Environ Sci Ecotechnol. 2024 Jul 8;22:100450. doi: 10.1016/j.ese.2024.100450. eCollection 2024 Nov.

ABSTRACT

Rainfall data with high spatial and temporal resolutions are essential for urban hydrological modeling. Ubiquitous surveillance cameras can continuously record rainfall events through video and audio, so they have been recognized as potential rain gauges to supplement professional rainfall observation networks. Since video-based rainfall estimation methods can be affected by variable backgrounds and lighting conditions, audio-based approaches could be a supplement without suffering from these conditions. However, most audio-based approaches focus on rainfall-level classification rather than rainfall intensity estimation. Here, we introduce a dataset named Surveillance Audio Rainfall Intensity Dataset (SARID) and a deep learning model for estimating rainfall intensity. First, we created the dataset through audio of six real-world rainfall events. This dataset's audio recordings are segmented into 12,066 pieces and annotated with rainfall intensity and environmental information, such as underlying surfaces, temperature, humidity, and wind. Then, we developed a deep learning-based baseline using Mel-Frequency Cepstral Coefficients (MFCC) and Transformer architecture to estimate rainfall intensity from surveillance audio. Validated from ground truth data, our baseline achieves a root mean absolute error of 0.88 mm h-1 and a coefficient of correlation of 0.765. Our findings demonstrate the potential of surveillance audio-based models as practical and effective tools for rainfall observation systems, initiating a new chapter in rainfall intensity estimation. It offers a novel data source for high-resolution hydrological sensing and contributes to the broader landscape of urban sensing, emergency response, and resilience.

PMID:39161573 | PMC:PMC11331698 | DOI:10.1016/j.ese.2024.100450

Front Genet. 2024 Aug 2;15:1378809. doi: 10.3389/fgene.2024.1378809. eCollection 2024.

ABSTRACT

Introduction: Developing effective breast cancer survival prediction models is critical to breast cancer prognosis. With the widespread use of next-generation sequencing technologies, numerous studies have focused on survival prediction. However, previous methods predominantly relied on single-omics data, and survival prediction using multi-omics data remains a significant challenge. Methods: In this study, considering the similarity of patients and the relevance of multi-omics data, we propose a novel multi-omics stacked fusion network (MSFN) based on a stacking strategy to predict the survival of breast cancer patients. MSFN first constructs a patient similarity network (PSN) and employs a residual graph neural network (ResGCN) to obtain correlative prognostic information from PSN. Simultaneously, it employs convolutional neural networks (CNNs) to obtain specificity prognostic information from multi-omics data. Finally, MSFN stacks the prognostic information from these networks and feeds into AdaboostRF for survival prediction. Results: Experiments results demonstrated that our method outperformed several state-of-the-art methods, and biologically validated by Kaplan-Meier and t-SNE.

PMID:39161422 | PMC:PMC11331006 | DOI:10.3389/fgene.2024.1378809

Front Bioeng Biotechnol. 2024 Aug 5;12:1454728. doi: 10.3389/fbioe.2024.1454728. eCollection 2024.

ABSTRACT

Jaw cyst is a fluid-containing cystic lesion that can occur in any part of the jaw and cause facial swelling, dental lesions, jaw fractures, and other associated issues. Due to the diversity and complexity of jaw images, existing deep-learning methods still have challenges in segmentation. To this end, we propose MARes-Net, an innovative multi-scale attentional residual network architecture. Firstly, the residual connection is used to optimize the encoder-decoder process, which effectively solves the gradient disappearance problem and improves the training efficiency and optimization ability. Secondly, the scale-aware feature extraction module (SFEM) significantly enhances the network's perceptual abilities by extending its receptive field across various scales, spaces, and channel dimensions. Thirdly, the multi-scale compression excitation module (MCEM) compresses and excites the feature map, and combines it with contextual information to obtain better model performance capabilities. Furthermore, the introduction of the attention gate module marks a significant advancement in refining the feature map output. Finally, rigorous experimentation conducted on the original jaw cyst dataset provided by Quzhou People's Hospital to verify the validity of MARes-Net architecture. The experimental data showed that precision, recall, IoU and F1-score of MARes-Net reached 93.84%, 93.70%, 86.17%, and 93.21%, respectively. Compared with existing models, our MARes-Net shows its unparalleled capabilities in accurately delineating and localizing anatomical structures in the jaw cyst image segmentation.

