal distances were measured at pre, 3 and 6 months post-operation. The aesthetic results were evaluated with patient visual analog scale (VAS) scores. RESULTS: The average intercanthal distance significantly decreased 3 months after the operation (32.7 ± 2.3 mm vs 36.5 ± 2.6 mm, p < 0.05, paired t test). Little retraction was noticed at 6 months after the operation (33.0 ± 2.4 vs 32.7 ± 2.3 mm, p < 0.05, paired t test). The mean patient VAS score associated with satisfaction of overall outcome was 4.2 at 6 months after operation (range 2.5-5.0). CONCLUSION: This new method provides an effective way to correct the medial epicanthus without leaving any scar in the medial canthal region. Patients with mild to moderate medial epicanthus are good candidates for this procedure. LEVEL OF EVIDENCE V: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

[Rat prostate glandular epithelial cells cultured in vitro and their barrier function].

OBJECTIVE: To culture rat prostate glandular epithelial cells and study their barrier functions in vitro. METHODS: Rat prostate glandular epithelial cells were cultured in vitro. The expression of the tight junction protein claudin-1 was determined by immunohistochemistry, the structure and composition of the epithelial cells observed under the inverted microscope and transmission electron microscope. The transepithelial electrical resistances (TEERs) were monitored with the Millicell system. The permeability of the prostate glandular epithelial cells was assessed by the phenol red leakage test. RESULTS: Compact monolayer cell structures were formed in the prostate glandular epithelial cells cultured in vitro. Immunohistochemistry showed the expression of the tight junction protein claudin-1 and transmission electron microscopy confirmed the formation of tight junctions between the adjacent glandular epithelial cells. The TEERs in the cultured prostate glandular epithelial cells reached the peak of about (201.3 ± 3.5) Ω/cm2 on the 8th day. The phenol red leakage test manifested a decreased permeability of the cell layers with the increase of TEERs. CONCLUSION: The structure and function of rat prostate glandular epithelial cells are similar to those of brain capillary endothelial cells, retinal capillary endothelial cells, and intestinal epithelial cells. In vitro cultured prostate glandular epithelial cells have the barrier function and can be used as a model for the study of blood prostate barrier in vitro.

Meta-analysis of contrast-enhanced ultrasound for the differentiation of benign and malignant breast lesions.

BACKGROUND: Contrast-enhanced ultrasound (CEUS) is a non-invasive method for the assessment of breast lesions. The accuracy of CEUS in diagnosing of breast cancer has never been systematically assessed. PURPOSE: To determine the overall performance of CEUS in the differentiation of benign and malignant breast lesions using meta-analysis. MATERIAL AND METHODS: PubMed, Embase, Cochrane Library, and article references published before October 2012 were searched. Published studies that used histopathologic results as golden reference to assess the diagnostic performance of CEUS in patients suspected of having breast cancer and the data necessary to calculate the diagnostic results were included. The qualities of eligible studies for final meta-analysis were assessed by using the quality assessment of diagnostic studies (QUADAS) instrument. Sensitivity, specificity, summary receiver-operating characteristic (sROC) curves, and area under the curve were calculated to examine the diagnostic performance of CEUS. RESULTS: Of 16 eligible studies, 957 breast lesions were included in the original meta-analysis, among which heterogeneity arising from factors other than threshold effect was explored. Meta-regression analysis confirmed the contrast agent was the most significant factor cause of heterogeneity (P = 0.0012, relative diagnostic odds ratio [DOR] = 7.06). The use of perfluoro containing microbubbles (Sonovue or Optison) significantly increased the diagnostic precision compared with Levovist. The pooled weighted estimates of sensitivity and specificity for CEUS in the diagnosis of breast lesions were 0.86 (95% confidence interval [CI], 0.83, 0.89) and 0.79 (95% CI, 0.75, 0.83), respectively. CONCLUSION: CEUS has good sensitivity and specificity in the characterization of breast lesions and can potentially help to select suspicious breast mass for surgery.

Association of ADAM33 gene polymorphisms with keloid scars in a northeastern Chinese population.

OBJECTIVE: To study the association between ADAM33 and keloid scars in the northeastern Chinese population. METHODS: A total of 283 keloid scar patients and a control group of 290 healthy volunteers were recruited for this study. Six polymorphic loci (V4, T+1, T2, T1, S2 and Q-1 ) of ADAM33 were selected for genotyping. Genotypes were determined by using the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method. RESULTS: We observed the frequency of the rs612709 A allele exhibited a significantly decreased frequency in cases than in controls(22 vs.39.6%, P<0.0001) We also found that the frequencies of H2 (GGAAGA) haplotypes was significantly higher in the case group than in the control group (P= 0.041). In contrast, the haplotype H8 (GGGAGG) was more common in the control group than in the case group (P=0.022). CONCLUSIONS: Our data suggest that the ADAM33 polymorphisms may be associated with keloid scars in the northeastern Chinese population.

Sonography for diagnosis of benign and malignant tumors of the nose and paranasal sinuses.

