Knee Surg Sports Traumatol Arthrosc. 2024 Mar 1. doi: 10.1002/ksa.12085. Online ahead of print.

ABSTRACT

Deep learning is a subset of artificial intelligence (AI) with enormous potential to transform orthopaedic surgery. As has already become evident with the deployment of Large Language Models (LLMs) like ChatGPT (OpenAI Inc.), deep learning can rapidly enter clinical and surgical practices. As such, it is imperative that orthopaedic surgeons acquire a deeper understanding of the technical terminology, capabilities and limitations associated with deep learning models. The focus of this series thus far has been providing surgeons with an overview of the steps needed to implement a deep learning-based pipeline, emphasizing some of the important technical details for surgeons to understand as they encounter, evaluate or lead deep learning projects. However, this series would be remiss without providing practical examples of how deep learning models have begun to be deployed and highlighting the areas where the authors feel deep learning may have the most profound potential. While computer vision applications of deep learning were the focus of Parts I and II, due to the enormous impact that natural language processing (NLP) has had in recent months, NLP-based deep learning models are also discussed in this final part of the series. In this review, three applications that the authors believe can be impacted the most by deep learning but with which many surgeons may not be familiar are discussed: (1) registry construction, (2) diagnostic AI and (3) data privacy. Deep learning-based registry construction will be essential for the development of more impactful clinical applications, with diagnostic AI being one of those applications likely to augment clinical decision-making in the near future. As the applications of deep learning continue to grow, the protection of patient information will become increasingly essential; as such, applications of deep learning to enhance data privacy are likely to become more important than ever before. Level of Evidence: Level IV.

PMID:38426614 | DOI:10.1002/ksa.12085

Med Phys. 2024 Mar 1. doi: 10.1002/mp.17012. Online ahead of print.

ABSTRACT

BACKGROUND: Deep learning based optical coherence tomography (OCT) segmentation methods have achieved excellent results, allowing quantitative analysis of large-scale data. However, OCT images are often acquired by different devices or under different imaging protocols, which leads to serious domain shift problem. This in turn results in performance degradation of segmentation models.

PURPOSE: Aiming at the domain shift problem, we propose a two-stage adversarial learning based network (TSANet) that accomplishes unsupervised cross-domain OCT segmentation.

METHODS: In the first stage, a Fourier transform based approach is adopted to reduce image style differences from the image level. Then, adversarial learning networks, including a segmenter and a discriminator, are designed to achieve inter-domain consistency in the segmentation output. In the second stage, pseudo labels of selected unlabeled target domain training data are used to fine-tune the segmenter, which further improves its generalization capability. The proposed method was tested on cross-domain datasets for choroid or retinoschisis segmentation tasks. For choroid segmentation, the model was trained on 400 images and validated on 100 images from the source domain, and then trained on 1320 unlabeled images and tested on 330 images from target domain I, and also trained on 400 unlabeled images and tested on 200 images from target domain II. For retinoschisis segmentation, the model was trained on 1284 images and validated on 312 images from the source domain, and then trained on 1024 unlabeled images and tested on 200 images from the target domain.

RESULTS: The proposed method achieved significantly improved results over that without domain adaptation, with improvement of 8.34%, 55.82% and 3.53% in intersection over union (IoU) respectively for the three test sets. The performance is better than some state-of-the-art domain adaptation methods.

CONCLUSIONS: The proposed TSANet, with image level adaptation, feature level adaptation and pseudo-label based fine-tuning, achieved excellent cross-domain generalization. This alleviates the burden of obtaining additional manual labels when adapting the deep learning model to new OCT data.

PMID:38426594 | DOI:10.1002/mp.17012

Brief Bioinform. 2024 Jan 22;25(2):bbae068. doi: 10.1093/bib/bbae068.

ABSTRACT

Cluster assignment is vital to analyzing single-cell RNA sequencing (scRNA-seq) data to understand high-level biological processes. Deep learning-based clustering methods have recently been widely used in scRNA-seq data analysis. However, existing deep models often overlook the interconnections and interactions among network layers, leading to the loss of structural information within the network layers. Herein, we develop a new self-supervised clustering method based on an adaptive multi-scale autoencoder, called scAMAC. The self-supervised clustering network utilizes the Multi-Scale Attention mechanism to fuse the feature information from the encoder, hidden and decoder layers of the multi-scale autoencoder, which enables the exploration of cellular correlations within the same scale and captures deep features across different scales. The self-supervised clustering network calculates the membership matrix using the fused latent features and optimizes the clustering network based on the membership matrix. scAMAC employs an adaptive feedback mechanism to supervise the parameter updates of the multi-scale autoencoder, obtaining a more effective representation of cell features. scAMAC not only enables cell clustering but also performs data reconstruction through the decoding layer. Through extensive experiments, we demonstrate that scAMAC is superior to several advanced clustering and imputation methods in both data clustering and reconstruction. In addition, scAMAC is beneficial for downstream analysis, such as cell trajectory inference. Our scAMAC model codes are freely available at https://github.com/yancy2024/scAMAC.

