Prehosp Disaster Med. 2024 May 17:1-11. doi: 10.1017/S1049023X24000414. Online ahead of print.

ABSTRACT

OBJECTIVE: The aim of this study was to summarize the literature on the applications of machine learning (ML) and their performance in Emergency Medical Services (EMS).

METHODS: Four relevant electronic databases were searched (from inception through January 2024) for all original studies that employed EMS-guided ML algorithms to enhance the clinical and operational performance of EMS. Two reviewers screened the retrieved studies and extracted relevant data from the included studies. The characteristics of included studies, employed ML algorithms, and their performance were quantitively described across primary domains and subdomains.

RESULTS: This review included a total of 164 studies published from 2005 through 2024. Of those, 125 were clinical domain focused and 39 were operational. The characteristics of ML algorithms such as sample size, number and type of input features, and performance varied between and within domains and subdomains of applications. Clinical applications of ML algorithms involved triage or diagnosis classification (n = 62), treatment prediction (n = 12), or clinical outcome prediction (n = 50), mainly for out-of-hospital cardiac arrest/OHCA (n = 62), cardiovascular diseases/CVDs (n = 19), and trauma (n = 24). The performance of these ML algorithms varied, with a median area under the receiver operating characteristic curve (AUC) of 85.6%, accuracy of 88.1%, sensitivity of 86.05%, and specificity of 86.5%. Within the operational studies, the operational task of most ML algorithms was ambulance allocation (n = 21), followed by ambulance detection (n = 5), ambulance deployment (n = 5), route optimization (n = 5), and quality assurance (n = 3). The performance of all operational ML algorithms varied and had a median AUC of 96.1%, accuracy of 90.0%, sensitivity of 94.4%, and specificity of 87.7%. Generally, neural network and ensemble algorithms, to some degree, out-performed other ML algorithms.

CONCLUSION: Triaging and managing different prehospital medical conditions and augmenting ambulance performance can be improved by ML algorithms. Future reports should focus on a specific clinical condition or operational task to improve the precision of the performance metrics of ML models.

PMID:38757150 | DOI:10.1017/S1049023X24000414

Bioinform Biol Insights. 2024 May 15;18:11779322241231108. doi: 10.1177/11779322241231108. eCollection 2024.

ABSTRACT

Gait analysis serves as a critical diagnostic tool for identifying neurologic and musculoskeletal damage. Traditional manual analysis of motion data, however, is labor-intensive and heavily reliant on the expertise and judgment of the therapist. This study introduces a binary classification method for the quantitative assessment of gait impairments, specifically focusing on Duchenne muscular dystrophy (DMD), a prevalent and fatal neuromuscular genetic disorder. The research compares spatiotemporal and sagittal kinematic gait features derived from 2D and 3D human pose estimation trajectories against concurrently recorded 3D motion capture (MoCap) data from healthy children. The proposed model leverages a novel benchmark dataset, collected from YouTube and publicly available datasets of their typically developed peers, to extract time-distance variables (e.g. speed, step length, stride time, and cadence) and sagittal joint angles of the lower extremity (e.g. hip, knee, and knee flexion angles). Machine learning and deep learning techniques are employed to discern patterns that can identify children exhibiting DMD gait disturbances. While the current model is capable of distinguishing between healthy subjects and those with DMD, it does not specifically differentiate between DMD patients and patients with other gait impairments. Experimental results validate the efficacy of our cost-effective method, which relies on recorded RGB video, in detecting gait abnormalities, achieving a prediction accuracy of 96.2% for Support Vector Machine (SVM) and 97% for the deep network.

PMID:38757143 | PMC:PMC11097739 | DOI:10.1177/11779322241231108

Cyborg Bionic Syst. 2024 May 16;5:0100. doi: 10.34133/cbsystems.0100. eCollection 2024.

