J Med Internet Res. 2024 Jul 3;26:e52139. doi: 10.2196/52139.

ABSTRACT

BACKGROUND: Although several biomarkers exist for patients with heart failure (HF), their use in routine clinical practice is often constrained by high costs and limited availability.

OBJECTIVE: We examined the utility of an artificial intelligence (AI) algorithm that analyzes printed electrocardiograms (ECGs) for outcome prediction in patients with acute HF.

METHODS: We retrospectively analyzed prospectively collected data of patients with acute HF at two tertiary centers in Korea. Baseline ECGs were analyzed using a deep-learning system called Quantitative ECG (QCG), which was trained to detect several urgent clinical conditions, including shock, cardiac arrest, and reduced left ventricular ejection fraction (LVEF).

RESULTS: Among the 1254 patients enrolled, in-hospital cardiac death occurred in 53 (4.2%) patients, and the QCG score for critical events (QCG-Critical) was significantly higher in these patients than in survivors (mean 0.57, SD 0.23 vs mean 0.29, SD 0.20; P<.001). The QCG-Critical score was an independent predictor of in-hospital cardiac death after adjustment for age, sex, comorbidities, HF etiology/type, atrial fibrillation, and QRS widening (adjusted odds ratio [OR] 1.68, 95% CI 1.47-1.92 per 0.1 increase; P<.001), and remained a significant predictor after additional adjustments for echocardiographic LVEF and N-terminal prohormone of brain natriuretic peptide level (adjusted OR 1.59, 95% CI 1.36-1.87 per 0.1 increase; P<.001). During long-term follow-up, patients with higher QCG-Critical scores (>0.5) had higher mortality rates than those with low QCG-Critical scores (<0.25) (adjusted hazard ratio 2.69, 95% CI 2.14-3.38; P<.001).

CONCLUSIONS: Predicting outcomes in patients with acute HF using the QCG-Critical score is feasible, indicating that this AI-based ECG score may be a novel biomarker for these patients.

TRIAL REGISTRATION: ClinicalTrials.gov NCT01389843; https://clinicaltrials.gov/study/NCT01389843.

PMID:38959500 | DOI:10.2196/52139

Dis Colon Rectum. 2024 Jul 3. doi: 10.1097/DCR.0000000000003335. Online ahead of print.

ABSTRACT

BACKGROUND: Iatrogenic ureteral injury is a serious complication of abdominopelvic surgery. Identifying the ureters intraoperatively is essential to avoid iatrogenic ureteral injury. Here, we developed a model that may minimize this complication.

IMPACT OF INNOVATION: We applied a deep learning-based semantic segmentation algorithm to the ureter recognition task and developed a deep learning model called UreterNet. This study aimed to verify whether the ureters could be identified in videos of laparoscopic colorectal surgery.

TECHNOLOGY MATERIALS AND METHODS: Semantic segmentation of the ureter area was performed using a convolutional neural network-based approach. Feature Pyramid Networks were used as the convolutional neural network architecture for semantic segmentation. Precision, recall, and the dice coefficient were used as the evaluation metrics in this study.

PRELIMINARY RESULTS: We created 14,069 annotated images from 304 videos, with 9537, 2266, and 2266 images in the training, validation, and test datasets, respectively. Concerning ureter recognition performance, precision, recall, and the Dice coefficient for the test data were 0.712, 0.722, and 0.716, respectively. Regarding the real-time performance on recorded videos, it took 71 ms for UreterNet to infer all pixels corresponding to the ureter from a single still image and 143 ms to output and display the inferred results as a segmentation mask on the laparoscopic monitor.

CONCLUSIONS AND FUTURE DIRECTIONS: UreterNet is a noninvasive method for identifying the ureter in videos of laparoscopic colorectal surgery and can potentially improve surgical safety. Although this could lead to the development of an image-navigated surgical system, it is necessary to verify whether UreterNet reduces the occurrence of iatrogenic ureteral injury.

PMID:38959453 | DOI:10.1097/DCR.0000000000003335

PLoS Comput Biol. 2024 Jul 3;20(7):e1011198. doi: 10.1371/journal.pcbi.1011198. Online ahead of print.

