Sci Rep. 2024 Oct 19;14(1):24605. doi: 10.1038/s41598-024-75242-2.

ABSTRACT

A variety of reasons have made it more difficult for educators and tutors to anticipate students' performance. Numerous researchers have used various predictive models to identify students who may be at-risk of dropping out early. Additionally, these methods were used to forecast final semester grades based on various datasets. However, these prediction models still fall short of meeting educational management requirements. In this paper, we propose the deep learning (DL) based model named students academic performance prediction network (SAPPNet) to predict the students' grades. We consider the questionnaire-based Jordan University dataset which contains demographic information, usage of digital tools before and after COVID-19, sleep times before and after COVID-19, social interaction, psychological state, and academic performance. SAPPNet consists of spatial convolution modules which are designed to extract spatial dependencies includes categorical and numerical attributes that represent static features (gender, level/year, age, digital tools used before and after COVID-19, psychological condition using prolonged e-learning tools) and temporal module for temporal dependencies involves sequences that capture changes before and after COVID-19. Additionally, we also try to implement classical machine learning (ML) models including support vector machine, k nearest neighbor, decision tree, and random forest, and DL models named artificial neural network, convolutional neural network, long short-term memory, and students learning prediction network. Simulation results show that SAPPNet achieved the best performance compared to state-of-the-art methods, with an accuracy, precision, recall, and an F1-score of 93 % . The proposed model with spatial and temporal modules improves the prediction performance, and it implies new aspect of the educational dataset.

PMID:39427025 | DOI:10.1038/s41598-024-75242-2

J Physiol. 2024 Oct 19. doi: 10.1113/JP287001. Online ahead of print.

ABSTRACT

The present study was designed to test the potential utility of regional cerebral oxygen saturation (rcSO2) in detecting term infants with brain injury. The study also examined whether quantitative rcSO2 features are associated with grade of hypoxic ischaemic encephalopathy (HIE). We analysed 58 term infants with HIE (>36 weeks of gestational age) enrolled in a prospective observational study. All newborn infants had a period of continuous rcSO2 monitoring and magnetic resonance imaging (MRI) assessment during the first week of life. rcSO2 Signals were pre-processed and quantitative features were extracted. Machine-learning and deep-learning models were developed to detect adverse outcome (brain injury on MRI or death in the first week) using the leave-one-out cross-validation approach and to assess the association between rcSO2 and HIE grade (modified Sarnat - at 1 h). The machine-learning model (rcSO2 excluding prolonged relative desaturations) significantly detected infant MRI outcome or death in the first week of life [area under the curve (AUC) = 0.73, confidence interval (CI) = 0.59-0.86, Matthew's correlation coefficient = 0.35]. In agreement, deep learning models detected adverse outcome with an AUC = 0.64, CI = 0.50-0.79. We also report a significant association between rcSO2 features and HIE grade using a machine learning approach (AUC = 0.81, CI = 0.73-0.90). We conclude that automated analysis of rcSO2 using machine learning methods in term infants with HIE was able to determine, with modest accuracy, infants with adverse outcome. De novo approaches to signal analysis of NIRS holds promise to aid clinical decision making in the future. KEY POINTS: Hypoxic-induced neonatal brain injury contributes to both short- and long-term functional deficits. Non-invasive continuous monitoring of brain oxygenation using near-infrared- spectroscopy offers a potential new insight to the development of serious injury. In this study, characteristics of the NIRS signal were summarised using either predefined features or data-driven feature extraction, both were combined with a machine learning approach to predict short-term brain injury. Using data from a cohort of term infants with hypoxic ischaemic encephalopathy, the present study illustrates that automated analysis of regional cerebral oxygen saturation rcSO2, using either machine learning or deep learning methods, was able to determine infants with adverse outcome.

PMID:39425751 | DOI:10.1113/JP287001

ACS Nano. 2024 Oct 19. doi: 10.1021/acsnano.4c09312. Online ahead of print.