OBJECTIVES: The purpose of this study was to demonstrate the reliability of sonography for diagnosis of nose and paranasal sinus tumors. METHODS: Ninety-six consecutive patients with tumors underwent sonography and computed tomography (CT) before surgical treatment. Tumor detectability and imaging findings were evaluated independently and then compared with pathologic findings. RESULTS: Of 96 tumors, 75 were detected by sonography, for a detectability rate of 78.1%; 93 tumors were detected by CT, for a detectability rate of 96.9%. By comparison, sonography showed a trend toward higher detectability of nasal vestibular tumors than CT (87.5% for sonography versus 50.0% for CT) and small lumps on the wing of the nose (78.8% for sonography versus 33.3% for CT). Among the sonographic features, boundary, shape, internal echo, calcification, bone invasion, vascular pattern, and cervical lymph node metastasis all had significantly positive correlations with malignancy (P < .05), but size did not (P = .324). In addition, the vascular resistive index for malignant tumors was significantly higher (mean ± SD, 0.66 ± 0.20) than the index for benign lesions (0.24 ± 0.30; P < .001). Moreover, the detection rate for grade 1-3 (small-large) blood flow in benign lesions was only 43.8%, whereas the rate for malignant tumors was 97.7% (P < .001). CONCLUSIONS: The vascular pattern may be a promising predictive indicator for distinguishing benign and malignant tumors of the nose and paranasal sinuses. Consequently, sonography has high value for diagnosis of benign and malignant tumors of the nose and paranasal sinuses, especially for nasal vestibular tumors and small lumps on the wing of the nose.

Diagnostic accuracy of sonography versus magnetic resonance imaging for primary nasopharyngeal carcinoma.

OBJECTIVES: The purpose of this study was to prospectively assess the accuracy of sonography versus magnetic resonance imaging (MRI) for a diagnosis of primary nasopharyngeal carcinoma. METHODS: A total of 150 patients suspected of having nasopharyngeal carcinoma underwent sonography and MRI. A diagnosis was obtained from an endoscopic biopsy that was collected from the suspected tumor or a normal nasopharynx. The diagnostic performance of sonography and MRI for nasopharyngeal carcinoma was evaluated by receiver operating characteristic curve analysis. The sensitivity and specificity of the two imaging methods were compared by the McNemar test. RESULTS: Nasopharyngeal carcinoma was present in 71 of 150 patients (47.3%) and absent in 79 (52.7%). The sensitivity, specificity, and accuracy of sonography versus MRI for these cases were 90.1%, 84.8%, and 87.3% for sonography and 97.2%, 89.9%, and 93.3% for MRI, respectively. Both sonography and MRI had good diagnostic performance for nasopharyngeal carcinoma, with area under the curve values of 0.958 and 0.987, respectively. There was no significant difference in the rate of tumor detectability between sonography and MRI (P = .12), and the specificities of sonography and MRI were similar (P = .22). CONCLUSIONS: Both sonography and MRI are useful tools for clinical screening of nasopharyngeal carcinoma. However, sonography is less expensive and easier to perform. The results of this study also suggest that nasopharyngeal sonography could be used for the initial investigation of primary cancer in patients suspected of having nasopharyngeal carcinoma.

A Personal Healthcare Knowledge Graph Framework for Diagnosis of Pelvic Masses Diseases.

A Personal Health Knowledge Graph (PHKG) facilitates the efficient integration of potential diagnostic clues from patients' electronic health records with medical knowledge, establishing diagnostic reasoning paths and ensuring accurate, individually interpretable results in the diagnosis of pelvic masses.

A few-shot disease diagnosis decision making model based on meta-learning for general practice.

BACKGROUND: Diagnostic errors have become the biggest threat to the safety of patients in primary health care. General practitioners, as the "gatekeepers" of primary health care, have a responsibility to accurately diagnose patients. However, many general practitioners have insufficient knowledge and clinical experience in some diseases. Clinical decision making tools need to be developed to effectively improve the diagnostic process in primary health care. The long-tailed class distributions of medical datasets are challenging for many popular decision making models based on deep learning, which have difficulty predicting few-shot diseases. Meta-learning is a new strategy for solving few-shot problems. METHODS AND MATERIALS: In this study, a few-shot disease diagnosis decision making model based on a model-agnostic meta-learning algorithm (FSDD-MAML) is proposed. The MAML algorithm is applied in a knowledge graph-based disease diagnosis model to find the optimal model parameters. Moreover, FSDD-MAML can learn learning rates for all modules of the knowledge graph-based disease diagnosis model. For n-way, k-shot learning tasks, the inner loop of FSDD-MAML performs multiple gradient update steps to learn internal features in disease classification tasks using n×k examples, and the outer loop of FSDD-MAML optimizes the meta-objective to find the associated optimal parameters and learning rates. FSDD-MAML is compared with the original knowledge graph-based disease diagnosis model and other meta-learning algorithms based on an abdominal disease dataset. RESULT: Meta-learning algorithms can greatly improve the performance of models in top-1 evaluation compared with top-3, top-5, and top-10 evaluations. The proposed decision making model FSDD-MAML outperforms all the other models, with a precision@1 of 90.02 %. We achieve state-of-the-art performance in the diagnosis of all diseases, and the prediction performance for few-shot diseases is greatly improved. For the two groups with the fewest examples of diseases, FSDD-MAML achieves relative increases in precision@1 of 29.13 % and 21.63 % compared with the original knowledge graph-based disease diagnosis model. In addition, we analyze the reasoning process of several few-shot disease predictions and provide an explanation for the results. CONCLUSION: The decision making model based on meta-learning proposed in this paper can support the rapid diagnosis of diseases in general practice and is especially capable of helping general practitioners diagnose few-shot diseases. This study is of profound significance for the exploration and application of meta-learning to few-shot disease assessment in general practice.