ABSTRACT

Interpreting transcriptome data is an important yet challenging aspect of bioinformatic analysis. While gene set enrichment analysis is a standard tool for interpreting regulatory changes, we utilize deep learning techniques, specifically autoencoder architectures, to learn latent variables that drive transcriptome signals. We investigate whether simple, variational autoencoder (VAE), and beta-weighted VAE are capable of learning reduced representations of transcriptomes that retain critical biological information. We propose a novel VAE that utilizes priors from biological data to direct the network to learn a representation of the transcriptome that is based on understandable biological concepts. After benchmarking five different autoencoder architectures, we found that each succeeded in reducing the transcriptomes to 50 latent dimensions, which captured enough variation for accurate reconstruction. The simple, fully connected autoencoder, performs best across the benchmarks, but lacks the characteristic of having directly interpretable latent dimensions. The beta-weighted, prior-informed VAE implementation is able to solve the benchmarking tasks, and provide semantically accurate latent features equating to biological pathways. This study opens a new direction for differential pathway analysis in transcriptomics with increased transparency and interpretability.

PMID:38959284 | DOI:10.1371/journal.pcbi.1011198

PLoS One. 2024 Jul 3;19(7):e0305864. doi: 10.1371/journal.pone.0305864. eCollection 2024.

ABSTRACT

This research aims to establish a practical stress detection framework by integrating physiological indicators and deep learning techniques. Utilizing a virtual reality (VR) interview paradigm mirroring real-world scenarios, our focus is on classifying stress states through accessible single-channel electroencephalogram (EEG) and galvanic skin response (GSR) data. Thirty participants underwent stress-inducing VR interviews, with biosignals recorded for deep learning models. Five convolutional neural network (CNN) architectures and one Vision Transformer model, including a multiple-column structure combining EEG and GSR features, showed heightened predictive capabilities and an enhanced area under the receiver operating characteristic curve (AUROC) in stress prediction compared to single-column models. Our experimental protocol effectively elicited stress responses, observed through fluctuations in stress visual analogue scale (VAS), EEG, and GSR metrics. In the single-column architecture, ResNet-152 excelled with a GSR AUROC of 0.944 (±0.027), while the Vision Transformer performed well in EEG, achieving peak AUROC values of 0.886 (±0.069) respectively. Notably, the multiple-column structure, based on ResNet-50, achieved the highest AUROC value of 0.954 (±0.018) in stress classification. Through VR-based simulated interviews, our study induced social stress responses, leading to significant modifications in GSR and EEG measurements. Deep learning models precisely classified stress levels, with the multiple-column strategy demonstrating superiority. Additionally, discreetly placing single-channel EEG measurements behind the ear enhances the convenience and accuracy of stress detection in everyday situations.

PMID:38959272 | DOI:10.1371/journal.pone.0305864

PLOS Digit Health. 2024 Jul 3;3(7):e0000541. doi: 10.1371/journal.pdig.0000541. eCollection 2024 Jul.

ABSTRACT

There is a substantial increase in sexually transmitted infections (STIs) among men who have sex with men (MSM) globally. Unprotected sexual practices, multiple sex partners, criminalization, stigmatisation, fear of discrimination, substance use, poor access to care, and lack of early STI screening tools are among the contributing factors. Therefore, this study applied multilayer perceptron (MLP), extremely randomized trees (ExtraTrees) and XGBoost machine learning models to predict STIs among MSM using bio-behavioural survey (BBS) data in Zimbabwe. Data were collected from 1538 MSM in Zimbabwe. The dataset was split into training and testing sets using the ratio of 80% and 20%, respectively. The synthetic minority oversampling technique (SMOTE) was applied to address class imbalance. Using a stepwise logistic regression model, the study revealed several predictors of STIs among MSM such as age, cohabitation with sex partners, education status and employment status. The results show that MLP performed better than STI predictive models (XGBoost and ExtraTrees) and achieved accuracy of 87.54%, recall of 97.29%, precision of 89.64%, F1-Score of 93.31% and AUC of 66.78%. XGBoost also achieved an accuracy of 86.51%, recall of 96.51%, precision of 89.25%, F1-Score of 92.74% and AUC of 54.83%. ExtraTrees recorded an accuracy of 85.47%, recall of 95.35%, precision of 89.13%, F1-Score of 92.13% and AUC of 60.21%. These models can be effectively used to identify highly at-risk MSM, for STI surveillance and to further develop STI infection screening tools to improve health outcomes of MSM.