ABSTRACT

Precise size and shape control in nanocrystal synthesis is essential for utilizing nanocrystals in various industrial applications, such as catalysis, sensing, and energy conversion. However, traditional ensemble measurements often overlook the subtle size and shape distributions of individual nanocrystals, hindering the establishment of robust structure-property relationships. In this study, we uncover intricate shape evolutions and growth mechanisms in Co3O4 nanocrystal synthesis at a subnanometer scale, enabled by deep-learning-assisted statistical characterization. By first controlling synthetic parameters such as cobalt precursor concentration and water amount then using high resolution electron microscopy imaging to identify the geometric features of individual nanocrystals, this study provides insights into the interplay between synthesis conditions and the size-dependent shape evolution in colloidal nanocrystals. Utilizing population-wide imaging data encompassing over 441,067 nanocrystals, we analyze their characteristics and elucidate previously unobserved size-resolved shape evolution. This high-throughput statistical analysis is essential for representing the entire population accurately and enables the study of the size dependency of growth regimes in shaping nanocrystals. Our findings provide experimental quantification of the growth regime transition based on the size of the crystals, specifically (i) for faceting and (ii) from thermodynamic to kinetic, as evidenced by transitions from convex to concave polyhedral crystals. Additionally, we introduce the concept of an "onset radius," which describes the critical size thresholds at which these transitions occur. This discovery has implications beyond achieving nanocrystals with desired morphology; it enables finely tuned correlation between geometry and material properties, advancing the field of colloidal nanocrystal synthesis and its applications.

PMID:39425689 | DOI:10.1021/acsnano.4c09312

Electromagn Biol Med. 2024 Oct 19:1-16. doi: 10.1080/15368378.2024.2415089. Online ahead of print.

ABSTRACT

Human being's biological processes and psychological activities are jointly connected to the brain. So, the examination of human activity is more significant for the well-being of humans. There are various models for brain activity detection considering neuroimaging for attaining decreased time requirement, increased control commands, and enhanced accuracy. Motor Imagery (MI)-based Brain-Computer Interface (BCI) systems create a way in which the brain can interact with the environment by processing Electroencephalogram (EEG) signals. Human Activity Recognition (HAR) deals with identifying the physiological activities of human beings based on sensory signals. This survey reviews the different methods available for HAR based on MI-EEG signals. A total of 50 research articles based on HAR from EEG signals are considered in this survey. This survey discusses the challenges faced by various techniques for HAR. Moreover, the papers are assessed considering various parameters, techniques, publication year, performance metrics, utilized tools, employed databases, etc. There were many techniques developed to solve the problem of HAR and they are classified as Machine Learning (ML) and Deep Learning (DL)models. At last, the research gaps and limitations of the techniques were discussed that contribute to developing an effective HAR.

PMID:39425602 | DOI:10.1080/15368378.2024.2415089

Acta Ophthalmol. 2024 Oct 19. doi: 10.1111/aos.16781. Online ahead of print.

ABSTRACT

PURPOSE: To validate the performance of autonomous diabetic retinopathy (DR) grading by comparing a human grader and a self-developed deep-learning (DL) algorithm with gold-standard evaluation.

METHODS: We included 500, 6-field retinal images graded by an expert ophthalmologist (gold standard) according to the International Clinical Diabetic Retinopathy Disease Severity Scale as represented with DR levels 0-4 (97, 100, 100, 103, 100, respectively). Weighted kappa was calculated to measure the DR classification agreement for (1) a certified human grader without, and (2) with assistance from a DL algorithm and (3) the DL operating autonomously. Using any DR (level 0 vs. 1-4) as a cutoff, we calculated sensitivity, specificity, as well as positive and negative predictive values (PPV and NPV). Finally, we assessed lesion discrepancies between Model 3 and the gold standard.

RESULTS: As compared to the gold standard, weighted kappa for Models 1-3 was 0.88, 0.89 and 0.72, sensitivities were 95%, 94% and 78% and specificities were 82%, 84% and 81%. Extrapolating to a real-world DR prevalence of 23.8%, the PPV were 63%, 64% and 57% and the NPV were 98%, 98% and 92%. Discrepancies between the gold standard and Model 3 were mainly incorrect detection of artefacts (n = 49), missed microaneurysms (n = 26) and inconsistencies between the segmentation and classification (n = 51).