PMID:38426327 | DOI:10.1093/bib/bbae068

Brief Bioinform. 2024 Jan 22;25(2):bbae061. doi: 10.1093/bib/bbae061.

ABSTRACT

Cancer is a complex and high-mortality disease regulated by multiple factors. Accurate cancer subtyping is crucial for formulating personalized treatment plans and improving patient survival rates. The underlying mechanisms that drive cancer progression can be comprehensively understood by analyzing multi-omics data. However, the high noise levels in omics data often pose challenges in capturing consistent representations and adequately integrating their information. This paper proposed a novel variational autoencoder-based deep learning model, named Deeply Integrating Latent Consistent Representations (DILCR). Firstly, multiple independent variational autoencoders and contrastive loss functions were designed to separate noise from omics data and capture latent consistent representations. Subsequently, an Attention Deep Integration Network was proposed to integrate consistent representations across different omics levels effectively. Additionally, we introduced the Improved Deep Embedded Clustering algorithm to make integrated variable clustering friendly. The effectiveness of DILCR was evaluated using 10 typical cancer datasets from The Cancer Genome Atlas and compared with 14 state-of-the-art integration methods. The results demonstrated that DILCR effectively captures the consistent representations in omics data and outperforms other integration methods in cancer subtyping. In the Kidney Renal Clear Cell Carcinoma case study, cancer subtypes were identified by DILCR with significant biological significance and interpretability.

PMID:38426322 | DOI:10.1093/bib/bbae061

Front Physiol. 2024 Feb 15;15:1329145. doi: 10.3389/fphys.2024.1329145. eCollection 2024.

ABSTRACT

Background: Manual bone age assessment (BAA) is associated with longer interpretation time and higher cost and variability, thus posing challenges in areas with restricted medical facilities, such as the high-altitude Tibetan Plateau. The application of artificial intelligence (AI) for automating BAA could facilitate resolving this issue. This study aimed to develop an AI-based BAA model for Han and Tibetan children. Methods: A model named "EVG-BANet" was trained using three datasets, including the Radiology Society of North America (RSNA) dataset (training set n = 12611, validation set n = 1425, and test set n = 200), the Radiological Hand Pose Estimation (RHPE) dataset (training set n = 5491, validation set n = 713, and test set n = 79), and a self-established local dataset [training set n = 825 and test set n = 351 (Han n = 216 and Tibetan n = 135)]. An open-access state-of-the-art model BoNet was used for comparison. The accuracy and generalizability of the two models were evaluated using the abovementioned three test sets and an external test set (n = 256, all were Tibetan). Mean absolute difference (MAD) and accuracy within 1 year were used as indicators. Bias was evaluated by comparing the MAD between the demographic groups. Results: EVG-BANet outperformed BoNet in the MAD on the RHPE test set (0.52 vs. 0.63 years, p < 0.001), the local test set (0.47 vs. 0.62 years, p < 0.001), and the external test set (0.53 vs. 0.66 years, p < 0.001) and exhibited a comparable MAD on the RSNA test set (0.34 vs. 0.35 years, p = 0.934). EVG-BANet achieved accuracy within 1 year of 97.7% on the local test set (BoNet 90%, p < 0.001) and 89.5% on the external test set (BoNet 85.5%, p = 0.066). EVG-BANet showed no bias in the local test set but exhibited a bias related to chronological age in the external test set. Conclusion: EVG-BANet can accurately predict the bone age (BA) for both Han children and Tibetan children living in the Tibetan Plateau with limited healthcare facilities.

PMID:38426209 | PMC:PMC10902452 | DOI:10.3389/fphys.2024.1329145

Front Physiol. 2024 Feb 15;15:1293380. doi: 10.3389/fphys.2024.1293380. eCollection 2024.