ABSTRACT

Three-dimensional skeleton-based action recognition (3D SAR) has gained important attention within the computer vision community, owing to the inherent advantages offered by skeleton data. As a result, a plethora of impressive works, including those based on conventional handcrafted features and learned feature extraction methods, have been conducted over the years. However, prior surveys on action recognition have primarily focused on video or red-green-blue (RGB) data-dominated approaches, with limited coverage of reviews related to skeleton data. Furthermore, despite the extensive application of deep learning methods in this field, there has been a notable absence of research that provides an introductory or comprehensive review from the perspective of deep learning architectures. To address these limitations, this survey first underscores the importance of action recognition and emphasizes the significance of 3-dimensional (3D) skeleton data as a valuable modality. Subsequently, we provide a comprehensive introduction to mainstream action recognition techniques based on 4 fundamental deep architectures, i.e., recurrent neural networks, convolutional neural networks, graph convolutional network, and Transformers. All methods with the corresponding architectures are then presented in a data-driven manner with detailed discussion. Finally, we offer insights into the current largest 3D skeleton dataset, NTU-RGB+D, and its new edition, NTU-RGB+D 120, along with an overview of several top-performing algorithms on these datasets. To the best of our knowledge, this research represents the first comprehensive discussion of deep learning-based action recognition using 3D skeleton data.

PMID:38757045 | PMC:PMC11096730 | DOI:10.34133/cbsystems.0100

Front Med (Lausanne). 2024 May 2;11:1360143. doi: 10.3389/fmed.2024.1360143. eCollection 2024.

ABSTRACT

INTRODUCTION: Deep learning-based methods can promote and save critical time for the diagnosis of pneumonia from computed tomography (CT) images of the chest, where the methods usually rely on large amounts of labeled data to learn good visual representations. However, medical images are difficult to obtain and need to be labeled by professional radiologists.

METHODS: To address this issue, a novel contrastive learning model with token projection, namely CoTP, is proposed for improving the diagnostic quality of few-shot chest CT images. Specifically, (1) we utilize solely unlabeled data for fitting CoTP, along with a small number of labeled samples for fine-tuning, (2) we present a new Omicron dataset and modify the data augmentation strategy, i.e., random Poisson noise perturbation for the CT interpretation task, and (3) token projection is utilized to further improve the quality of the global visual representations.

RESULTS: The ResNet50 pre-trained by CoTP attained accuracy (ACC) of 92.35%, sensitivity (SEN) of 92.96%, precision (PRE) of 91.54%, and the area under the receiver-operating characteristics curve (AUC) of 98.90% on the presented Omicron dataset. On the contrary, the ResNet50 without pre-training achieved ACC, SEN, PRE, and AUC of 77.61, 77.90, 76.69, and 85.66%, respectively.

CONCLUSION: Extensive experiments reveal that a model pre-trained by CoTP greatly outperforms that without pre-training. The CoTP can improve the efficacy of diagnosis and reduce the heavy workload of radiologists for screening of Omicron pneumonia.

PMID:38756944 | PMC:PMC11096503 | DOI:10.3389/fmed.2024.1360143

Front Artif Intell. 2024 May 2;7:1376042. doi: 10.3389/frai.2024.1376042. eCollection 2024.

ABSTRACT

Wetware computing and organoid intelligence is an emerging research field at the intersection of electrophysiology and artificial intelligence. The core concept involves using living neurons to perform computations, similar to how Artificial Neural Networks (ANNs) are used today. However, unlike ANNs, where updating digital tensors (weights) can instantly modify network responses, entirely new methods must be developed for neural networks using biological neurons. Discovering these methods is challenging and requires a system capable of conducting numerous experiments, ideally accessible to researchers worldwide. For this reason, we developed a hardware and software system that allows for electrophysiological experiments on an unmatched scale. The Neuroplatform enables researchers to run experiments on neural organoids with a lifetime of even more than 100 days. To do so, we streamlined the experimental process to quickly produce new organoids, monitor action potentials 24/7, and provide electrical stimulations. We also designed a microfluidic system that allows for fully automated medium flow and change, thus reducing the disruptions by physical interventions in the incubator and ensuring stable environmental conditions. Over the past three years, the Neuroplatform was utilized with over 1,000 brain organoids, enabling the collection of more than 18 terabytes of data. A dedicated Application Programming Interface (API) has been developed to conduct remote research directly via our Python library or using interactive compute such as Jupyter Notebooks. In addition to electrophysiological operations, our API also controls pumps, digital cameras and UV lights for molecule uncaging. This allows for the execution of complex 24/7 experiments, including closed-loop strategies and processing using the latest deep learning or reinforcement learning libraries. Furthermore, the infrastructure supports entirely remote use. Currently in 2024, the system is freely available for research purposes, and numerous research groups have begun using it for their experiments. This article outlines the system's architecture and provides specific examples of experiments and results.