PMID:39161348 | PMC:PMC11330813 | DOI:10.3389/fbioe.2024.1454728

Phys Rev E. 2024 Jul;110(1-1):014121. doi: 10.1103/PhysRevE.110.014121.

ABSTRACT

Deriving closed-form analytical expressions for reduced-order models, and judiciously choosing the closures leading to them, has long been the strategy of choice for studying phase- and noise-induced transitions for agent-based models (ABMs). In this paper, we propose a data-driven framework that pinpoints phase transitions for an ABM-the Desai-Zwanzig model-in its mean-field limit, using a smaller number of variables than traditional closed-form models. To this end, we use the manifold learning algorithm Diffusion Maps to identify a parsimonious set of data-driven latent variables, and we show that they are in one-to-one correspondence with the expected theoretical order parameter of the ABM. We then utilize a deep learning framework to obtain a conformal reparametrization of the data-driven coordinates that facilitates, in our example, the identification of a single parameter-dependent ordinary differential equation (ODE) in these coordinates. We identify this ODE through a residual neural network inspired by a numerical integration scheme (forward Euler). We then use the identified ODE-enabled through an odd symmetry transformation-to construct the bifurcation diagram exhibiting the phase transition.

PMID:39160966 | DOI:10.1103/PhysRevE.110.014121

Healthc Inform Res. 2024 Jul;30(3):234-243. doi: 10.4258/hir.2024.30.3.234. Epub 2024 Jul 31.

ABSTRACT

OBJECTIVES: This study aimed to optimize early coronary heart disease (CHD) prediction using a genetic algorithm (GA)-based convolutional neural network (CNN) feature engineering approach. We sought to overcome the limitations of traditional hyperparameter optimization techniques by leveraging a GA for superior predictive performance in CHD detection.

METHODS: Utilizing a GA for hyperparameter optimization, we navigated a complex combinatorial space to identify optimal configurations for a CNN model. We also employed information gain for feature selection optimization, transforming the CHD datasets into an image-like input for the CNN architecture. The efficacy of this method was benchmarked against traditional optimization strategies.

RESULTS: The advanced GA-based CNN model outperformed traditional methods, achieving a substantial increase in accuracy. The optimized model delivered a promising accuracy range, with a peak of 85% in hyperparameter optimization and 100% accuracy when integrated with machine learning algorithms, namely naïve Bayes, support vector machine, decision tree, logistic regression, and random forest, for both binary and multiclass CHD prediction tasks.

CONCLUSIONS: The integration of a GA into CNN feature engineering is a powerful technique for improving the accuracy of CHD predictions. This approach results in a high degree of predictive reliability and can significantly contribute to the field of AI-driven healthcare, with the possibility of clinical deployment for early CHD detection. Future work will focus on expanding the approach to encompass a wider set of CHD data and potential integration with wearable technology for continuous health monitoring.

PMID:39160782 | DOI:10.4258/hir.2024.30.3.234

Sci Rep. 2024 Aug 19;14(1):19139. doi: 10.1038/s41598-024-70115-0.

ABSTRACT

In coal mining areas, surface subsidence poses significant risks to human life and property. Fortunately, surface subsidence caused by coal mining can be monitored and predicted by using various methods, e.g., probability integral method and deep learning (DL) methods. Although DL methods show promise in predicting subsidence, they often lack accuracy due to insufficient consideration of spatial correlation and temporal nonlinearity. Considering this issue, we propose a novel DL-based approach for predicting mining surface subsidence. Our method employs K-means clustering to partition spatial data, allowing the application of a gate recurrent unit (GRU) model to capture nonlinear relationships in subsidence time series within each partition. Optimization using snake optimization (SO) further enhances model accuracy globally. Validation shows our method outperforms traditional Long Short-Term Memory (LSTM) and GRU models, achieving 99.1% of sample pixels with less than 8 mm absolute error.

PMID:39160327 | DOI:10.1038/s41598-024-70115-0

Sci Rep. 2024 Aug 16;14(1):18990. doi: 10.1038/s41598-024-69814-5.