ABSTRACT

Background and Purpose: Precisely assessing the likelihood of an intracranial aneurysm rupturing is critical for guiding clinical decision-making. The objective of this study is to construct and validate a deep learning framework utilizing point clouds to forecast the likelihood of aneurysm rupturing. Methods: The dataset included in this study consisted of a total of 623 aneurysms, with 211 of them classified as ruptured and 412 as unruptured, which were obtained from two separate projects within the AneuX morphology database. The HUG project, which included 124 ruptured aneurysms and 340 unruptured aneurysms, was used to train and internally validate the model. For external validation, another project named @neurIST was used, which included 87 ruptured and 72 unruptured aneurysms. A standardized method was employed to isolate aneurysms and a segment of their parent vessels from the original 3D vessel models. These models were then converted into a point cloud format using open3d package to facilitate training of the deep learning network. The PointNet++ architecture was utilized to process the models and generate risk scores through a softmax layer. Finally, two models, the dome and cut1 model, were established and then subjected to a comprehensive comparison of statistical indices with the LASSO regression model built by the dataset authors. Results: The cut1 model outperformed the dome model in the 5-fold cross-validation, with the mean AUC values of 0.85 and 0.81, respectively. Furthermore, the cut1 model beat the morphology-based LASSO regression model with an AUC of 0.82. However, as the original dataset authors stated, we observed potential generalizability concerns when applying trained models to datasets with different selection biases. Nevertheless, our method outperformed the LASSO regression model in terms of generalizability, with an AUC of 0.71 versus 0.67. Conclusion: The point cloud, as a 3D visualization technique for intracranial aneurysms, can effectively capture the spatial contour and morphological aspects of aneurysms. More structural features between the aneurysm and its parent vessels can be exposed by keeping a portion of the parent vessels, enhancing the model's performance. The point cloud-based deep learning model exhibited good performance in predicting rupture risk while also facing challenges in generalizability.

PMID:38426204 | PMC:PMC10901972 | DOI:10.3389/fphys.2024.1293380

Front Neurol. 2024 Feb 15;15:1359033. doi: 10.3389/fneur.2024.1359033. eCollection 2024.

ABSTRACT

INTRODUCTION: T2 mapping can characterize peripheral neuropathy and muscle denervation due to axonal damage. Three-dimensional double echo steady-state (DESS) can simultaneously provide 3D qualitative information and T2 maps with equivalent spatial resolution. However, insufficient signal-to-noise ratio may bias DESS-T2 values. Deep learning reconstruction (DLR) techniques can reduce noise, and hence may improve quantitation of high-resolution DESS-T2. This study aims to (i) evaluate the effect of DLR methods on DESS-T2 values, and (ii) to evaluate the feasibility of using DESS-T2 maps to differentiate abnormal from normal nerves and muscles in the upper extremities, with abnormality as determined by electromyography.

METHODS AND RESULTS: Analysis of images from 25 subjects found that DLR decreased DESS-T2 values in abnormal muscles (DLR = 37.71 ± 9.11 msec, standard reconstruction = 38.56 ± 9.44 msec, p = 0.005) and normal muscles (DLR: 27.18 ± 6.34 msec, standard reconstruction: 27.58 ± 6.34 msec, p < 0.001) consistent with a noise reduction bias. Mean DESS-T2, both with and without DLR, was higher in abnormal nerves (abnormal = 75.99 ± 38.21 msec, normal = 35.10 ± 9.78 msec, p < 0.001) and muscles (abnormal = 37.71 ± 9.11 msec, normal = 27.18 ± 6.34 msec, p < 0.001). A higher DESS-T2 in muscle was associated with electromyography motor unit recruitment (p < 0.001).

DISCUSSION: These results suggest that quantitative DESS-T2 is improved by DLR and can differentiate the nerves and muscles involved in peripheral neuropathies from those uninvolved.

PMID:38426170 | PMC:PMC10902120 | DOI:10.3389/fneur.2024.1359033

PeerJ. 2024 Feb 26;12:e17006. doi: 10.7717/peerj.17006. eCollection 2024.