PMID:38959248 | DOI:10.1371/journal.pdig.0000541

IEEE/ACM Trans Comput Biol Bioinform. 2024 Jul 3;PP. doi: 10.1109/TCBB.2024.3422288. Online ahead of print.

ABSTRACT

The goal of protein structure refinement is to enhance the precision of predicted protein models, particularly at the residue level of the local structure. Existing refinement approaches primarily rely on physics, whereas molecular simulation methods are resource-intensive and time-consuming. In this study, we employ deep learning methods to extract structural constraints from protein structure residues to assist in protein structure refinement. We introduce a novel method, AnglesRefine, which focuses on a protein's secondary structure and employs transformer to refine various protein structure angles (psi, phi, omega, CA_C_N_angle, C_N_CA_angle, N_CA_C_angle), ultimately generating a superior protein model based on the refined angles. We evaluate our approach against other cutting-edge methods using the CASP11-14 and CASP15 datasets. Experimental outcomes indicate that our method generally surpasses other techniques on the CASP11-14 test dataset, while performing comparably or marginally better on the CASP15 test dataset. Our method consistently demonstrates the least likelihood of model quality degradation, e.g., the degradation percentage of our method is less than 10%, while other methods are about 50%. Furthermore, as our approach eliminates the need for conformational search and sampling, it significantly reduces computational time compared to existing refinement methods.

PMID:38959143 | DOI:10.1109/TCBB.2024.3422288

IEEE Trans Med Imaging. 2024 Jul 3;PP. doi: 10.1109/TMI.2024.3422027. Online ahead of print.

ABSTRACT

One of the primary sources of suboptimal image quality in ultrasound imaging is phase aberration. It is caused by spatial changes in sound speed over a heterogeneous medium, which disturbs the transmitted waves and prevents coherent summation of echo signals. Obtaining non-aberrated ground truths in real-world scenarios can be extremely challenging, if not impossible. This challenge hinders the performance of deep learning-based techniques due to the domain shift between simulated and experimental data. Here, for the first time, we propose a deep learning-based method that does not require ground truth to correct the phase aberration problem and, as such, can be directly trained on real data. We train a network wherein both the input and target output are randomly aberrated radio frequency (RF) data. Moreover, we demonstrate that a conventional loss function such as mean square error is inadequate for training such a network to achieve optimal performance. Instead, we propose an adaptive mixed loss function that employs both B-mode and RF data, resulting in more efficient convergence and enhanced performance. Finally, we publicly release our dataset, comprising over 180,000 aberrated single plane-wave images (RF data), wherein phase aberrations are modeled as near-field phase screens. Although not utilized in the proposed method, each aberrated image is paired with its corresponding aberration profile and the non-aberrated version, aiming to mitigate the data scarcity problem in developing deep learning-based techniques for phase aberration correction. Source code and trained model are also available along with the dataset at http://code.sonography.ai/main-aaa.

PMID:38959140 | DOI:10.1109/TMI.2024.3422027

JMIR Hum Factors. 2024 Jul 3;11:e55964. doi: 10.2196/55964.

ABSTRACT

BACKGROUND: Artificial intelligence (AI) has the potential to enhance physical activity (PA) interventions. However, human factors (HFs) play a pivotal role in the successful integration of AI into mobile health (mHealth) solutions for promoting PA. Understanding and optimizing the interaction between individuals and AI-driven mHealth apps is essential for achieving the desired outcomes.

OBJECTIVE: This study aims to review and describe the current evidence on the HFs in AI-driven digital solutions for increasing PA.