ABSTRACT

Temporomandibular joint disorders are prevalent causes of orofacial discomfort. Diagnosis predominantly relies on assessing the configuration and positions of temporomandibular joint components in magnetic resonance images. The complex anatomy of the temporomandibular joint, coupled with the variability in magnetic resonance image quality, often hinders an accurate diagnosis. To surmount this challenge, we developed deep learning models tailored to the automatic segmentation of temporomandibular joint components, including the temporal bone, disc, and condyle. These models underwent rigorous training and validation utilizing a dataset of 3693 magnetic resonance images from 542 patients. Upon evaluation, our ensemble model, which combines five individual models, yielded average Dice similarity coefficients of 0.867, 0.733, 0.904, and 0.952 for the temporal bone, disc, condyle, and background class during internal testing. In the external validation, the average Dice similarity coefficients values for the temporal bone, disc, condyle, and background were 0.720, 0.604, 0.800, and 0.869, respectively. When applied in a clinical setting, these artificial intelligence-augmented tools enhanced the diagnostic accuracy of physicians, especially when discerning between temporomandibular joint anterior disc displacement and osteoarthritis. In essence, automated temporomandibular joint segmentation by our deep learning approach, stands as a promising aid in refining temporomandibular joint disorders diagnosis and treatment strategies.

PMID:39160234 | DOI:10.1038/s41598-024-69814-5

Med Eng Phys. 2024 Aug;130:104206. doi: 10.1016/j.medengphy.2024.104206. Epub 2024 Jul 5.

ABSTRACT

Epilepsy is one of the most common brain diseases, characterised by repeated seizures that occur on a regular basis. During a seizure, a patient's muscles flex uncontrollably, causing a loss of mobility and balance, which can be harmful or even fatal. Developing an automatic approach for warning patients of oncoming seizures necessitates substantial research. Analyzing the electroencephalogram (EEG) output from the human brain's scalp region can help predict seizures. EEG data were analyzed to extract time domain features such as Hurst exponent (Hur), Tsallis entropy (TsEn), enhanced permutation entropy (impe), and amplitude-aware permutation entropy (AAPE). In order to automatically diagnose epileptic seizure in children from normal children, this study conducted two sessions. In the first session, the extracted features from the EEG dataset were classified using three machine learning (ML)-based models, including support vector machine (SVM), K nearest neighbor (KNN), or decision tree (DT), and in the second session, the dataset was classified using three deep learning (DL)-based recurrent neural network (RNN) classifiers in The EEG dataset was obtained from the Neurology Clinic of the Ibn Rushd Training Hospital. In this regard, extensive explanations and research from the time domain and entropy characteristics demonstrate that employing GRU, LSTM, and BiLSTM RNN deep learning classifiers on the All-time-entropy fusion feature improves the final classification results.

PMID:39160030 | DOI:10.1016/j.medengphy.2024.104206

Med Eng Phys. 2024 Aug;130:104196. doi: 10.1016/j.medengphy.2024.104196. Epub 2024 Jun 15.

ABSTRACT

The 12-lead electrocardiogram (ECG) is widely used for diagnosing cardiovascular diseases in clinical practice. Recently, deep learning methods have become increasingly effective for automatically classifying ECG signals. However, most current research simply combines the 12-lead ECG signals into a matrix without fully considering the intrinsic relationships between the leads and the heart's structure. To better utilize medical domain knowledge, we propose a multi-branch network for multi-label ECG classification and introduce an intuitive and effective lead grouping strategy. Correspondingly, we design multi-branch networks where each branch employs a multi-scale convolutional network structure to extract more comprehensive features, with each branch corresponding to a lead combination. To better integrate features from different leads, we propose a feature weighting fusion module. We evaluate our method on the PTB-XL dataset for classifying 4 arrhythmia types and normal rhythm, and on the China Physiological Signal Challenge 2018 (CPSC2018) database for classifying 8 arrhythmia types and normal rhythm. Experimental results on multiple multi-label datasets demonstrate that our proposed multi-branch network outperforms state-of-the-art networks in multi-label classification tasks.

PMID:39160024 | DOI:10.1016/j.medengphy.2024.104196

Med Eng Phys. 2024 Aug;130:104209. doi: 10.1016/j.medengphy.2024.104209. Epub 2024 Jul 17.