ABSTRACT

Single-cell omics sequencing has rapidly advanced, enabling the quantification of diverse omics profiles at a single-cell resolution. To facilitate comprehensive biological insights, such as cellular differentiation trajectories, precise annotation of cell subtypes is essential. Conventional methods involve clustering cells and manually assigning subtypes based on canonical markers, a labor-intensive and expert-dependent process. Hence, an automated computational prediction framework is crucial. While several classification frameworks for predicting cell subtypes from single-cell RNA sequencing datasets exist, these methods solely rely on single-omics data, offering insights at a single molecular level. They often miss inter-omic correlations and a holistic understanding of cellular processes. To address this, the integration of multi-omics datasets from individual cells is essential for accurate subtype annotation. This article introduces moSCminer, a novel framework for classifying cell subtypes that harnesses the power of single-cell multi-omics sequencing datasets through an attention-based neural network operating at the omics level. By integrating three distinct omics datasets-gene expression, DNA methylation, and DNA accessibility-while accounting for their biological relationships, moSCminer excels at learning the relative significance of each omics feature. It then transforms this knowledge into a novel representation for cell subtype classification. Comparative evaluations against standard machine learning-based classifiers demonstrate moSCminer's superior performance, consistently achieving the highest average performance on real datasets. The efficacy of multi-omics integration is further corroborated through an in-depth analysis of the omics-level attention module, which identifies potential markers for cell subtype annotation. To enhance accessibility and scalability, moSCminer is accessible as a user-friendly web-based platform seamlessly connected to a cloud system, publicly accessible at http://203.252.206.118:5568. Notably, this study marks the pioneering integration of three single-cell multi-omics datasets for cell subtype identification.

PMID:38426141 | PMC:PMC10903350 | DOI:10.7717/peerj.17006

Front Cardiovasc Med. 2024 Feb 15;11:1327179. doi: 10.3389/fcvm.2024.1327179. eCollection 2024.

ABSTRACT

BACKGROUND: Artificial intelligence (AI) has shown promise in the early detection of various cardiac conditions from a standard 12-lead electrocardiogram (ECG). However, the ability of AI to identify abnormalities from single-lead recordings across a range of pathological conditions remains to be systematically investigated. This study aims to assess the performance of a convolutional neural network (CNN) using a single-lead (D1) rather than a standard 12-lead setup for accurate identification of ECG abnormalities.

METHODS: We designed and trained a lightweight CNN to identify 20 different cardiac abnormalities on ECGs, using data from the PTB-XL dataset. With a relatively simple architecture, the network was designed to accommodate different combinations of leads as input (<100,000 learnable parameters). We compared various lead setups such as the standard 12-lead, D1 alone, and D1 paired with an additional lead.

RESULTS: The CNN based on single-lead ECG (D1) outperformed the one based on the standard 12-lead framework [with an average percentage difference of the area under the curve (AUC) of -8.7%]. Notably, for certain diagnostic classes, there was no difference in the diagnostic AUC between the single-lead and the standard 12-lead setups. When a second lead was detected in the CNN in addition to D1, the AUC gap was further reduced to an average percentage difference of -2.8% compared with that of the standard 12-lead setup.

CONCLUSIONS: A relatively lightweight CNN can predict different classes of cardiac abnormalities from D1 alone and the standard 12-lead ECG. Considering the growing availability of wearable devices capable of recording a D1-like single-lead ECG, we discuss how our findings contribute to the foundation of a large-scale screening of cardiac abnormalities.

PMID:38426118 | PMC:PMC10901971 | DOI:10.3389/fcvm.2024.1327179

Front Radiol. 2024 Feb 15;4:1339612. doi: 10.3389/fradi.2024.1339612. eCollection 2024.

ABSTRACT

Image-to-text radiology report generation aims to automatically produce radiology reports that describe the findings in medical images. Most existing methods focus solely on the image data, disregarding the other patient information accessible to radiologists. In this paper, we present a novel multi-modal deep neural network framework for generating chest x-rays reports by integrating structured patient data, such as vital signs and symptoms, alongside unstructured clinical notes. We introduce a conditioned cross-multi-head attention module to fuse these heterogeneous data modalities, bridging the semantic gap between visual and textual data. Experiments demonstrate substantial improvements from using additional modalities compared to relying on images alone. Notably, our model achieves the highest reported performance on the ROUGE-L metric compared to relevant state-of-the-art models in the literature. Furthermore, we employed both human evaluation and clinical semantic similarity measurement alongside word-overlap metrics to improve the depth of quantitative analysis. A human evaluation, conducted by a board-certified radiologist, confirms the model's accuracy in identifying high-level findings, however, it also highlights that more improvement is needed to capture nuanced details and clinical context.

PMID:38426080 | PMC:PMC10902135 | DOI:10.3389/fradi.2024.1339612

Data Brief. 2024 Feb 13;53:110182. doi: 10.1016/j.dib.2024.110182. eCollection 2024 Apr.

