Timely and precise identification of pulmonary nodules via CT imaging plays a pivotal role in effective lung cancer screening and diagnosis. However, traditional nodule detection largely relies on radiologists’ visual inspection, which is not only labor‐intensive but also subjective. This paper introduces a novel two‐stage three‐dimensional convolutional neural network for automated pulmonary nodule detection. In the first stage, an adaptive multiscale dual attention (AMDA) mechanism is introduced to enhance feature representation across varying spatial scales. In the second stage, a Kolmogorov–Arnold layer (KAL) is integrated to strengthen nonlinear feature modeling and improve discrimination between true nodules and hard negative samples. Based on extensive experiments conducted on the LUNA16 dataset, the results indicate that the proposed framework achieves competitive performance across a wide range of false‐positive rates. Furthermore, cross‐dataset evaluation on the LIDC‐IDRI dataset validates the generalization capability of the proposed method. Our approach not only advances automated nodule detection but also provides valuable support for radiologists in clinical settings.

A 0.52% nonlinearity,$44.9zF/\sqrt {Hz}$input capacitance noise capacitance-to-voltage (C/V) converter for MEMS accelerometer is proposed in this study. The multi-step successive approximation (SA) output mechanism is proposed to replace the conventional single-step conversion. Meanwhile, combining with the modulation-demodulation (MD) technique in the traditional continuous-time (CT) C/V converter, a much lower nonlinearity and noise CT C/V converter without increasing power consumption is achieved in this work. In addition, a short time reset signal between modulation and demodulation is proposed to reset the integration capacitors in each conversion cycle, which lets the C/V converter achieve an approximate twice dynamic range enhancement (DRE). The proposed C/V converter is implemented in a standard 180 nm CMOS process. It incorporates an on-chip capacitor signal generator (CSG), whose capacitance is tuned from −80 fF to 80 fF using clock signals at frequencies of 10Hz, 1 kHz, 2 kHz, 3 kHz, 4 kHz, and 5 kHz respectively, where measurement results demonstrate that the converter has correct output voltage across this frequency range, with a conversion error of less than$350~ppm/\sqrt {Hz}$. The proposed C/V converter is also tested together with an integrated a MEMS accelerometer element. It achieves a maximum nonlinearity as low as 0.52% within ±10 g acceleration and obtains equivalent input noise power spectral density (PSD) as low as$44.9~zF/\sqrt {Hz}$when the frequency is beyond 50 Hz.

In this paper, the acoustic signal generated by the damaged water supply pipeline is deeply analyzed by Ansys simulation software. By constructing a three-dimensional model of water supply pipeline, the acoustic signals generated in the pipeline under different damage forms are simulated, and effective acoustic characteristic parameters are successfully extracted. In the research process, firstly, the three-dimensional model of water supply pipeline is established, and many common damage forms are simulated. Then, using the acoustic analysis module of Ansys software, the acoustic signal generated at the damaged part is simulated, and the simulation results are analyzed in detail by signal processing technology. By comparing the characteristic parameters of acoustic signals under different damage forms, the differences and laws between them are found, which provides a new theoretical basis for the damage detection of water supply pipelines.

Conventional knowledge graphs (KGs) are composed solely of entities, attributes, and relationships, which poses challenges for enhancing multimodal knowledge representation and reasoning. To address the issue, this article proposes a multimodal deep learning-based approach to build a multimodal knowledge base (MMKB) for better multimodal feature (MMF) utilization. First, we construct a multimodal computation sequence (MCS) model for structured multimodal data storage. Then, we propose multimodal node, relationship, and dictionary models to enhance multimodal knowledge representation. Various feature extractors are used to extract MMFs from text, audio, image, and video data. Finally, we leverage generative adversarial networks (GANs) to facilitate MMF representation and update the MMKB dynamically. We examine the performance of the proposed method by using three multimodal datasets. BOW-, LBP-, Volume-, and VGGish-based feature extractors outperform the other methods by reducing at least 1.13%, 22.14%, 39.87, and 5.65% of the time cost, respectively. The average time costs of creating multimodal indexes improve by approximately 55.07% and 68.60% exact matching rates compared with the baseline method, respectively. The deep learning-based autoencoder method reduces the search time cost by 98.90% after using the trained model, outperforming the state-of-the-art methods. In terms of multimodal data representation, the GAN-CNN models achieve an average correct rate of 82.70%. Our open-source work highlights the importance of flexible MMF utilization in multimodal KGs, leading to more powerful and diverse applications that can leverage different types of data.