METHODS: We conducted a scoping review by searching for publications containing terms related to PA, HFs, and AI in the titles and abstracts across 3 databases-PubMed, Embase, and IEEE Xplore-and Google Scholar. Studies were included if they were primary studies describing an AI-based solution aimed at increasing PA, and results from testing the solution were reported. Studies that did not meet these criteria were excluded. Additionally, we searched the references in the included articles for relevant research. The following data were extracted from included studies and incorporated into a qualitative synthesis: bibliographic information, study characteristics, population, intervention, comparison, outcomes, and AI-related information. The certainty of the evidence in the included studies was evaluated using GRADE (Grading of Recommendations Assessment, Development, and Evaluation).

RESULTS: A total of 15 studies published between 2015 and 2023 involving 899 participants aged approximately between 19 and 84 years, 60.7% (546/899) of whom were female participants, were included in this review. The interventions lasted between 2 and 26 weeks in the included studies. Recommender systems were the most commonly used AI technology in digital solutions for PA (10/15 studies), followed by conversational agents (4/15 studies). User acceptability and satisfaction were the HFs most frequently evaluated (5/15 studies each), followed by usability (4/15 studies). Regarding automated data collection for personalization and recommendation, most systems involved fitness trackers (5/15 studies). The certainty of the evidence analysis indicates moderate certainty of the effectiveness of AI-driven digital technologies in increasing PA (eg, number of steps, distance walked, or time spent on PA). Furthermore, AI-driven technology, particularly recommender systems, seems to positively influence changes in PA behavior, although with very low certainty evidence.

CONCLUSIONS: Current research highlights the potential of AI-driven technologies to enhance PA, though the evidence remains limited. Longer-term studies are necessary to assess the sustained impact of AI-driven technologies on behavior change and habit formation. While AI-driven digital solutions for PA hold significant promise, further exploration into optimizing AI's impact on PA and effectively integrating AI and HFs is crucial for broader benefits. Thus, the implications for innovation management involve conducting long-term studies, prioritizing diversity, ensuring research quality, focusing on user experience, and understanding the evolving role of AI in PA promotion.

PMID:38959064 | DOI:10.2196/55964

Proc Natl Acad Sci U S A. 2024 Jul 9;121(28):e2320870121. doi: 10.1073/pnas.2320870121. Epub 2024 Jul 3.

ABSTRACT

Efficient storage and sharing of massive biomedical data would open up their wide accessibility to different institutions and disciplines. However, compressors tailored for natural photos/videos are rapidly limited for biomedical data, while emerging deep learning-based methods demand huge training data and are difficult to generalize. Here, we propose to conduct Biomedical data compRession with Implicit nEural Function (BRIEF) by representing the target data with compact neural networks, which are data specific and thus have no generalization issues. Benefiting from the strong representation capability of implicit neural function, BRIEF achieves 2[Formula: see text]3 orders of magnitude compression on diverse biomedical data at significantly higher fidelity than existing techniques. Besides, BRIEF is of consistent performance across the whole data volume, and supports customized spatially varying fidelity. BRIEF's multifold advantageous features also serve reliable downstream tasks at low bandwidth. Our approach will facilitate low-bandwidth data sharing and promote collaboration and progress in the biomedical field.

PMID:38959033 | DOI:10.1073/pnas.2320870121

Ultrasound Q. 2024 Jul 3;40(3):e00683. doi: 10.1097/RUQ.0000000000000683. eCollection 2024 Sep 1.