ABSTRACT

Plant diseases pose a significant obstacle to global agricultural productivity, impacting crop quality yield and causing substantial economic losses for farmers. Watermelon, a commonly cultivated succulent vine plant, is rich in hydration and essential nutrients. However, it is susceptible to various diseases due to unfavorable environmental conditions and external factors, leading to compromised quality and substantial financial setbacks. Swift identification and management of crop diseases are imperative to minimize losses, enhance yield, reduce costs, and bolster agricultural output. Conventional disease diagnosis methods are often labor-intensive, time-consuming, ineffective, and prone to subjectivity. As a result, there is a critical need to advance research into machine-based models for disease detection in watermelons. This paper presents a large dataset of watermelons that can be used to train a machine vision-based illness detection model. Images of healthy and diseased watermelons from the Mosaic Virus, Anthracnose, and Downy Mildew Disease are included in the dataset's five separate classifications. Images were painstakingly collected on June 25, 2023, in close cooperation with agricultural experts from the highly regarded Regional Horticulture Research Station in Lebukhali, Patuakhali.

PMID:38425879 | PMC:PMC10904151 | DOI:10.1016/j.dib.2024.110182

Data Brief. 2024 Feb 20;53:110218. doi: 10.1016/j.dib.2024.110218. eCollection 2024 Apr.

ABSTRACT

Concrete is a prominent construction material globally, owing to its reputed attributes such as robustness, endurance, optimal functionality, and adaptability. Formulating concrete mixtures poses a formidable challenge, mainly when introducing novel materials and additives and evaluating diverse design resistances. Recent methodologies for projecting concrete performance in fundamental aspects, including compressive strength, flexural strength, tensile strength, and durability (encompassing homogeneity, porosity, and internal structure), exist. However, actual approaches need more diversity in the materials and properties considered in their analyses. This dataset outlines the outcomes of an extensive 10-year laboratory investigation into concrete materials involving mechanical tests and non-destructive assessments within a comprehensive dataset denoted as ConcreteXAI. This dataset encompasses evaluations of mechanical performances and non-destructive tests. ConcreteXAI integrates a spectrum of analyzed mixtures comprising twelve distinct concrete formulations incorporating diverse additives and aggregate types. The dataset encompasses 18,480 data points, establishing itself as a cutting-edge resource for concrete analysis. ConcreteXAI acknowledges the influence of artificial intelligence techniques in various science fields. Emphatically, deep learning emerges as a precise methodology for analyzing and constructing predictive models. ConcreteXAI is designed to seamlessly integrate with deep learning models, enabling direct application of these models to predict or estimate desired attributes. Consequently, this dataset offers a resourceful avenue for researchers to develop high-quality prediction models for both mechanical and non-destructive tests on concrete elements, employing advanced deep learning techniques.

PMID:38425877 | PMC:PMC10904185 | DOI:10.1016/j.dib.2024.110218

Cancer Imaging. 2024 Feb 29;24(1):30. doi: 10.1186/s40644-024-00675-x.

ABSTRACT

BACKGROUND: Prostate-specific membrane antigen (PSMA) PET/CT imaging is widely used for quantitative image analysis, especially in radioligand therapy (RLT) for metastatic castration-resistant prostate cancer (mCRPC). Unknown features influencing PSMA biodistribution can be explored by analyzing segmented organs at risk (OAR) and lesions. Manual segmentation is time-consuming and labor-intensive, so automated segmentation methods are desirable. Training deep-learning segmentation models is challenging due to the scarcity of high-quality annotated images. Addressing this, we developed shifted windows UNEt TRansformers (Swin UNETR) for fully automated segmentation. Within a self-supervised framework, the model's encoder was pre-trained on unlabeled data. The entire model was fine-tuned, including its decoder, using labeled data.

METHODS: In this work, 752 whole-body [68Ga]Ga-PSMA-11 PET/CT images were collected from two centers. For self-supervised model pre-training, 652 unlabeled images were employed. The remaining 100 images were manually labeled for supervised training. In the supervised training phase, 5-fold cross-validation was used with 64 images for model training and 16 for validation, from one center. For testing, 20 hold-out images, evenly distributed between two centers, were used. Image segmentation and quantification metrics were evaluated on the test set compared to the ground-truth segmentation conducted by a nuclear medicine physician.