PMID:38756757 | PMC:PMC11097343 | DOI:10.3389/frai.2024.1376042

Front Oncol. 2024 May 2;14:1359148. doi: 10.3389/fonc.2024.1359148. eCollection 2024.

ABSTRACT

OBJECTIVE: Neoadjuvant chemotherapy (NAC) is a key element of treatment for locally advanced breast cancer (LABC). Predicting the response to NAC for patients with Locally Advanced Breast Cancer (LABC) before treatment initiation could be beneficial to optimize therapy, ensuring the administration of effective treatments. The objective of the work here was to develop a predictive model to predict tumor response to NAC for LABC using deep learning networks and computed tomography (CT).

MATERIALS AND METHODS: Several deep learning approaches were investigated including ViT transformer and VGG16, VGG19, ResNet-50, Res-Net-101, Res-Net-152, InceptionV3 and Xception transfer learning networks. These deep learning networks were applied on CT images to assess the response to NAC. Performance was evaluated based on balanced_accuracy, accuracy, sensitivity and specificity classification metrics. A ViT transformer was applied to utilize the attention mechanism in order to increase the weight of important part image which leads to better discrimination between classes.

RESULTS: Amongst the 117 LABC patients studied, 82 (70%) had clinical-pathological response and 35 (30%) had no response to NAC. The ViT transformer obtained the best performance range (accuracy = 71 ± 3% to accuracy = 77 ± 4%, specificity = 86 ± 6% to specificity = 76 ± 3%, sensitivity = 56 ± 4% to sensitivity = 52 ± 4%, and balanced_accuracy=69 ± 3% to balanced_accuracy=69 ± 3%) depending on the split ratio of train-data and test-data. Xception network obtained the second best results (accuracy = 72 ± 4% to accuracy = 65 ± 4, specificity = 81 ± 6% to specificity = 73 ± 3%, sensitivity = 55 ± 4% to sensitivity = 52 ± 5%, and balanced_accuracy = 66 ± 5% to balanced_accuracy = 60 ± 4%). The worst results were obtained using VGG-16 transfer learning network.

CONCLUSION: Deep learning networks in conjunction with CT imaging are able to predict the tumor response to NAC for patients with LABC prior to start. A ViT transformer could obtain the best performance, which demonstrated the importance of attention mechanism.

PMID:38756659 | PMC:PMC11096486 | DOI:10.3389/fonc.2024.1359148

J Gastrointest Oncol. 2024 Apr 30;15(2):535-543. doi: 10.21037/jgo-24-158. Epub 2024 Apr 29.

ABSTRACT

BACKGROUND: There have been studies on the application of computer-aided diagnosis (CAD) in the endoscopic diagnosis of early esophageal cancer (EEC), but there is still a significant gap from clinical application. We developed an endoscopic CAD system for EEC based on the AutoGluon framework, aiming to explore the feasibility of automatic deep learning (DL) in clinical application.

METHODS: The endoscopic pictures of normal esophagus, esophagitis, and EEC were collected from The First Affiliated Hospital of Soochow University (September 2015 to December 2021) and the Norwegian HyperKvasir database. All images of non-cancerous esophageal lesions and EEC in this study were pathologically examined. There were three tasks: task A was normal vs. lesion classification under non-magnifying endoscopy (n=932 vs. 1,092); task B was non-cancer lesion vs. EEC classification under non-magnifying endoscopy (n=594 vs. 429); and task C was non-cancer lesion vs. EEC classification under magnifying endoscopy (n=505 vs. 824). In all classification tasks, we took 100 pictures as the verification set, and the rest comprised as the training set. The CAD system was established based on the AutoGluon framework. Diagnostic performance of the model was compared with that of endoscopists grouped according to years of experience (senior >15 years; junior <5 years). Model evaluation indicators included accuracy, recall rate, precision, F1 value, interpretation time, and the area under the receiver operating characteristic (ROC) curve (AUC).