ABSTRACT

The objective of the study was to use a deep learning model to differentiate between benign and malignant sentinel lymph nodes (SLNs) in patients with breast cancer compared to radiologists' assessments.Seventy-nine women with breast cancer were enrolled and underwent lymphosonography and contrast-enhanced ultrasound (CEUS) examination after subcutaneous injection of ultrasound contrast agent around their tumor to identify SLNs. Google AutoML was used to develop image classification model. Grayscale and CEUS images acquired during the ultrasound examination were uploaded with a data distribution of 80% for training/20% for testing. The performance metric used was area under precision/recall curve (AuPRC). In addition, 3 radiologists assessed SLNs as normal or abnormal based on a clinical established classification. Two-hundred seventeen SLNs were divided in 2 for model development; model 1 included all SLNs and model 2 had an equal number of benign and malignant SLNs. Validation results model 1 AuPRC 0.84 (grayscale)/0.91 (CEUS) and model 2 AuPRC 0.91 (grayscale)/0.87 (CEUS). The comparison between artificial intelligence (AI) and readers' showed statistical significant differences between all models and ultrasound modes; model 1 grayscale AI versus readers, P = 0.047, and model 1 CEUS AI versus readers, P < 0.001. Model 2 r grayscale AI versus readers, P = 0.032, and model 2 CEUS AI versus readers, P = 0.041.The interreader agreement overall result showed κ values of 0.20 for grayscale and 0.17 for CEUS.In conclusion, AutoML showed improved diagnostic performance in balance volume datasets. Radiologist performance was not influenced by the dataset's distribution.

PMID:38958999 | DOI:10.1097/RUQ.0000000000000683

Abdom Radiol (NY). 2024 Jul 3. doi: 10.1007/s00261-024-04468-5. Online ahead of print.

ABSTRACT

OBJECTIVE: To assess impact of image quality on prostate cancer extraprostatic extension (EPE) detection on MRI using a deep learning-based AI algorithm.

MATERIALS AND METHODS: This retrospective, single institution study included patients who were imaged with mpMRI and subsequently underwent radical prostatectomy from June 2007 to August 2022. One genitourinary radiologist prospectively evaluated each patient using the NCI EPE grading system. Each T2WI was classified as low- or high-quality by a previously developed AI algorithm. Fisher's exact tests were performed to compare EPE detection metrics between low- and high-quality images. Univariable and multivariable analyses were conducted to assess the predictive value of image quality for pathological EPE.

RESULTS: A total of 773 consecutive patients (median age 61 [IQR 56-67] years) were evaluated. At radical prostatectomy, 23% (180/773) of patients had EPE at pathology, and 41% (131/318) of positive EPE calls on mpMRI were confirmed to have EPE. The AI algorithm classified 36% (280/773) of T2WIs as low-quality and 64% (493/773) as high-quality. For EPE grade ≥ 1, high-quality T2WI significantly improved specificity for EPE detection (72% [95% CI 67-76%] vs. 63% [95% CI 56-69%], P = 0.03), but did not significantly affect sensitivity (72% [95% CI 62-80%] vs. 75% [95% CI 63-85%]), positive predictive value (44% [95% CI 39-49%] vs. 38% [95% CI 32-43%]), or negative predictive value (89% [95% CI 86-92%] vs. 89% [95% CI 85-93%]). Sensitivity, specificity, PPV, and NPV for EPE grades ≥ 2 and ≥ 3 did not show significant differences attributable to imaging quality. For NCI EPE grade 1, high-quality images (OR 3.05, 95% CI 1.54-5.86; P < 0.001) demonstrated a stronger association with pathologic EPE than low-quality images (OR 1.76, 95% CI 0.63-4.24; P = 0.24).

CONCLUSION: Our study successfully employed a deep learning-based AI algorithm to classify image quality of prostate MRI and demonstrated that better quality T2WI was associated with more accurate prediction of EPE at final pathology.

PMID:38958754 | DOI:10.1007/s00261-024-04468-5

Eur J Nucl Med Mol Imaging. 2024 Jul 3. doi: 10.1007/s00259-024-06818-3. Online ahead of print.

ABSTRACT

PURPOSE: While sedation is routinely used in pediatric PET examinations to preserve diagnostic quality, it may result in side effects and may affect the radiotracer's biodistribution. This study aims to investigate the feasibility of sedation-free pediatric PET imaging using ultra-fast total-body (TB) PET scanners and deep learning (DL)-based attenuation and scatter correction (ASC).

METHODS: This retrospective study included TB PET (uExplorer) imaging of 35 sedated pediatric patients under four years old to determine the minimum effective scanning time. A DL-based ASC method was applied to enhance PET quantification. Both quantitative and qualitative assessments were conducted to evaluate the image quality of ultra-fast DL-ASC PET. Five non-sedated pediatric patients were subsequently used to validate the proposed approach.