RESULTS: The model generates high-quality OARs and lesion segmentation in lesion-positive cases, including mCRPC. The results show that self-supervised pre-training significantly improved the average dice similarity coefficient (DSC) for all classes by about 3%. Compared to nnU-Net, a well-established model in medical image segmentation, our approach outperformed with a 5% higher DSC. This improvement was attributed to our model's combined use of self-supervised pre-training and supervised fine-tuning, specifically when applied to PET/CT input. Our best model had the lowest DSC for lesions at 0.68 and the highest for liver at 0.95.

CONCLUSIONS: We developed a state-of-the-art neural network using self-supervised pre-training on whole-body [68Ga]Ga-PSMA-11 PET/CT images, followed by fine-tuning on a limited set of annotated images. The model generates high-quality OARs and lesion segmentation for PSMA image analysis. The generalizable model holds potential for various clinical applications, including enhanced RLT and patient-specific internal dosimetry.

PMID:38424612 | DOI:10.1186/s40644-024-00675-x

Fluids Barriers CNS. 2024 Feb 29;21(1):21. doi: 10.1186/s12987-024-00525-9.

ABSTRACT

BACKGROUND: The choroid plexus functions as the blood-cerebrospinal fluid (CSF) barrier, plays an important role in CSF production and circulation, and has gained increased attention in light of the recent elucidation of CSF circulation dysfunction in neurodegenerative conditions. However, methods for routinely quantifying choroid plexus volume are suboptimal and require technical improvements and validation. Here, we propose three deep learning models that can segment the choroid plexus from commonly-acquired anatomical MRI data and report performance metrics and changes across the adult lifespan.

METHODS: Fully convolutional neural networks were trained from 3D T1-weighted, 3D T2-weighted, and 2D T2-weighted FLAIR MRI using gold-standard manual segmentations in control and neurodegenerative participants across the lifespan (n = 50; age = 21-85 years). Dice coefficients, 95% Hausdorff distances, and area-under-curve (AUCs) were calculated for each model and compared to segmentations from FreeSurfer using two-tailed Wilcoxon tests (significance criteria: p < 0.05 after false discovery rate multiple comparisons correction). Metrics were regressed against lateral ventricular volume using generalized linear models to assess model performance for varying levels of atrophy. Finally, models were applied to an expanded cohort of adult controls (n = 98; age = 21-89 years) to provide an exemplar of choroid plexus volumetry values across the lifespan.

RESULTS: Deep learning results yielded Dice coefficient = 0.72, Hausdorff distance = 1.97 mm, AUC = 0.87 for T1-weighted MRI, Dice coefficient = 0.72, Hausdorff distance = 2.22 mm, AUC = 0.87 for T2-weighted MRI, and Dice coefficient = 0.74, Hausdorff distance = 1.69 mm, AUC = 0.87 for T2-weighted FLAIR MRI; values did not differ significantly between MRI sequences and were statistically improved compared to current commercially-available algorithms (p < 0.001). The intraclass coefficients were 0.95, 0.95, and 0.96 between T1-weighted and T2-weighted FLAIR, T1-weighted and T2-weighted, and T2-weighted and T2-weighted FLAIR models, respectively. Mean lateral ventricle choroid plexus volume across all participants was 3.20 ± 1.4 cm3; a significant, positive relationship (R2 = 0.54-0.60) was observed between participant age and choroid plexus volume for all MRI sequences (p < 0.001).

CONCLUSIONS: Findings support comparable performance in choroid plexus delineation between standard, clinically available, non-contrasted anatomical MRI sequences. The software embedding the evaluated models is freely available online and should provide a useful tool for the growing number of studies that desire to quantitatively evaluate choroid plexus structure and function ( https://github.com/hettk/chp_seg ).

PMID:38424598 | DOI:10.1186/s12987-024-00525-9

BioData Min. 2024 Feb 29;17(1):8. doi: 10.1186/s13040-024-00359-z.

ABSTRACT

BACKGROUND: Breast cancer is the most common malignancy among women worldwide. Despite advances in treating breast cancer over the past decades, drug resistance and adverse effects remain challenging. Recent therapeutic progress has shifted toward using drug combinations for better treatment efficiency. However, with a growing number of potential small-molecule cancer inhibitors, in silico strategies to predict pharmacological synergy before experimental trials are required to compensate for time and cost restrictions. Many deep learning models have been previously proposed to predict the synergistic effects of drug combinations with high performance. However, these models heavily relied on a large number of drug chemical structural fingerprints as their main features, which made model interpretation a challenge.