RESULTS: In tasks A and B, the accuracies of medium-performance CAD and high-performance CAD were lower than those of junior doctors and senior doctors. In task C, the medium-performance and high-performance CAD accuracies were close to those of junior doctors and senior doctors. The high-performance CAD model outperformed the junior doctors in both task A (0.850 vs. 0.830) and task C (0.840 vs. 0.830) in sensitivity comparison, but there was still a large gap between high-performance CAD models and doctors in sensitivity comparison. In task A, with the aid of CAD pre-interpretation, the accuracy of junior and senior physicians were significantly improved (from 0.880 to 0.915 and from 0.920 to 0.945, respectively); the time spent on film reading was significantly shortened (junior: from 11.3 to 8.7 s; senior: from 6.7 to 5.5 s). In task C, with the aid of CAD pre-interpretation, the accuracy of junior and senior physicians were significantly improved (from 0.850 to 0.865 and from 0.915 to 0.935, respectively); the reading time was significantly shortened (junior: from 9.5 to 7.7 s; senior: from 5.6 to 3.0 s).

CONCLUSIONS: The CAD system based on the AutoGluon framework can assist doctors to improve the diagnostic accuracy and reading time of EEC under endoscopy. This study reveals that automatic DL methods are promising in clinical application.

PMID:38756633 | PMC:PMC11094492 | DOI:10.21037/jgo-24-158

Front Big Data. 2024 May 2;7:1366415. doi: 10.3389/fdata.2024.1366415. eCollection 2024.

ABSTRACT

Chest X-ray (CXR) imaging is widely employed by radiologists to diagnose thoracic diseases. Recently, many deep learning techniques have been proposed as computer-aided diagnostic (CAD) tools to assist radiologists in minimizing the risk of incorrect diagnosis. From an application perspective, these models have exhibited two major challenges: (1) They require large volumes of annotated data at the training stage and (2) They lack explainable factors to justify their outcomes at the prediction stage. In the present study, we developed a class activation mapping (CAM)-based ensemble model, called Ensemble-CAM, to address both of these challenges via weakly supervised learning by employing explainable AI (XAI) functions. Ensemble-CAM utilizes class labels to predict the location of disease in association with interpretable features. The proposed work leverages ensemble and transfer learning with class activation functions to achieve three objectives: (1) minimizing the dependency on strongly annotated data when locating thoracic diseases, (2) enhancing confidence in predicted outcomes by visualizing their interpretable features, and (3) optimizing cumulative performance via fusion functions. Ensemble-CAM was trained on three CXR image datasets and evaluated through qualitative and quantitative measures via heatmaps and Jaccard indices. The results reflect the enhanced performance and reliability in comparison to existing standalone and ensembled models.

PMID:38756502 | PMC:PMC11096460 | DOI:10.3389/fdata.2024.1366415

Proc SPIE Int Soc Opt Eng. 2024 Feb;12933:129330A. doi: 10.1117/12.3006418. Epub 2024 Apr 3.

ABSTRACT

Current deep learning methods in histopathology are limited by the small amount of available data and time consumption in labeling the data. Colorectal cancer (CRC) tumor budding quantification performed using H&E-stained slides is crucial for cancer staging and prognosis but is subject to labor-intensive annotation and human bias. Thus, acquiring a large-scale, fully annotated dataset for training a tumor budding (TB) segmentation/detection system is difficult. Here, we present a DatasetGAN-based approach that can generate essentially an unlimited number of images with TB masks from a moderate number of unlabeled images and a few annotated images. The images generated by our model closely resemble the real colon tissue on H&E-stained slides. We test the performance of this model by training a downstream segmentation model, UNet++, on the generated images and masks. Our results show that the trained UNet++ model can achieve reasonable TB segmentation performance, especially at the instance level. This study demonstrates the potential of developing an annotation-efficient segmentation model for automatic TB detection and quantification.