RESULTS: Comparisons between standard 300-second and ultra-fast 15-second imaging, CT-ASC and DL-ASC ultra-fast 15-second images, as well as DL-ASC ultra-fast 15-second images in non-sedated and sedated patients, showed no significant differences in qualitative scoring, lesion detectability, and quantitative Standard Uptake Value (SUV) (P = ns).

CONCLUSIONS: This study demonstrates that pediatric PET imaging can be effectively performed without sedation by combining ultra-fast imaging techniques with a DL-based ASC. This advancement in sedation-free ultra-fast PET imaging holds potential for broader clinical adoption.

PMID:38958680 | DOI:10.1007/s00259-024-06818-3

ACS Appl Mater Interfaces. 2024 Jul 3. doi: 10.1021/acsami.4c04491. Online ahead of print.

ABSTRACT

Polyacrylonitrile (PAN) is an important commercial polymer, bearing atactic stereochemistry resulting from nonselective radical polymerization. As such, an accurate, fundamental understanding of governing interactions among PAN molecular units is indispensable for advancing the design principles of final products at reduced processability costs. While ab initio molecular dynamics (AIMD) simulations can provide the necessary accuracy for treating key interactions in polar polymers, such as dipole-dipole interactions and hydrogen bonding, and analyzing their influence on the molecular orientation, their implementation is limited to small molecules only. Herein, we show that the neural network interatomic potentials (NNIPs) that are trained on the small-scale AIMD data (acquired for oligomers) can be efficiently employed to examine the structures and properties at large scales (polymers). NNIP provides critical insight into intra- and interchain hydrogen-bonding and dipolar correlations and accurately predicts the amorphous bulk PAN structure validated by modeling the experimental X-ray structure factor. Furthermore, the NNIP-predicted PAN properties, such as density and elastic modulus, are in good agreement with their experimental values. Overall, the trend in the elastic modulus is found to correlate strongly with the PAN structural orientations encoded in the Hermans orientation factor. This study enables the ability to predict the structure-property relations for PAN and analogues with sustainable ab initio accuracy across scales.

PMID:38958640 | DOI:10.1021/acsami.4c04491

J Acoust Soc Am. 2024 Jul 1;156(1):151-163. doi: 10.1121/10.0026474.

ABSTRACT

This paper explores the application of deep reinforcement learning for autonomously designing noise-mitigating structures. Specifically, deep Q- and double deep Q-networks are employed to find material distributions that result in broadband noise mitigation for reflection and transmission problems. Unlike conventional deep learning approaches which require prior knowledge for data labeling, the double deep Q-network algorithm learns configurations that result in broadband noise mitigations without prior knowledge by utilizing pixel-based inputs. By employing unified hyperparameters and network architectures for transmission and reflection problems, the capability of the algorithms to generalize over different environments is demonstrated. In addition, a comparison with a genetic algorithm highlights the potential for generalized design in complex environments, despite the algorithms tending to predict local maxima. Furthermore, we examine the impact of hyperparameters and environment types on agent performance. The autonomous design approach offers generalized learning while avoiding restrictions to specific shapes or prior knowledge of the task.

PMID:38958582 | DOI:10.1121/10.0026474

Crit Care Med. 2024 Jul 3. doi: 10.1097/CCM.0000000000006359. Online ahead of print.

ABSTRACT

OBJECTIVES: To evaluate the transferability of deep learning (DL) models for the early detection of adverse events to previously unseen hospitals.

DESIGN: Retrospective observational cohort study utilizing harmonized intensive care data from four public datasets.

SETTING: ICUs across Europe and the United States.

PATIENTS: Adult patients admitted to the ICU for at least 6 hours who had good data quality.

INTERVENTIONS: None.

MEASUREMENTS AND MAIN RESULTS: Using carefully harmonized data from a total of 334,812 ICU stays, we systematically assessed the transferability of DL models for three common adverse events: death, acute kidney injury (AKI), and sepsis. We tested whether using more than one data source and/or algorithmically optimizing for generalizability during training improves model performance at new hospitals. We found that models achieved high area under the receiver operating characteristic (AUROC) for mortality (0.838-0.869), AKI (0.823-0.866), and sepsis (0.749-0.824) at the training hospital. As expected, AUROC dropped when models were applied at other hospitals, sometimes by as much as -0.200. Using more than one dataset for training mitigated the performance drop, with multicenter models performing roughly on par with the best single-center model. Dedicated methods promoting generalizability did not noticeably improve performance in our experiments.