CONCLUSION: While the autonomous DL algorithm for DR classification only performed on par with a human grader for some measures in a high-risk population, extrapolations to a real-world population demonstrated an excellent 92% NPV, which could make it clinically feasible to use autonomously to identify non-DR patients.

PMID:39425597 | DOI:10.1111/aos.16781

BMC Med Inform Decis Mak. 2024 Oct 18;24(1):304. doi: 10.1186/s12911-024-02695-w.

ABSTRACT

BACKGROUND: Determining the optimal timing of surgical intervention for Neonatal necrotizing enterocolitis (NEC) poses significant challenges. This study develops a predictive model using the long short-term memory network (LSTM) with a focal loss (FL) to identify infants at risk of developing Bell IIB + NEC early and issue timely surgical warnings.

METHODS: Data from 791 neonates diagnosed with NEC are gathered from the Neonatal Intensive Care Unit (NICU), encompassing 35 selected features. Infants are categorized into those requiring surgical intervention (n = 257) and those managed medically (n = 534) based on the Mod-Bell criteria. A fivefold cross-validation approach is employed for training and testing. The LSTM algorithm is utilized to capture and utilize temporal relationships in the dataset, with FL employed as a loss function to address class imbalance. Model performance metrics include precision, recall, F1 score, and average precision (AP).

RESULTS: The model tested on a real dataset demonstrated high performance. Predicting surgical risk 1 day in advance achieved precision (0.913 ± 0.034), recall (0.841 ± 0.053), F1 score (0.874 ± 0.029), and AP (0.917 ± 0.025). The 2-days-in-advance predictions yielded (0.905 ± 0.036), recall (0.815 ± 0.057), F1 score (0.857 ± 0.035), and AP (0.905 ± 0.029).

CONCLUSION: The LSTM model with FL exhibits high precision and recall in forecasting the need for surgical intervention 1 or 2 days ahead. This predictive capability holds promise for enhancing infants' outcomes by facilitating timely clinical decisions.

PMID:39425161 | DOI:10.1186/s12911-024-02695-w

BMC Med Educ. 2024 Oct 18;24(1):1166. doi: 10.1186/s12909-024-06076-9.

ABSTRACT

BACKGROUND: Artificial intelligence (AI) is transforming health profession education (HPE) through personalized learning technologies. HPE students must also learn about AI to understand its impact on healthcare delivery. We examined HPE students' AI-related knowledge and attitudes, and perceived challenges in integrating AI in HPE.

METHODS: This cross-sectional included medical, nursing, physiotherapy, and clinical nutrition students from four public universities in Jordan, the Kingdom of Saudi Arabia (KSA), the United Arab Emirates (UAE), and Egypt. Data were collected between February and October 2023 via an online survey that covered five main domains: benefits of AI in healthcare, negative impact on patient trust, negative impact on the future of healthcare professionals, inclusion of AI in HPE curricula, and challenges hindering integration of AI in HPE.

RESULTS: Of 642 participants, 66.4% reported low AI knowledge levels. The UAE had the largest proportion of students with low knowledge (72.7%). The majority (54.4%) of participants had learned about AI outside their curriculum, mainly through social media (66%). Overall, 51.2% expressed positive attitudes toward AI, with Egypt showing the largest proportion of positive attitudes (59.1%). Although most participants viewed AI in healthcare positively (91%), significant variations were observed in other domains. The majority (77.6%) supported integrating AI in HPE, especially in Egypt (82.3%). A perceived negative impact of AI on patient trust was expressed by 43.5% of participants, particularly in Egypt (54.7%). Only 18.1% of participants were concerned about the impact of AI on future healthcare professionals, with the largest proportion from Egypt (33.0%). Some participants (34.4%) perceived AI integration as challenging, notably in the UAE (47.6%). Common barriers included lack of expert training (53%), awareness (50%), and interest in AI (41%).