PMID:38756441 | PMC:PMC11098595 | DOI:10.1117/12.3006418

J Med Imaging (Bellingham). 2024 May;11(3):034001. doi: 10.1117/1.JMI.11.3.034001. Epub 2024 May 15.

ABSTRACT

PURPOSE: Automatic comprehensive reporting of coronary artery disease (CAD) requires anatomical localization of the coronary artery pathologies. To address this, we propose a fully automatic method for extraction and anatomical labeling of the coronary artery tree using deep learning.

APPROACH: We include coronary CT angiography (CCTA) scans of 104 patients from two hospitals. Reference annotations of coronary artery tree centerlines and labels of coronary artery segments were assigned to 10 segment classes following the American Heart Association guidelines. Our automatic method first extracts the coronary artery tree from CCTA, automatically placing a large number of seed points and simultaneous tracking of vessel-like structures from these points. Thereafter, the extracted tree is refined to retain coronary arteries only, which are subsequently labeled with a multi-resolution ensemble of graph convolutional neural networks that combine geometrical and image intensity information from adjacent segments.

RESULTS: The method is evaluated on its ability to extract the coronary tree and to label its segments, by comparing the automatically derived and the reference labels. A separate assessment of tree extraction yielded an F1 score of 0.85. Evaluation of our combined method leads to an average F1 score of 0.74.

CONCLUSIONS: The results demonstrate that our method enables fully automatic extraction and anatomical labeling of coronary artery trees from CCTA scans. Therefore, it has the potential to facilitate detailed automatic reporting of CAD.

PMID:38756439 | PMC:PMC11095121 | DOI:10.1117/1.JMI.11.3.034001

BMC Med Imaging. 2024 May 16;24(1):112. doi: 10.1186/s12880-024-01296-3.

ABSTRACT

Accurate preoperative differentiation of the chronic rhinosinusitis (CRS) endotype between eosinophilic CRS (eCRS) and non-eosinophilic CRS (non-eCRS) is an important topic in predicting postoperative outcomes and administering personalized treatment. To this end, we have constructed a sinus CT dataset, which comprises CT scan data and pathological biopsy results from 192 patients of chronic rhinosinusitis with nasal polyps (CRSwNP), treated at the Second Affiliated Hospital of Shantou University Medical College between 2020 and 2022. To differentiate CRSwNP endotype on preoperative CT and improve efficiency at the same time, we developed a multi-view fusion model that contains a mini-architecture with each network of 10 layers by modifying the deep residual neural network. The proposed model is trained on a training set and evaluated on a test set. The multi-view deep learning fusion model achieved the area under the receiver-operating characteristics curve (AUC) of 0.991, accuracy of 0.965 and F1-Score of 0.970 in test set. We compared the performance of the mini-architecture with other lightweight networks on the same Sinus CT dataset. The experimental results demonstrate that the developed ResMini architecture contribute to competitive CRSwNP endotype identification modeling in terms of accuracy and parameter number.

PMID:38755567 | DOI:10.1186/s12880-024-01296-3

BMC Med Inform Decis Mak. 2024 May 16;24(1):126. doi: 10.1186/s12911-024-02529-9.

ABSTRACT

BACKGROUND: Chest X-ray imaging based abnormality localization, essential in diagnosing various diseases, faces significant clinical challenges due to complex interpretations and the growing workload of radiologists. While recent advances in deep learning offer promising solutions, there is still a critical issue of domain inconsistency in cross-domain transfer learning, which hampers the efficiency and accuracy of diagnostic processes. This study aims to address the domain inconsistency problem and improve autonomic abnormality localization performance of heterogeneous chest X-ray image analysis, particularly in detecting abnormalities, by developing a self-supervised learning strategy called "BarlwoTwins-CXR".