RESULTS: This study developed a deep neural network model that predicts synergy between small-molecule pairs based on their inhibitory activities against 13 selected key proteins. The synergy prediction model achieved a Pearson correlation coefficient between model predictions and experimental data of 0.63 across five breast cancer cell lines. BT-549 and MCF-7 achieved the highest correlation of 0.67 when considering individual cell lines. Despite achieving a moderate correlation compared to previous deep learning models, our model offers a distinctive advantage in terms of interpretability. Using the inhibitory activities against key protein targets as the main features allowed a straightforward interpretation of the model since the individual features had direct biological meaning. By tracing the synergistic interactions of compounds through their target proteins, we gained insights into the patterns our model recognized as indicative of synergistic effects.

CONCLUSIONS: The framework employed in the present study lays the groundwork for future advancements, especially in model interpretation. By combining deep learning techniques and target-specific models, this study shed light on potential patterns of target-protein inhibition profiles that could be exploited in breast cancer treatment.

PMID:38424554 | DOI:10.1186/s13040-024-00359-z

Eur Arch Psychiatry Clin Neurosci. 2024 Feb 29. doi: 10.1007/s00406-024-01758-3. Online ahead of print.

ABSTRACT

As the brain ages, it almost invariably accumulates vascular pathology, which differentially affects the cerebral white matter. A rich body of research has investigated the link between vascular risk factors and the brain. One of the less studied questions is that among various modifiable vascular risk factors, which is the most debilitating one for white matter health? A white matter specific brain age was developed to evaluate the overall white matter health from diffusion weighted imaging, using a three-dimensional convolutional neural network deep learning model in both cross-sectional UK biobank participants (n = 37,327) and a longitudinal subset (n = 1409). White matter brain age gap (WMBAG) was the difference between the white matter age and the chronological age. Participants with one, two, and three or more vascular risk factors, compared to those without any, showed an elevated WMBAG of 0.54, 1.23, and 1.94 years, respectively. Diabetes was most strongly associated with an increased WMBAG (1.39 years, p < 0.001) among all risk factors followed by hypertension (0.87 years, p < 0.001) and smoking (0.69 years, p < 0.001). Baseline WMBAG was associated significantly with processing speed, executive and global cognition. Significant associations of diabetes and hypertension with poor processing speed and executive function were found to be mediated through the WMBAG. White matter specific brain age can be successfully targeted for the examination of the most relevant risk factors and cognition, and for tracking an individual's cerebrovascular ageing process. It also provides clinical basis for the better management of specific risk factors.

PMID:38424358 | DOI:10.1007/s00406-024-01758-3

J Imaging Inform Med. 2024 Feb 29. doi: 10.1007/s10278-024-01059-0. Online ahead of print.

ABSTRACT

This study aimed to develop and evaluate a CT-based deep learning radiomics model for differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB). A total of 330 patients with pathologically confirmed as CD or ITB from the First Affiliated Hospital of Zhengzhou University were divided into the validation dataset one (CD: 167; ITB: 57) and validation dataset two (CD: 78; ITB: 28). Based on the validation dataset one, the synthetic minority oversampling technique (SMOTE) was adopted to create balanced dataset as training data for feature selection and model construction. The handcrafted and deep learning (DL) radiomics features were extracted from the arterial and venous phases images, respectively. The interobserver consistency analysis, Spearman's correlation, univariate analysis, and the least absolute shrinkage and selection operator (LASSO) regression were used to select features. Based on extracted multi-phase radiomics features, six logistic regression models were finally constructed. The diagnostic performances of different models were compared using ROC analysis and Delong test. The arterial-venous combined deep learning radiomics model for differentiating between CD and ITB showed a high prediction quality with AUCs of 0.885, 0.877, and 0.800 in SMOTE dataset, validation dataset one, and validation dataset two, respectively. Moreover, the deep learning radiomics model outperformed the handcrafted radiomics model in same phase images. In validation dataset one, the Delong test results indicated that there was a significant difference in the AUC of the arterial models (p = 0.037), while not in venous and arterial-venous combined models (p = 0.398 and p = 0.265) as comparing deep learning radiomics models and handcrafted radiomics models. In our study, the arterial-venous combined model based on deep learning radiomics analysis exhibited good performance in differentiating between CD and ITB.

PMID:38424279 | DOI:10.1007/s10278-024-01059-0

J Imaging Inform Med. 2024 Feb 29. doi: 10.1007/s10278-023-00934-6. Online ahead of print.