CONCLUSION: This study clarified key considerations when integrating AI in HPE. Enhancing students' awareness and fostering innovation in an AI-driven medical landscape are crucial for effectively incorporating AI in HPE curricula.

PMID:39425151 | DOI:10.1186/s12909-024-06076-9

BMC Med Imaging. 2024 Oct 18;24(1):281. doi: 10.1186/s12880-024-01459-2.

ABSTRACT

BACKGROUND: This study aimed to evaluate the effectiveness of DeepLabv3+with Squeeze-and-Excitation (DeepLabv3+SE) architectures for segmenting the choroid in optical coherence tomography (OCT) images of patients with diabetic retinopathy.

METHODS: A total of 300 B-scans were selected from 21 patients with mild to moderate diabetic retinopathy. Six DeepLabv3+SE variants, each utilizing a different pre-trained convolutional neural network (CNN) for feature extraction, were compared. Segmentation performance was assessed using the Jaccard index, Dice score (DSC), precision, recall, and F1-score. Binarization and Bland-Altman analysis were employed to evaluate the agreement between automated and manual measurements of choroidal area, luminal area (LA), and Choroidal Vascularity Index (CVI).

RESULTS: DeepLabv3+SE with EfficientNetB0 achieved the highest segmentation performance, with a Jaccard index of 95.47, DSC of 98.29, precision of 98.80, recall of 97.41, and F1-score of 98.10 on the validation set. Bland-Altman analysis indicated good agreement between automated and manual measurements of LA and CVI.

CONCLUSIONS: DeepLabv3+SE with EfficientNetB0 demonstrates promise for accurate choroid segmentation in OCT images. This approach offers a potential solution for automated CVI calculation in diabetic retinopathy patients. Further evaluation of the proposed method on a larger and more diverse dataset can strengthen its generalizability and clinical applicability.

PMID:39425019 | DOI:10.1186/s12880-024-01459-2

BMC Bioinformatics. 2024 Oct 18;25(1):333. doi: 10.1186/s12859-024-05951-y.

ABSTRACT

BACKGROUND: Relation extraction (RE) plays a crucial role in biomedical research as it is essential for uncovering complex semantic relationships between entities in textual data. Given the significance of RE in biomedical informatics and the increasing volume of literature, there is an urgent need for advanced computational models capable of accurately and efficiently extracting these relationships on a large scale.

RESULTS: This paper proposes a novel approach, SARE, combining ensemble learning Stacking and attention mechanisms to enhance the performance of biomedical relation extraction. By leveraging multiple pre-trained models, SARE demonstrates improved adaptability and robustness across diverse domains. The attention mechanisms enable the model to capture and utilize key information in the text more accurately. SARE achieved performance improvements of 4.8, 8.7, and 0.8 percentage points on the PPI, DDI, and ChemProt datasets, respectively, compared to the original BERT variant and the domain-specific PubMedBERT model.

CONCLUSIONS: SARE offers a promising solution for improving the accuracy and efficiency of relation extraction tasks in biomedical research, facilitating advancements in biomedical informatics. The results suggest that combining ensemble learning with attention mechanisms is effective for extracting complex relationships from biomedical texts. Our code and data are publicly available at: https://github.com/GS233/Biomedical .

PMID:39425010 | DOI:10.1186/s12859-024-05951-y

Nat Chem Biol. 2024 Oct 18. doi: 10.1038/s41589-024-01754-7. Online ahead of print.

NO ABSTRACT

PMID:39424957 | DOI:10.1038/s41589-024-01754-7

Sci Rep. 2024 Oct 18;14(1):24441. doi: 10.1038/s41598-024-75531-w.