METHODS: We utilized two publicly available datasets: the NIH Chest X-ray Dataset and the VinDr-CXR. The BarlowTwins-CXR approach was conducted in a two-stage training process. Initially, self-supervised pre-training was performed using an adjusted Barlow Twins algorithm on the NIH dataset with a Resnet50 backbone pre-trained on ImageNet. This was followed by supervised fine-tuning on the VinDr-CXR dataset using Faster R-CNN with Feature Pyramid Network (FPN). The study employed mean Average Precision (mAP) at an Intersection over Union (IoU) of 50% and Area Under the Curve (AUC) for performance evaluation.

RESULTS: Our experiments showed a significant improvement in model performance with BarlowTwins-CXR. The approach achieved a 3% increase in mAP50 accuracy compared to traditional ImageNet pre-trained models. In addition, the Ablation CAM method revealed enhanced precision in localizing chest abnormalities. The study involved 112,120 images from the NIH dataset and 18,000 images from the VinDr-CXR dataset, indicating robust training and testing samples.

CONCLUSION: BarlowTwins-CXR significantly enhances the efficiency and accuracy of chest X-ray image-based abnormality localization, outperforming traditional transfer learning methods and effectively overcoming domain inconsistency in cross-domain scenarios. Our experiment results demonstrate the potential of using self-supervised learning to improve the generalizability of models in medical settings with limited amounts of heterogeneous data. This approach can be instrumental in aiding radiologists, particularly in high-workload environments, offering a promising direction for future AI-driven healthcare solutions.

PMID:38755563 | DOI:10.1186/s12911-024-02529-9

Sci Rep. 2024 May 16;14(1):11185. doi: 10.1038/s41598-024-61915-5.

ABSTRACT

The brain presents age-related structural and functional changes in the human life, with different extends between subjects and groups. Brain age prediction can be used to evaluate the development and aging of human brain, as well as providing valuable information for neurodevelopment and disease diagnosis. Many contributions have been made for this purpose, resorting to different machine learning methods. To solve this task and reduce memory resource consumption, we develop a mini architecture of only 10 layers by modifying the deep residual neural network (ResNet), named ResNet mini architecture. To support the ResNet mini architecture in brain age prediction, the brain age dataset (OpenNeuro #ds000228) that consists of 155 study participants (three classes) and the Alzheimer MRI preprocessed dataset that consists of 6400 images (four classes) are employed. We compared the performance of the ResNet mini architecture with other popular networks using the two considered datasets. Experimental results show that the proposed architecture exhibits generality and robustness with high accuracy and less parameter number.

PMID:38755275 | DOI:10.1038/s41598-024-61915-5

Sci Rep. 2024 May 16;14(1):11233. doi: 10.1038/s41598-024-62278-7.

ABSTRACT

Automated disease diagnosis and prediction, powered by AI, play a crucial role in enabling medical professionals to deliver effective care to patients. While such predictive tools have been extensively explored in resource-rich languages like English, this manuscript focuses on predicting disease categories automatically from symptoms documented in the Afaan Oromo language, employing various classification algorithms. This study encompasses machine learning techniques such as support vector machines, random forests, logistic regression, and Naïve Bayes, as well as deep learning approaches including LSTM, GRU, and Bi-LSTM. Due to the unavailability of a standard corpus, we prepared three data sets with different numbers of patient symptoms arranged into 10 categories. The two feature representations, TF-IDF and word embedding, were employed. The performance of the proposed methodology has been evaluated using accuracy, recall, precision, and F1 score. The experimental results show that, among machine learning models, the SVM model using TF-IDF had the highest accuracy and F1 score of 94.7%, while the LSTM model using word2vec embedding showed an accuracy rate of 95.7% and F1 score of 96.0% from deep learning models. To enhance the optimal performance of each model, several hyper-parameter tuning settings were used. This study shows that the LSTM model verifies to be the best of all the other models over the entire dataset.

PMID:38755269 | DOI:10.1038/s41598-024-62278-7

Healthc Inform Res. 2024 Apr;30(2):140-146. doi: 10.4258/hir.2024.30.2.140. Epub 2024 Apr 30.