ABSTRACT

As the number of patients with cardiovascular diseases (CVDs) increases annually, a reliable and automated system for detecting electrocardiogram (ECG) abnormalities is becoming increasingly essential. Scholars have developed numerous methods of arrhythmia classification using machine learning or deep learning. However, the issue of low classification rates of individual classes in inter-patient heartbeat classification remains a challenge. This study proposes a method for inter-patient heartbeat classification by fusing dual-channel squeeze-and-excitation residual neural networks (SE-ResNet) and expert features. In the preprocessing stage, ECG heartbeats extracted from both leads of ECG signals are filtered and normalized. Additionally, nine features representing waveform morphology and heartbeat contextual information are selected to be fused with the deep neural networks. Using different filter and kernel sizes for each block, the SE-residual block-based model can effectively learn long-term features between heartbeats. The divided ECG heartbeats and extracted features are then input to the improved SE-ResNet for training and testing according to the inter-patient scheme. The focal loss is utilized to handle the heartbeat of the imbalance category. The proposed arrhythmia classification method is evaluated on three open-source databases, and it achieved an overall F1-score of 83.39 % in the MIT-BIH database. This system can be applied in the scenario of daily monitoring of ECG and plays a significant role in diagnosing arrhythmias.

PMID:39160018 | DOI:10.1016/j.medengphy.2024.104209

Radiother Oncol. 2024 Aug 17:110483. doi: 10.1016/j.radonc.2024.110483. Online ahead of print.

ABSTRACT

INTRODUCTION: New radiotherapy machines such as Halcyon are capable of delivering dose-rate of 600 monitor-units per minute, allowing large numbers of patients treated per day. However, patient-specific quality assurance (QA) is still required, which dramatically decrease machine availability. Innovative artificial intelligence (AI) algorithms could predict QA result based on complexity metrics. However, no AI solution exists for Halcyon machines and the complexity metrics to be used have not been definitively determined. The aim of this study was to develop an AI solution capable of firstly determining the complexity indices to be obtained and secondly predicting patient-specific QA in a routine clinical setting.

METHODS: Three hundred and eighteen beams from 56 patients with breast cancer were used. The seven complexity indices named Modulation-Complexity-Score (MCS), Small-Aperture-Score (SAS10), Beam-Area (BA), Beam-Irregularity (BI), Beam-Modulation (BM), Gantry and Collimator angles were used as input to the AI model. Machine learning (ML) and deep learning (DL) models using tensorflow were set up to predict DreamDose QA conformance.

RESULTS: MCS, BI, gantry and collimator angle are not correlated with QA compliance. Therefore, ML and DL models were trained using SAS10, BA and BM complexity indices. ROC analyses enabled to find best predicted probability threshold to increase specificity and sensitivity. ML models did not show satisfactory performance with an area under-the-curve (AUC) of 0.75 and specificity and sensitivity of 0.88 and 0.86. However, optimised DL model showed better performance with an AUC of 0.95 and specificity and sensitivity of 0.98 and 0.97.

CONCLUSION: The DL model demonstrated a high degree of accuracy in its predictions of the quality assurance (QA) results. Our online predictive QA-platform offers significant time savings in terms of accelerator occupancy and working time.

PMID:39159677 | DOI:10.1016/j.radonc.2024.110483

Accid Anal Prev. 2024 Aug 18;207:107748. doi: 10.1016/j.aap.2024.107748. Online ahead of print.

ABSTRACT

Driving risk prediction emerges as a pivotal technology within the driving safety domain, facilitating the formulation of targeted driving intervention strategies to enhance driving safety. The driving safety undergoes continuous evolution in response to the complexities of the traffic environment, representing a dynamic and ongoing serialization process. The evolutionary trend of this sequence offers valuable information pertinent to driving safety research. However, existing research on driving risk prediction has primarily concentrated on forecasting a single index, such as the driving safety level or the extreme value within a specified future timeframe. This approach often neglects the intrinsic properties that characterize the temporal evolution of driving safety. Leveraging the high-D natural driving dataset, this study employs the multi-step time series forecasting methodology to predict the risk evolution sequence throughout the car-following process, elucidates the benefits of the multi-step time series forecasting approach, and contrasts the predictive efficacy on driving safety levels across various temporal windows. The empirical findings demonstrate that the time series prediction model proficiently captures essential dynamics such as risk evolution trends, amplitudes, and turning points. Consequently, it provides predictions that are significantly more robust and comprehensive than those obtained from a single risk index. The TsLeNet proposed in this study integrates a 2D convolutional network architecture with a dual attention mechanism, adeptly capturing and synthesizing multiple features across time steps. This integration significantly enhances the prediction precision at each temporal interval. Comparative analyses with other mainstream models reveal that TsLeNet achieves the best performance in terms of prediction accuracy and efficiency. Concurrently, this research undertakes a comprehensive analysis of the temporal distribution of errors, the impact pattern of features on risk sequence, and the applicability of interaction features among surrounding vehicles. The adoption of multi-step time series forecasting approach not only offers a novel perspective for analyzing and exploring driving safety, but also furnishes the design and development of targeted driving intervention systems.