ABSTRACT

Cardiac arrhythmias are the leading cause of death and pose a huge health and economic burden globally. Electrocardiography (ECG) is an effective technique for the diagnosis of cardiovascular diseases because of its noninvasive and cost-effective advantages. However, traditional ECG analysis relies heavily on the clinical experience of physicians, which can be challenging and time-consuming to produce valid diagnostic results. This work proposes a new hybrid deep learning model that combines convolutional neural network (CNN) and bidirectional gated recurrent unit (BiGRU) with multi-head attention (CBGM model). Specifically, the model consists of seven convolutional layers with varying filter sizes (4, 16, 32, and 64) and three pooling layers, respectively, while the BiGRU module includes two layers with 64 units each followed by multi-head attention (8-heads). The combination of CNN and BiGRU effectively captures spatio-temporal features of ECG signals, with multi-head attention comprehensively extracted global correlations among multiple segments of ECG signals. The validation in the MIT-BIH arrhythmia database achieved an accuracy of 99.41%, a precision of 99.15%, a specificity of 99.68%, and an F1-Score of 99.21%, indicating its robust performance across different evaluation metrics. Additionally, the model's performance was evaluated on the PTB Diagnostic ECG Database, where it achieved an accuracy of 98.82%, demonstrating its generalization capability. Comparative analysis against previous methods revealed that our proposed CBGM model exhibits more higher performance in automatic classification of arrhythmia and can be helpful for assisting clinicians by enabling real-time detection of cardiac arrhythmias during routine ECG screenings.

PMID:39424921 | DOI:10.1038/s41598-024-75531-w

J Phys Chem A. 2024 Oct 18. doi: 10.1021/acs.jpca.4c03074. Online ahead of print.

ABSTRACT

Melting in the deep rocky parts of planets plays an important role in geological processes such as planet formation, seismicity, magnetic field generation, thermal convection, and crustal evolution. Such processes are the key way to understanding the dynamics of planetary interiors and the history as well as mechanisms of planetary evolution. We herein investigate the melting curves and pressure-temperature (P-T)-phase diagrams for CaO3, a candidate mineral for the lower mantle, by means of the deep learning potential model. Using first-principles, molecular dynamics, and quasi-harmonic approximation, the reliability of the deep learning potential model is verified by calculating the high-temperature and high-pressure equations of state and phase transition pressures for the orthorhombic and tetragonal structures of CaO3 described by space groups Aea2 and P4̅21m, respectively. The melting temperatures of 975, 850, and 755 K at zero pressure are obtained by the single-phase, void, and two-phase methods, respectively, and their melting temperatures are analyzed by the radial distribution function and mean-square displacement to analyze the melting process. Finally, the melting phase diagrams of CaO3 at 0-135 GPa were obtained by the two-phase method.

PMID:39423322 | DOI:10.1021/acs.jpca.4c03074

Bioinformatics. 2024 Oct 18:btae617. doi: 10.1093/bioinformatics/btae617. Online ahead of print.

ABSTRACT

MOTIVATION: Embryo selection is one of the critical factors in determining the success of pregnancy in in vitro fertilization (IVF) procedures. Using artificial intelligence to aid in embryo selection could effectively address the current time-consuming, expensive, subjectively influenced process of embryo assessment by trained embryologists. However, current deep learning-based methods often focus on blastocyst segmentation, grading, or predicting cell development via time-lapse videos, often overlooking morphokinetic parameters or lacking interpretability. Given the significance of both morphokinetic and morphological evaluation in predicting the implantation potential of cleavage-stage embryos, as emphasized by previous research, there is a necessity for an automated method to segment cleavage-stage embryos to improve this process.

RESULTS: In this article, we introduce the SAM-based Dual Branch Segmentation Pipeline for automated segmentation of blastomeres in cleavage-stage embryos. Leveraging the powerful segmentation capability of SAM, the instance branch conducts instance segmentation of blastomeres, while the semantic branch performs semantic segmentation of fragments. Due to the lack of publicly available datasets, we construct the CleavageEmbryo dataset, the first dataset of human cleavage-stage embryos with pixel-level annotations containing fragment information. We train and test a series of state-of-the-art segmentation algorithms on CleavageEmbryo. Our experiments demonstrate that our method outperforms existing algorithms in terms of objective metrics (mAP 0.748 on blastomeres, Dice 0.694 on fragments) and visual quality, enabling more accurate segmentation of cleavage-stage embryos.