ABSTRACT

OBJECTIVES: Skin cancer is a prevalent type of malignancy, necessitating efficient diagnostic tools. This study aimed to develop an automated skin lesion classification model using the dynamically expandable representation (DER) incremental learning algorithm. This algorithm adapts to new data and expands its classification capabilities, with the goal of creating a scalable and efficient system for diagnosing skin cancer.

METHODS: The DER model with incremental learning was applied to the HAM10000 and ISIC 2019 datasets. Validation involved two steps: initially, training and evaluating the HAM10000 dataset against a fixed ResNet-50; subsequently, performing external validation of the trained model using the ISIC 2019 dataset. The model's performance was assessed using precision, recall, the F1-score, and area under the precision-recall curve.

RESULTS: The developed skin lesion classification model demonstrated high accuracy and reliability across various types of skin lesions, achieving a weighted-average precision, recall, and F1-score of 0.918, 0.808, and 0.847, respectively. The model's discrimination performance was reflected in an average area under the curve (AUC) value of 0.943. Further external validation with the ISIC 2019 dataset confirmed the model's effectiveness, as shown by an AUC of 0.911.

CONCLUSIONS: This study presents an optimized skin lesion classification model based on the DER algorithm, which shows high performance in disease classification with the potential to expand its classification range. The model demonstrated robust results in external validation, indicating its adaptability to new disease classes.

PMID:38755104 | DOI:10.4258/hir.2024.30.2.140

Schizophr Bull. 2024 May 16:sbae069. doi: 10.1093/schbul/sbae069. Online ahead of print.

ABSTRACT

BACKGROUND AND HYPOTHESIS: Schizophrenia (SZ) is a prevalent mental disorder that imposes significant health burdens. Diagnostic accuracy remains challenging due to clinical subjectivity. To address this issue, we explore magnetic resonance imaging (MRI) as a tool to enhance SZ diagnosis and provide objective references and biomarkers. Using deep learning with graph convolution, we represent MRI data as graphs, aligning with brain structure, and improving feature extraction, and classification. Integration of multiple modalities is expected to enhance classification.

STUDY DESIGN: Our study enrolled 683 SZ patients and 606 healthy controls from 7 hospitals, collecting structural MRI and functional MRI data. Both data types were represented as graphs, processed by 2 graph attention networks, and fused for classification. Grad-CAM with graph convolution ensured interpretability, and partial least squares analyzed gene expression in brain regions.

STUDY RESULTS: Our method excelled in the classification task, achieving 83.32% accuracy, 83.41% sensitivity, and 83.20% specificity in 10-fold cross-validation, surpassing traditional methods. And our multimodal approach outperformed unimodal methods. Grad-CAM identified potential brain biomarkers consistent with gene analysis and prior research.

CONCLUSIONS: Our study demonstrates the effectiveness of deep learning with graph attention networks, surpassing previous SZ diagnostic methods. Multimodal MRI's superiority over unimodal MRI confirms our initial hypothesis. Identifying potential brain biomarkers alongside gene biomarkers holds promise for advancing objective SZ diagnosis and research in SZ.

PMID:38754993 | DOI:10.1093/schbul/sbae069

J Mol Biol. 2024 May 14:168610. doi: 10.1016/j.jmb.2024.168610. Online ahead of print.

ABSTRACT

The executors of organismal functions are proteins, and the transition from RNA to protein is subject to post-transcriptional regulation; therefore, considering both RNA and surface protein expression simultaneously can provide additional evidence of biological processes. Cellular indexing of transcriptomes and epitopes by sequencing (CITE-Seq) technology can measure both RNA and protein expression in single cells, but these experiments are expensive and time-consuming. Due to the lack of computational tools for predicting surface proteins, we used datasets obtained with CITE-seq technology to design a deep generative prediction method based on diffusion models and to find biological discoveries through the prediction results. In our method, the scDM, which predicts protein expression values from RNA expression values of individual cells, uses a novel way of encoding the data into a model and generates predicted samples by introducing Gaussian noise to gradually remove the noise to learn the data distribution during the modelling process. Comprehensive evaluation across different datasets demonstrated that our predictions yielded satisfactory results and further demonstrated the effectiveness of incorporating information from single-cell multiomics data into diffusion models for biological studies. We also found that new directions for discovering therapeutic drug targets could be provided by jointly analysing the predictive value of surface protein expression and cancer cell drug scores.