PMID:39159592 | DOI:10.1016/j.aap.2024.107748

J Biomech. 2024 Aug 13;174:112271. doi: 10.1016/j.jbiomech.2024.112271. Online ahead of print.

ABSTRACT

Bone and cartilage tissues are physiologically dynamic organs that are systematically regulated by mechanical inputs. At cellular level, mechanical stimulation engages an intricate network where mechano-sensors and transmitters cooperate to manipulate downstream signaling. Despite accumulating evidence, there is a notable underutilization of available information, due to limited integration and analysis. In this context, we conceived an interactive web tool named MechanoBone to introduce a new avenue of literature-based discovery. Initially, we compiled a literature database by sourcing content from Pubmed and processing it through the Natural Language Toolkit project, Pubtator, and a custom library. We identified direct co-occurrence among entities based on existing evidence, archiving in a relational database via SQLite. Latent connections were then quantified by leveraging the Link Prediction algorithm. Secondly, mechanobiological pathway maps were generated, and an entity-pathway correlation scoring system was established through weighted algorithm based on our database, String, and KEGG, predicting potential functions of specific entities. Additionally, we established a mechanical circumstance-based exploration to sort genes by their relevance based on big data, revealing the potential mechanically sensitive factors in bone research and future clinical applications. In conclusion, MechanoBone enables: 1) interpreting mechanobiological processes; 2) identifying correlations and crosstalk among molecules and pathways under specific mechanical conditions; 3) connecting clinical applications with mechanobiological processes in bone research. It offers a literature mining tool with visualization and interactivity, facilitating targeted molecule navigation and prediction within the mechanobiological framework of bone-related cells, thereby enhancing knowledge sharing and big data analysis in the biomedical realm.

PMID:39159585 | DOI:10.1016/j.jbiomech.2024.112271

Comput Biol Med. 2024 Aug 18;180:109025. doi: 10.1016/j.compbiomed.2024.109025. Online ahead of print.

ABSTRACT

INTRODUCTION: In the treatment of malocclusion, continuous monitoring of the three-dimensional relationship between dental roots and the surrounding alveolar bone is essential for preventing complications from orthodontic procedures. Cone-beam computed tomography (CBCT) provides detailed root and bone data, but its high radiation dose limits its frequent use, consequently necessitating an alternative for ongoing monitoring.

OBJECTIVES: We aimed to develop a deep learning-based cross-temporal multimodal image fusion system for acquiring root and jawbone information without additional radiation, enhancing the ability of orthodontists to monitor risk.

METHODS: Utilizing CBCT and intraoral scans (IOSs) as cross-temporal modalities, we integrated deep learning with multimodal fusion technologies to develop a system that includes a CBCT segmentation model for teeth and jawbones. This model incorporates a dynamic kernel prior model, resolution restoration, and an IOS segmentation network optimized for dense point clouds. Additionally, a coarse-to-fine registration module was developed. This system facilitates the integration of IOS and CBCT images across varying spatial and temporal dimensions, enabling the comprehensive reconstruction of root and jawbone information throughout the orthodontic treatment process.

RESULTS: The experimental results demonstrate that our system not only maintains the original high resolution but also delivers outstanding segmentation performance on external testing datasets for CBCT and IOSs. CBCT achieved Dice coefficients of 94.1 % and 94.4 % for teeth and jawbones, respectively, and it achieved a Dice coefficient of 91.7 % for the IOSs. Additionally, in the context of real-world registration processes, the system achieved an average distance error (ADE) of 0.43 mm for teeth and 0.52 mm for jawbones, significantly reducing the processing time.