Diagnostic error refers to a missed, delayed, or wrong diagnosis, which seriously affects diagnostic reliability and safety. Identifying diagnostic errors is crucial for diagnostic error research. As an important method of identifying diagnostic errors, electronic triggers (e-triggers) have several limitations, including the performance improvement bottleneck, the inability to predict diagnostic errors, and the narrow trigger range caused by inadequate data utilization. This article proposes a novel approach for identifying diagnostic errors in electronic medical records (EMRs) based on machine learning (ML) techniques. First, we design four stages for our approach, i.e., EMR utilization, ML, evaluation, and prediction. By learning the implicit identification pattern of diagnostic errors from historical EMRs, the proposed approach can find potential diagnostic error cases and predict the probability of upcoming diagnostic errors. Second, in the evaluation stage, we introduce various metrics to evaluate its performance accurately and effectively. Metrics for measuring the abilities to identify diagnostic errors and resist false negatives are introduced. Third, the study utilizes real-world EMRs of leukemia patients to illustrate and verify the proposed approach. The experimental results indicate that our approach accurately identifies most diagnostic error cases and efficiently reduces the number of false negatives, which will help protect diagnostic reliability and safety in the clinic.

As acquired seismic data is usually incomplete and noisy, simultaneous reconstruction and denoising is an extremely important step for the accurate interpretation of seismic data and subsequent processing. We propose a hybrid low-rank and sparsity constraint method with Hankel structure preservation to improve the performance of simultaneous reconstruction and denoising. The proposed method combines the advantages of high efficiency pertaining to sparsity-promoting transforms and the strong data adaptability of rank reduction methods. Meanwhile, a structure-preserving matrix is constructed to preserve the predefined Hankel structure of the twofold Hankel matrix to further improve the accuracy and efficiency of simultaneous reconstruction and denoising. Moreover, weighted nuclear norm minimization (WNNM) is introduced to adaptively assign weights to different singular values. Experimental results in both synthetic and field seismic data compared with other state-of-the-art methods demonstrate the superior performance of the proposed method.

Analyzing massive VAERS reports without medical context may lead to incorrect conclusions about vaccine adverse events (VAE). Facilitating VAE detection promotes continual safety improvement for new vaccines. This study proposes a multi-label classification method with various term-and topic-based label selection strategies to improve the accuracy and efficiency of VAE detection. Topic modeling methods are first used to generate rule-based label dependencies from Medical Dictionary for Regulatory Activities terms in VAE reports with two hyper-parameters. Multiple label selection strategies, namely one-vs-rest (OvsR), problem transformation (PT), algorithm adaption (AA), and deep learning (DL) methods, are used in multi-label classification to examine the model performance, respectively. Experimental results indicated that the topic-based PT methods improve the accuracy by up to 33.69% using a COVID-19 VAE reporting data set, which improves the robustness and interpretability of our models. In addition, the topic-based OvsR methods achieve an optimal accuracy of up to 98.88%. The accuracy of the AA methods with topic-based labels increased by up to 87.36%. By contrast, the state-of-art LSTM- and BERT-based DL methods have relatively poor performance with accuracy rates of 71.89% and 64.63%, respectively. Our findings reveal that the proposed method effectively improves the model accuracy and strengthens VAE interpretability by using different label selection strategies and domain knowledge in multi-label classification for VAE detection.

Identifying the model of a cell-phone with which an audio recording is made serves an important forensic purpose. In this paper, we propose a channel attention mechanism based on subband awareness to focus on the most relevant parts of the frequency bands to produce more efficient feature representations for cell-phone recognition task. A multi-stream network is introduced to fully exploit difference among frequency bands that provide critical clues to identify the fingerprints from the built-in microphones, and thus is able to recognize cell-phones from different manufacturers and even different models from the same manufacturer. The effectiveness of the proposed design is validated by a collection of speaker-independent audio recordings from 20 models of cell-phones made by 5 major manufacturers. In particular, the fusion of data augmentation and attention strategy greatly increases the robustness of the scheme when additive noise is present in the recorded audios. Finally, the salient regions infer the critical subbands for the recognition of different cell-phone models.