ABSTRACT

Low-dose computer tomography (LDCT) has been widely used in medical diagnosis. Various denoising methods have been presented to remove noise in LDCT scans. However, existing methods cannot achieve satisfactory results due to the difficulties in (1) distinguishing the characteristics of structures, textures, and noise confused in the image domain, and (2) representing local details and global semantics in the hierarchical features. In this paper, we propose a novel denoising method consisting of (1) a 2D dual-domain restoration framework to reconstruct noise-free structure and texture signals separately, and (2) a 3D multi-depth reinforcement U-Net model to further recover image details with enhanced hierarchical features. In the 2D dual-domain restoration framework, the convolutional neural networks are adopted in both the image domain where the image structures are well preserved through the spatial continuity, and the sinogram domain where the textures and noise are separately represented by different wavelet coefficients and processed adaptively. In the 3D multi-depth reinforcement U-Net model, the hierarchical features from the 3D U-Net are enhanced by the cross-resolution attention module (CRAM) and dual-branch graph convolution module (DBGCM). The CRAM preserves local details by integrating adjacent low-level features with different resolutions, while the DBGCM enhances global semantics by building graphs for high-level features in intra-feature and inter-feature dimensions. Experimental results on the LUNA16 dataset and 2016 NIH-AAPM-Mayo Clinic LDCT Grand Challenge dataset illustrate the proposed method outperforms the state-of-the-art methods on removing noise from LDCT images with clear structures and textures, proving its potential in clinical practice.

PMID:38424278 | DOI:10.1007/s10278-023-00934-6

J Imaging Inform Med. 2024 Feb 29. doi: 10.1007/s10278-024-01042-9. Online ahead of print.

ABSTRACT

Automatic breast ultrasound image segmentation plays an important role in medical image processing. However, current methods for breast ultrasound segmentation suffer from high computational complexity and large model parameters, particularly when dealing with complex images. In this paper, we take the Unext network as a basis and utilize its encoder-decoder features. And taking inspiration from the mechanisms of cellular apoptosis and division, we design apoptosis and division algorithms to improve model performance. We propose a novel segmentation model which integrates the division and apoptosis algorithms and introduces spatial and channel convolution blocks into the model. Our proposed model not only improves the segmentation performance of breast ultrasound tumors, but also reduces the model parameters and computational resource consumption time. The model was evaluated on the breast ultrasound image dataset and our collected dataset. The experiments show that the SC-Unext model achieved Dice scores of 75.29% and accuracy of 97.09% on the BUSI dataset, and on the collected dataset, it reached Dice scores of 90.62% and accuracy of 98.37%. Meanwhile, we conducted a comparison of the model's inference speed on CPUs to verify its efficiency in resource-constrained environments. The results indicated that the SC-Unext model achieved an inference speed of 92.72 ms per instance on devices equipped only with CPUs. The model's number of parameters and computational resource consumption are 1.46M and 2.13 GFlops, respectively, which are lower compared to other network models. Due to its lightweight nature, the model holds significant value for various practical applications in the medical field.

PMID:38424276 | DOI:10.1007/s10278-024-01042-9

Sci Rep. 2024 Feb 29;14(1):4981. doi: 10.1038/s41598-024-55677-3.

ABSTRACT

Developing a deep-learning-based diagnostic model demands extensive labor for medical image labeling. Attempts to reduce the labor often lead to incomplete or inaccurate labeling, limiting the diagnostic performance of models. This paper (i) constructs an attention-guiding framework that enhances the diagnostic performance of jaw bone pathology by utilizing attention information with partially labeled data; (ii) introduces an additional loss to minimize the discrepancy between network attention and its label; (iii) introduces a trapezoid augmentation method to maximize the utility of minimally labeled data. The dataset includes 716 panoramic radiograph data for jaw bone lesions and normal cases collected and labeled by two radiologists from January 2019 to February 2021. Experiments show that guiding network attention with even 5% of attention-labeled data can enhance the diagnostic accuracy for pathology from 92.41 to 96.57%. Furthermore, ablation studies reveal that the proposed augmentation methods outperform prior preprocessing and augmentation combinations, achieving an accuracy of 99.17%. The results affirm the capability of the proposed framework in fine-grained diagnosis using minimally labeled data, offering a practical solution to the challenges of medical image analysis.

PMID:38424124 | DOI:10.1038/s41598-024-55677-3