AVAILABILITY AND IMPLEMENTATION: The code and sample data in this study can be found at: Https://github.com/12austincc/Cleavage-StageEmbryoSegmentation.

SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PMID:39423150 | DOI:10.1093/bioinformatics/btae617

IEEE Trans Med Imaging. 2024 Oct 18;PP. doi: 10.1109/TMI.2024.3483451. Online ahead of print.

ABSTRACT

4D cone-beam computed tomography (CBCT) plays a critical role in adaptive radiation therapy for lung cancer. However, extremely sparse sampling projection data will cause severe streak artifacts in 4D CBCT images. Existing deep learning (DL) methods heavily rely on large labeled training datasets which are difficult to obtain in practical scenarios. Restricted by this dilemma, DL models often struggle with simultaneously retaining dynamic motions, removing streak degradations, and recovering fine details. To address the above challenging problem, we introduce a Deep Prior Image Constrained Motion Compensation framework (DPI-MoCo) that decouples the 4D CBCT reconstruction into two sub-tasks including coarse image restoration and structural detail fine-tuning. In the first stage, the proposed DPI-MoCo combines the prior image guidance, generative adversarial network, and contrastive learning to globally suppress the artifacts while maintaining the respiratory movements. After that, to further enhance the local anatomical structures, the motion estimation and compensation technique is adopted. Notably, our framework is performed without the need for paired datasets, ensuring practicality in clinical cases. In the Monte Carlo simulation dataset, the DPI-MoCo achieves competitive quantitative performance compared to the state-of-the-art (SOTA) methods. Furthermore, we test DPI-MoCo in clinical lung cancer datasets, and experiments validate that DPI-MoCo not only restores small anatomical structures and lesions but also preserves motion information.

PMID:39423082 | DOI:10.1109/TMI.2024.3483451

IEEE Trans Neural Netw Learn Syst. 2024 Oct 18;PP. doi: 10.1109/TNNLS.2024.3476439. Online ahead of print.

ABSTRACT

Hyperparameter recommendation via meta-learning has shown great promise in various studies. The main challenge for meta-learning is how to develop an effective meta-learner (learning algorithm) that can capture the intrinsic relationship between dataset characteristics and the empirical performance of hyperparameters. Existing meta-learners are mostly based on traditional machine-learning models that only learn data representations with a single layer, which are incapable of learning complex features from the data and often cannot capture those properties deeply embedded in data. To address this issue, in this article, we propose hyperparameter recommendation approaches by integrating the learning model with convolutional neural networks (CNNs). Specifically, we first formulate the recommendation task as a regression problem, where dataset characteristics are treated as predictors and the historical performance of hyperparameters as responses. We establish a CNN-based learning model with feature selection capability to serve as the regressor. We then develop a convolutional denoising autoencoder (ConvDAE) that can leverage the spatial structure of the entire hyperparameter performance space and evaluate the performance of hyperparameters via denoising when the performance of partial hyperparameters is available under the multidimensional framework. To make our approach being flexible in applications, we establish a comprehensive two-branch CNN model that can utilize both dataset characteristics and partial evaluations to make effective recommendations. We conduct extensive experiments on 400 real classification problems and the well-known SVM. Our proposed approaches outperform existing meta-learning baselines as well as various search algorithms, demonstrating the high effectiveness in hyperparameter recommendations via deep learning.

PMID:39423079 | DOI:10.1109/TNNLS.2024.3476439

IEEE J Biomed Health Inform. 2024 Oct 18;PP. doi: 10.1109/JBHI.2024.3483301. Online ahead of print.