The International Classification of Diseases (ICD) not only serves as the bedrock for health statistics but also provides a holistic overview of every health aspect of life. This study aims to facilitate the computer-assisted coding of the 11th revision of the ICD (ICD-11) by leveraging the data structures of ICD-11 and semantics in WordNet. First, a computer-assisted coding framework using WordNet and ICD-11 application programming interface (API) is proposed. Secondly, a network based on entity relations in ICD-11 and synonym sets in WordNet, called CodeNet, is developed. Thirdly, an algorithm for generating ICD-11 code candidates from CodeNet with two tuning parameters on the basis of the input of disease-related text is illustrated. Finally, the discharge summaries in the Medical Information Mart for Intensive Care III database and textual information from ICD-11 entities are used to evaluate the proposed method. Experimental results indicate that the proposed coding method achieves a precision of 84% and a recall of 89% relative to a precision of 65% and a recall of 81% achieved with the existing ICD-11 API. The proposed method also outperforms other methods in the literature by reducing a failure rate of up to 8% in ICD-11 coding. The proposed thresholds of similarity and percentage can be applied to tuning the performance of our method to meet different coding needs. In sum, improving the new structures of ICD-11 with the use of semantics in WordNet can help develop more reliable computer-aided coding systems for ICD-11 coders.

Chronic diseases are associated with high morbidity and mortality, and they cannot be quickly determined and identified during their early stages. Therefore, early identification and warning of chronic diseases are important. This study proposed an early warning model (EWM) based on electronic health records (EHRs). The model included comprehensive methods that identify whether the patient is expected to suffer from chronic disease and provide early warning based on the undiagnosed narrative text of EHRs. A professional medical terminology library called Systematized Nomenclature of Medicine-Clinical Terms (SNOMED CT) was utilized to improve the accuracy of early warning. The model utilized semantic relationships in the SNOMED CT to expand the effects of related terms of renal cancer. We utilized 1,300 medical records and more than 1,400 progress notes in the EHRs in our experiments. The proposed EWM of renal cancer achieved a precision of 90%, recall of 91%, and F-measure of 91%.

Current access control mechanisms of the hospital information system can hardly identify the real access intention of system users. A relaxed access control increases the risk of compromise of patient privacy. To reduce unnecessary access of patient information by hospital staff, this paper proposes a knowledge-constrained role-based access control (KC-RBAC) model in which a variety of medical domain knowledge is considered in access control. Based on the proposed Purpose Tree and knowledge-involved algorithms, the model can dynamically define the boundary of access to the patient information according to the context, which helps to protect patient privacy by controlling access. Compared with the role-based access control model, KC-RBAC can effectively protect patient information according to the results of the experiments.

In recent years, many online health communities (OHCs) are established to provide the patients with the services of disease prevention and self-management. Patients in those online health communities discuss their health conditions and share their experiences with other patients using narrative texts in the posts. Those posts contain a vast amount of patients' information, including drugs, symptoms, conditions, etc. They are really valuable for clinical research. However, it is hard for information systems to automatically extract the clinical knowledge from those posts to provide knowledge-based services to online patients. This paper investigates the characteristics of those post contents and accordingly proposes an encoding method with Systematized Nomenclature of Medicine - Clinical Terms (SNOMED CT) to process the texts into encoded and structured contents to help the analytic system in the OHCs to discover the biomedical knowledge. Based on our experimental result, the proposed method can effectively extract the biomedical knowledge from the posts and enhance the ability of clinical knowledge discovery online to improve the patient support.

The intuitionistic fuzzy set which is characterized by a membership degree and a non-membership degree, is a very powerful and useful tool to cope with fuzziness and uncertainty. Recently, the Pythagorean fuzzy set which is an extension of the intuitionistic fuzzy set has been introduced. In this paper, we focus on multi-attribute decision making with Pythagorean fuzzy information. First of all, considering the Bonferroni mean aggregation operator can take the interrelationship between arguments being fused into account, we introduce some Bonferroni mean-based aggregation operators for fusing Pythagorean fuzzy information. Then, a novel approach to multi-attribute decision making problems is proposed and discussed. Finally, an instance of supply chain management is provided to illustrate the validity of the new approach.

Access control is an important technical method to protect the sensitive data in the information system. This paper mainly focuses on the issue of privacy preserving for patients' information in HIS. On the basis of providing hospital employees necessary patient information that can support the treatment, the proposed Knowledge-Constrained Role Based Access Control (KC-RBAC) model tries to reduce the scope of patients' information that can be accessed by hospital employees. Compared with the traditional RBAC model, the medical knowledge and nonmedical knowledge lying in the process of treatment are introduced into KC-RBAC, which outline the boundary of accessible data for different users in the system.