PMID:38754773 | DOI:10.1016/j.jmb.2024.168610

Magn Reson Imaging. 2024 May 14:S0730-725X(24)00158-9. doi: 10.1016/j.mri.2024.05.009. Online ahead of print.

ABSTRACT

Accurately studying structural connectivity requires precise tract segmentation strategies. The U-Net network has been widely recognized for its exceptional capacity in image segmentation tasks and provides remarkable results in large tract segmentation when high-quality diffusion-weighted imaging (DWI) data are used. However, short tracts, which are associated with various neurological diseases, pose specific challenges, particularly when high-quality DWI data acquisition within clinical settings is concerned. Here, we aimed to evaluate the U-Net network ability to segment short tracts by using DWI data acquired in different experimental conditions. To this end, we conducted three types of training experiments involving 350 healthy subjects and 11 white matter tracts, including the anterior, posterior, and hippocampal commissure, fornix, and uncinated fasciculus. In the first experiment, the model was exclusively trained with high-quality data of the Human Connectome Project (HCP) dataset. The second experiment focused on images of healthy subjects acquired from a local hospital dataset, representing a typical clinical routine acquisition. In the third experiment, a hybrid training approach was employed, combining data of the HCP and local hospital datasets. Then, the best model was also tested in unseen DWIs of 10 epilepsy patients of the local hospital and 10 healthy subjects acquired on a scanner from another company. The outcomes of the third experiment demonstrated a notable enhancement in performance when contrasted with the preceding trials. Specifically, the short tracts within the local hospital dataset achieved Dice scores ranging between 0.60 and 0.65. Similar intervals were obtained with HCP data in the first experiment, and a substantial improvement compared to the scores between 0.37 and 0.50 obtained with the local hospital dataset at the same experiment. This improvement persisted when the method was applied to diverse scenarios, including different scanner acquisitions and epilepsy patients. These results indicate that combining datasets from different sources, coupled with resolution standardization strengthens the neural network ability to generalize predictions across a spectrum of datasets. Nevertheless, short tract segmentation performance is intricately linked to the training composition, to validation, and to testing data. Moreover, curved tracts have intricate structural nature, which adds complexities to their segmenting. Although the network training approach tested herein has provided promising results, caution must be taken when extrapolating its application to datasets acquired under distinct experimental conditions, even in the case of higher-quality data or analysis of long or short tracts.

PMID:38754751 | DOI:10.1016/j.mri.2024.05.009

J Microbiol Methods. 2024 May 14:106955. doi: 10.1016/j.mimet.2024.106955. Online ahead of print.

ABSTRACT

We aim to objectify the evaluation criteria of agglutination rate estimation in the Microscopic Agglutination Test (MAT). This study proposes a deep learning method that extracts free leptospires from dark-field microscopic images and calculates the agglutination rate. The experiments show the effect of objectification with real pictures.

PMID:38754481 | DOI:10.1016/j.mimet.2024.106955

Brief Bioinform. 2024 Mar 27;25(3):bbae227. doi: 10.1093/bib/bbae227.

ABSTRACT

Predicting cancer drug response using both genomics and drug features has shown some success compared to using genomics features alone. However, there has been limited research done on how best to combine or fuse the two types of features. Using a visible neural network with two deep learning branches for genes and drug features as the base architecture, we experimented with different fusion functions and fusion points. Our experiments show that injecting multiplicative relationships between gene and drug latent features into the original concatenation-based architecture DrugCell significantly improved the overall predictive performance and outperformed other baseline models. We also show that different fusion methods respond differently to different fusion points, indicating that the relationship between drug features and different hierarchical biological level of gene features is optimally captured using different methods. Considering both predictive performance and runtime speed, tensor product partial is the best-performing fusion function to combine late-stage representations of drug and gene features to predict cancer drug response.

PMID:38754407 | DOI:10.1093/bib/bbae227