CONCLUSIONS: Our results emphasize the importance of diverse training data for DL-based risk prediction. They suggest that as data from more hospitals become available for training, models may become increasingly generalizable. Even so, good performance at a new hospital still depended on the inclusion of compatible hospitals during training.

PMID:38958568 | DOI:10.1097/CCM.0000000000006359

J Chem Inf Model. 2024 Jul 3. doi: 10.1021/acs.jcim.4c00072. Online ahead of print.

ABSTRACT

In drug discovery, molecular docking methods face challenges in accurately predicting energy. Scoring functions used in molecular docking often fail to simulate complex protein-ligand interactions fully and accurately leading to biases and inaccuracies in virtual screening and target predictions. We introduce the "Docking Score ML", developed from an analysis of over 200,000 docked complexes from 155 known targets for cancer treatments. The scoring functions used are founded on bioactivity data sourced from ChEMBL and have been fine-tuned using both supervised machine learning and deep learning techniques. We validated our approach extensively using multiple data sets such as validation of selectivity mechanism, the DUDE, DUD-AD, and LIT-PCBA data sets, and performed a multitarget analysis on drugs like sunitinib. To enhance prediction accuracy, feature fusion techniques were explored. By merging the capabilities of the Graph Convolutional Network (GCN) with multiple docking functions, our results indicated a clear superiority of our methodologies over conventional approaches. These advantages demonstrate that Docking Score ML is an efficient and accurate tool for virtual screening and reverse docking.

PMID:38958413 | DOI:10.1021/acs.jcim.4c00072

JMIR Ment Health. 2024 Jul 2;11:e56569. doi: 10.2196/56569.

ABSTRACT

Large language model (LLM)-powered services are gaining popularity in various applications due to their exceptional performance in many tasks, such as sentiment analysis and answering questions. Recently, research has been exploring their potential use in digital health contexts, particularly in the mental health domain. However, implementing LLM-enhanced conversational artificial intelligence (CAI) presents significant ethical, technical, and clinical challenges. In this viewpoint paper, we discuss 2 challenges that affect the use of LLM-enhanced CAI for individuals with mental health issues, focusing on the use case of patients with depression: the tendency to humanize LLM-enhanced CAI and their lack of contextualized robustness. Our approach is interdisciplinary, relying on considerations from philosophy, psychology, and computer science. We argue that the humanization of LLM-enhanced CAI hinges on the reflection of what it means to simulate "human-like" features with LLMs and what role these systems should play in interactions with humans. Further, ensuring the contextualization of the robustness of LLMs requires considering the specificities of language production in individuals with depression, as well as its evolution over time. Finally, we provide a series of recommendations to foster the responsible design and deployment of LLM-enhanced CAI for the therapeutic support of individuals with depression.

PMID:38958218 | DOI:10.2196/56569

EngMedicine. 2024 Jun;1(1):100010. doi: 10.1016/j.engmed.2024.100010. Epub 2024 May 15.