CONCLUSION: We developed the first deep learning-based cross-temporal multimodal fusion system, addressing the critical challenge of continuous risk monitoring in orthodontic treatments without additional radiation exposure. We hope that this study will catalyze transformative advancements in risk management strategies and treatment modalities, fundamentally reshaping the landscape of future orthodontic practice.

PMID:39159544 | DOI:10.1016/j.compbiomed.2024.109025

Breast Cancer Res. 2024 Aug 19;26(1):124. doi: 10.1186/s13058-024-01863-0.

ABSTRACT

BACKGROUND: Human epidermal growth factor receptor 2 (HER2)-low breast cancer has emerged as a new subtype of tumor, for which novel antibody-drug conjugates have shown beneficial effects. Assessment of HER2 requires several immunohistochemistry tests with an additional in situ hybridization test if a case is classified as HER2 2+. Therefore, novel cost-effective methods to speed up the HER2 assessment are highly desirable.

METHODS: We used a self-supervised attention-based weakly supervised method to predict HER2-low directly from 1437 histopathological images from 1351 breast cancer patients. We built six distinct models to explore the ability of classifiers to distinguish between the HER2-negative, HER2-low, and HER2-high classes in different scenarios. The attention-based model was used to comprehend the decision-making process aimed at relevant tissue regions.

RESULTS: Our results indicate that the effectiveness of classification models hinges on the consistency and dependability of assay-based tests for HER2, as the outcomes from these tests are utilized as the baseline truth for training our models. Through the use of explainable AI, we reveal histologic patterns associated with the HER2 subtypes.

CONCLUSION: Our findings offer a demonstration of how deep learning technologies can be applied to identify HER2 subgroup statuses, potentially enriching the toolkit available for clinical decision-making in oncology.

PMID:39160593 | DOI:10.1186/s13058-024-01863-0

BMC Med Imaging. 2024 Aug 19;24(1):218. doi: 10.1186/s12880-024-01389-z.

ABSTRACT

Uterine fibroids are common benign tumors originating from the uterus's smooth muscle layer, often leading to symptoms such as pelvic pain, and reproductive issues. Early detection is crucial to prevent complications such as infertility or the need for invasive treatments like hysterectomy. One of the main challenges in diagnosing uterine fibroids is the lack of specific symptoms, which can mimic other gynecological conditions. This often leads to under-diagnosis or misdiagnosis, delaying appropriate management. In this research, an attention based fine-tuned EfficientNetB0 model is proposed for the classification of uterine fibroids from ultrasound images. Attention mechanisms, permit the model to focus on particular parts of an image and move forward the model's execution by empowering it to specifically go to imperative highlights whereas overlooking irrelevant ones. The proposed approach has used a total of 1990 images divided into two classes: Non-uterine fibroid and uterine fibroid. The data augmentation methods have been connected to improve generalization and strength by exposing it to a wider range of varieties within the training data. The proposed model has obtained the value of accuracy as 0.99. Future research should focus on improving the accuracy and efficiency of diagnostic techniques, as well as evaluating their effectiveness in diverse populations with higher sensitivity and specificity for the detection of uterine fibroids, as well as biomarkers to aid in diagnosis.

PMID:39160500 | DOI:10.1186/s12880-024-01389-z

J Imaging Inform Med. 2024 Aug 19. doi: 10.1007/s10278-024-01209-4. Online ahead of print.

ABSTRACT

Anemia is a significant global health issue, affecting over a billion people worldwide, according to the World Health Organization. Generally, the gold standard for diagnosing anemia relies on laboratory measurements of hemoglobin. To meet the need in clinical practice, physicians often rely on visual examination of specific areas, such as conjunctiva, to assess pallor. However, this method is subjective and relies on the physician's experience. Therefore, we proposed a deep learning prediction model based on three input images from different body parts, namely, conjunctiva, palm, and fingernail. By incorporating additional body part labels and employing a fusion attention mechanism, the model learns and enhances the salient features of each body part during training, enabling it to produce reliable results. Additionally, we employ a dual loss function that allows the regression model to benefit from well-established classification methods, thereby achieving stable handling of minority samples. We used a retrospective data set (EYES-DEFY-ANEMIA) to develop this model called Body-Part-Anemia Network (BPANet). The BPANet showed excellent performance in detecting anemia, with accuracy of 0.849 and an F1-score of 0.828. Our multi-body-part model has been validated on a prospectively collected data set of 101 patients in National Taiwan University Hospital. The prediction accuracy as well as F1-score can achieve as high as 0.716 and 0.788, respectively. To sum up, we have developed and validated a novel non-invasive hemoglobin prediction model based on image input from multiple body parts, with the potential of real-time use at home and in clinical settings.