ABSTRACT

This study introduces an innovative deep-learning model for cuffless blood pressure estimation using PPG and ECG signals, demonstrating state-of-the-art performance on the largest clean dataset, PulseDB. The rU-Net architecture, a fusion of U-Net and ResNet, enhances both generalization and feature extraction accuracy. Accurate multi-scale feature capture is facilitated by short-time Fourier transform (STFT) time-frequency distributions and multi-head attention mechanisms, allowing data-driven feature selection. The inclusion of demographic parameters as supervisory information further elevates performance. On the calibration-based dataset, our model excels, achieving outstanding accuracy (SBP MAE ± std: 4.49 ± 4.86 mmHg, DBP MAE ± std: 2.69 ± 3.10 mmHg), surpassing AAMI standards and earning a BHS Grade A rating. Addressing the challenge of calibration-free data, we propose a fine-tuning-based transfer learning approach. Remarkably, with only 10% data transfer, our model attains exceptional accuracy (SBP MAE ± std: 4.14 ± 5.01 mmHg, DBP MAE ± std: 2.48 ± 2.93 mmHg). This study sets the stage for the development of highly accurate and reliable wearable cuffless blood pressure monitoring devices.

PMID:39423074 | DOI:10.1109/JBHI.2024.3483301

IEEE J Biomed Health Inform. 2024 Oct 18;PP. doi: 10.1109/JBHI.2024.3483316. Online ahead of print.

ABSTRACT

Tumor heterogeneity presents a significant challenge in predicting drug responses, especially as missense mutations within the same gene can lead to varied outcomes such as drug resistance, enhanced sensitivity, or therapeutic ineffectiveness. These complex relationships highlight the need for advanced analytical approaches in oncology. Due to their powerful ability to handle heterogeneous data, graph convolutional networks (GCNs) represent a promising approach for predicting drug responses. However, simple bipartite graphs cannot accurately capture the complex relationships involved in missense mutation and drug response. Furthermore, Deep learning models for drug response are often considered "black boxes", and their interpretability remains a widely discussed issue. To address these challenges, we propose an Interpretable Dynamic Directed Graph Convolutional Network (IDDGCN) framework, which incorporates four key features: (1) the use of directed graphs to differentiate between sensitivity and resistance relationships, (2) the dynamic updating of node weights based on node-specific interactions, (3) the exploration of associations between different mutations within the same gene and drug response, and (4) the enhancement of interpretability models through the integration of a weighted mechanism that accounts for the biological significance, alongside a ground truth construction method to evaluate prediction transparency. The experimental results demonstrate that IDDGCN outperforms existing state-of-the-art models, exhibiting excellent predictive power. Both qualitative and quantitative evaluations of its interpretability further highlight its ability to explain predictions, offering a fresh perspective for precision oncology and targeted drug development.

PMID:39423073 | DOI:10.1109/JBHI.2024.3483316

Med Phys. 2024 Sep;51(9):6534. doi: 10.1002/mp.17332. Epub 2024 Jul 30.

NO ABSTRACT

PMID:39423010 | DOI:10.1002/mp.17332

J Med Internet Res. 2024 Oct 18;26:e51110. doi: 10.2196/51110.

ABSTRACT

BACKGROUND: The COVID-19 pandemic has profoundly affected mental health, leading to an increased prevalence of depression and insomnia. Currently, artificial intelligence (AI) and deep learning have thoroughly transformed health care-related mobile apps, offered more effective mental health support, and alleviated the psychological stress that may have emerged during the pandemic. Early reviews outlined the use of mobile apps for dealing with depression and insomnia separately. However, there is now an urgent need for a systematic evaluation of mobile apps that address both depression and insomnia to reveal new applications and research gaps.

OBJECTIVE: This study aims to systematically review and evaluate mobile apps targeting depression and insomnia, highlighting their features, effectiveness, and gaps in the current research.

METHODS: We systematically searched PubMed, Scopus, and Web of Science for peer-reviewed journal articles published between 2017 and 2023. The inclusion criteria were studies that (1) focused on mobile apps addressing both depression and insomnia, (2) involved young people or adult participants, and (3) provided data on treatment efficacy. Data extraction was independently conducted by 2 reviewers. Title and abstract screening, as well as full-text screening, were completed in duplicate. Data were extracted by a single reviewer and verified by a second reviewer, and risk of bias assessments were completed accordingly.