ABSTRACT

Kidney failure is particularly common in the United States, where it affects over 700,000 individuals. It is typically treated through repeated sessions of hemodialysis to filter and clean the blood. Hemodialysis requires vascular access, in about 70% of cases through an arteriovenous fistula (AVF) surgically created by connecting an artery and vein. AVF take 6 weeks or more to mature. Mature fistulae often require intervention, most often percutaneous transluminal angioplasty (PTA), also known as fistulaplasty, to maintain the patency of the fistula. PTA is also the first-line intervention to restore blood flow and prolong the use of an AVF, and many patients undergo the procedure multiple times. Although PTA is important for AVF maturation and maintenance, research into predictive models of AVF function following PTA has been limited. Therefore, in this paper we hypothesize that based on patient-specific information collected during PTA, a predictive model can be created to help improve treatment planning. We test a set of rich, multimodal data from 28 patients that includes medical history, AVF blood flow, and interventional angiographic imaging (specifically excluding any post-PTA measurements) and build deep hybrid neural networks. A hybrid model combining a 3D convolutional neural network with a multi-layer perceptron to classify AVF was established. We found using this model that we were able to identify the association between different factors and evaluate whether the PTA procedure can maintain primary patency for more than 3 months. The testing accuracy achieved was 0.75 with a weighted F1-score of 0.75, and AUROC of 0.75. These results indicate that evaluating multimodal clinical data using artificial neural networks can predict the outcome of PTA. These initial findings suggest that the hybrid model combining clinical data, imaging and hemodynamic analysis can be useful to treatment planning for hemodialysis. Further study based on a large cohort is needed to refine the accuracy and model efficiency.

PMID:38957294 | PMC:PMC11218659 | DOI:10.1016/j.engmed.2024.100010

Cureus. 2024 Jun 2;16(6):e61523. doi: 10.7759/cureus.61523. eCollection 2024 Jun.

ABSTRACT

This review aims to explore the potential of artificial intelligence (AI) in coronary CT angiography (CCTA), a key tool for diagnosing coronary artery disease (CAD). Because CAD is still a major cause of death worldwide, effective and accurate diagnostic methods are required to identify and manage the condition. CCTA certainly is a noninvasive alternative for diagnosing CAD, but it requires a large amount of data as input. We intend to discuss the idea of incorporating AI into CCTA, which enhances its diagnostic accuracy and operational efficiency. Using such AI technologies as machine learning (ML) and deep learning (DL) tools, CCTA images are automated to perfection and the analysis is significantly refined. It enables the characterization of a plaque, assesses the severity of the stenosis, and makes more accurate risk stratifications than traditional methods, with pinpoint accuracy. Automating routine tasks through AI-driven CCTA will reduce the radiologists' workload considerably, which is a standard benefit of such technologies. More importantly, it would enable radiologists to allocate more time and expertise to complex cases, thereby improving overall patient care. However, the field of AI in CCTA is not without its challenges, which include data protection, algorithm transparency, as well as criteria for standardization encoding. Despite such obstacles, it appears that the integration of AI technology into CCTA in the future holds great promise for keeping CAD itself in check, thereby aiding the fight against this disease and begetting better clinical outcomes and more optimized modes of healthcare. Future research on AI algorithms for CCTA, making ethical use of AI, and thereby overcoming the technical and clinical barriers to widespread adoption of this new tool, will hopefully pave the way for profound AI-driven transformations in healthcare.

PMID:38957241 | PMC:PMC11218716 | DOI:10.7759/cureus.61523

Proc SPIE Int Soc Opt Eng. 2024 Feb;12930:129301T. doi: 10.1117/12.3006771. Epub 2024 Apr 2.

ABSTRACT

Organ segmentation is a crucial task in various medical imaging applications. Many deep learning models have been developed to do this, but they are slow and require a lot of computational resources. To solve this problem, attention mechanisms are used which can locate important objects of interest within medical images, allowing the model to segment them accurately even when there is noise or artifact. By paying attention to specific anatomical regions, the model becomes better at segmentation. Medical images have unique features in the form of anatomical information, which makes them different from natural images. Unfortunately, most deep learning methods either ignore this information or do not use it effectively and explicitly. Combined natural intelligence with artificial intelligence, known as hybrid intelligence, has shown promising results in medical image segmentation, making models more robust and able to perform well in challenging situations. In this paper, we propose several methods and models to find attention regions in medical images for deep learning-based segmentation via non-deep-learning methods. We developed these models and trained them using hybrid intelligence concepts. To evaluate their performance, we tested the models on unique test data and analyzed metrics including false negatives quotient and false positives quotient. Our findings demonstrate that object shape and layout variations can be explicitly learned to create computational models that are suitable for each anatomic object. This work opens new possibilities for advancements in medical image segmentation and analysis.

PMID:38957740 | PMC:PMC11218901 | DOI:10.1117/12.3006771