PMID:39160365 | DOI:10.1007/s10278-024-01209-4

Invest Radiol. 2024 Aug 20. doi: 10.1097/RLI.0000000000001114. Online ahead of print.

ABSTRACT

Recent technological advancements have revolutionized routine brain magnetic resonance imaging (MRI) sequences, offering enhanced diagnostic capabilities in intracranial disease evaluation. This review explores 2 pivotal breakthrough areas: deep learning reconstruction (DLR) and quantitative MRI techniques beyond conventional structural imaging. DLR using deep neural networks facilitates accelerated imaging with improved signal-to-noise ratio and spatial resolution, enhancing image quality with short scan times. DLR focuses on supervised learning applied to clinical implementation and applications. Quantitative MRI techniques, exemplified by 2D multidynamic multiecho, 3D quantification using interleaved Look-Locker acquisition sequences with T2 preparation pulses, and magnetic resonance fingerprinting, enable precise calculation of brain-tissue parameters and further advance diagnostic accuracy and efficiency. Potential DLR instabilities and quantification and bias limitations will be discussed. This review underscores the synergistic potential of DLR and quantitative MRI, offering prospects for improved brain imaging beyond conventional methods.

PMID:39159333 | DOI:10.1097/RLI.0000000000001114

IEEE Trans Neural Syst Rehabil Eng. 2024 Aug 19;PP. doi: 10.1109/TNSRE.2024.3445936. Online ahead of print.

ABSTRACT

Auditory Brainstem Response (ABR) is an evoked potential in the brainstem's neural centers in response to sound stimuli. Clinically, characteristic waves, especially Wave V latency, extracted from ABR can objectively indicate auditory loss and diagnose diseases. Several methods have been developed for the extraction of characteristic waves. To ensure the effectiveness of the method, most of the methods are time-consuming and rely on the heavy workloads of clinicians. To reduce the workload of clinicians, automated extraction methods have been developed. However, the above methods also have limitations. This study introduces a novel deep learning network for automatic extraction of Wave V latency, named ABR-Attention. ABR-Attention model includes a self-attention module, first and second-derivative attention module, and regressor module. Experiments are conducted on the accuracy with 10-fold cross-validation, the effects on different sound pressure levels (SPLs), the effects of different error scales and the effects of ablation. ABR-Attention shows efficacy in extracting Wave V latency of ABR, with an overall accuracy of 96.76±0.41% and an error scale of 0.1ms, and provides a new solution for objective localization of ABR characteristic waves.

PMID:39159023 | DOI:10.1109/TNSRE.2024.3445936

IEEE Trans Med Imaging. 2024 Aug 19;PP. doi: 10.1109/TMI.2024.3446450. Online ahead of print.

ABSTRACT

MRI, a widespread non-invasive medical imaging modality, is highly sensitive to patient motion. Despite many attempts over the years, motion correction remains a difficult problem and there is no general method applicable to all situations. We propose a retrospective method for motion estimation and correction to tackle the problem of in-plane rigid-body motion, apt for classical 2D Spin-Echo scans of the brain, which are regularly used in clinical practice. Due to the sequential acquisition of k-space, motion artifacts are well localized. The method leverages the power of deep neural networks to estimate motion parameters in k-space and uses a model-based approach to restore degraded images to avoid "hallucinations". Notable advantages are its ability to estimate motion occurring in high spatial frequencies without the need of a motion-free reference. The proposed method operates on the whole k-space dynamic range and is moderately affected by the lower SNR of higher harmonics. As a proof of concept, we provide models trained using supervised learning on 600k motion simulations based on motion-free scans of 43 different subjects. Generalization performance was tested with simulations as well as in-vivo. Qualitative and quantitative evaluations are presented for motion parameter estimations and image reconstruction. Experimental results show that our approach is able to obtain good generalization performance on simulated data and in-vivo acquisitions. We provide a Python implementation at https://gitlab.unige.ch/Oscar.Dabrowski/sismik_mri/.

PMID:39159019 | DOI:10.1109/TMI.2024.3446450