RESULTS: Of the initial 383 studies we found, 365 were excluded after title, abstract screening, and removal of duplicates. Eventually, 18 full-text articles met our criteria and underwent full-text screening. The analysis revealed that mobile apps related to depression and insomnia were primarily utilized for early detection, assessment, and screening (n=5 studies); counseling and psychological support (n=3 studies); and cognitive behavioral therapy (CBT; n=10 studies). Among the 10 studies related to depression, our findings showed that chatbots demonstrated significant advantages in improving depression symptoms, a promising development in the field. Additionally, 2 studies evaluated the effectiveness of mobile apps as alternative interventions for depression and sleep, further expanding the potential applications of this technology.

CONCLUSIONS: The integration of AI and deep learning into mobile apps, particularly chatbots, is a promising avenue for personalized mental health support. Through innovative features, such as early detection, assessment, counseling, and CBT, these apps significantly contribute toward improving sleep quality and addressing depression. The reviewed chatbots leveraged advanced technologies, including natural language processing, machine learning, and generative dialog, to provide intelligent and autonomous interactions. Compared with traditional face-to-face therapies, their feasibility, acceptability, and potential efficacy highlight their user-friendly, cost-effective, and accessible nature with the aim of enhancing sleep and mental health outcomes.

PMID:39423009 | DOI:10.2196/51110

Med Phys. 2024 Sep;51(9):5911-5926. doi: 10.1002/mp.17252. Epub 2024 Jun 22.

ABSTRACT

BACKGROUND: Auxiliary diagnosis of different types of cystic lung diseases (CLDs) is important in the clinic and is instrumental in facilitating early and specific treatments. Current clinical methods heavily depend on accumulated experience, restricting their applicability in regions with less developed medical resources. Thus, how to realize the computer-aided diagnosis of CLDs is of great clinical value.

PURPOSE: This work proposed a deep learning-based method for automatically segmenting the lung parenchyma in computed tomography (CT) slice images and accurately diagnosing the CLDs using CT scans.

METHODS: A two-stage deep learning method was proposed for the automatic classification of normal cases and five different CLDs using CT scans. Lung parenchyma segmentation is the foundation of CT image analysis and auxiliary diagnosis. To meet the requirements of different sizes of the lung parenchyma, an adaptive region-growing and improved U-Net model was employed for mask acquisition and automatic segmentation. The former was achieved by a self-designed adaptive seed point selection method based on similarity measurement, and the latter introduced multiscale input and multichannel output into the original U-Net model and effectively achieved the lightweight design by adjusting the structure and parameters. After that, the middle 30 consecutive CT slice images of each sample were segmented to obtain lung parenchyma, which was employed for training and testing the proposed multichannel parallel input recursive MLP-Mixer network (MPIRMNet) model, achieving the computer-aided diagnosis of CLDs.

RESULTS: A total of 4718 and 16 290 CT slice images collected from 543 patients were employed to validate the proposed segmentation and classification methods, respectively. Experimental results showed that the improved U-Net model can accurately segment the lung parenchyma in CT slice images, with the Dice, precision, volumetric overlap error, and relative volume difference of 0.96 ± 0.01, 0.93 ± 0.04, 0.05 ± 0.02, and 0.05 ± 0.03, respectively. Meanwhile, the proposed MPIRMNet model achieved appreciable classification effect for normal cases and different CLDs, with the accuracy, sensitivity, specificity, and F1 score of 0.8823 ± 0.0324, 0.8897 ± 0.0325, 0.9746 ± 0.0078, and 0.8831 ± 0.0334, respectively. Compared with classical machine learning and convolutional neural networks-based methods for this task, the proposed classification method had a preferable performance, with a significant improvement of accuracy of 10.74%.

CONCLUSIONS: The work introduced a two-stage deep learning method, which can achieve the segmentation of lung parenchyma and the classification of CLDs. Compared to previous diagnostic tasks targeting single CLD, this work can achieve various CLDs' diagnosis in the early stage, thereby achieving targeted treatment and increasing the potential and value of clinical applications.

PMID:39422997 | DOI:10.1002/mp.17252