Deep Learning Platform Research Articles

Deep generative design of RNA aptamers using structural predictions.

RNAs represent a class of programmable biomolecules capable of performing diverse biological functions. Recent studies have developed accurate RNA three-dimensional structure prediction methods, which may enable new RNAs to be designed in a structure-guided manner. Here, we develop a structure-to-sequence deep learning platform for the de novo generative design of RNA aptamers. We show that our approach can design RNA aptamers that are predicted to be structurally similar, yet sequence dissimilar, to known light-up aptamers that fluoresce in the presence of small molecules. We experimentally validate several generated RNA aptamers to have fluorescent activity, show that these aptamers can be optimized for activity in silico, and find that they exhibit a mechanism of fluorescence similar to that of known light-up aptamers. Our results demonstrate how structural predictions can guide the targeted and resource-efficient design of new RNA sequences.

Artificial Intelligence Assisted Surgical Scene Recognition: A Comparative Study Amongst Healthcare Professionals.

This study aimed to compare the ability of a deep-learning platform (the MACSSwin-T model) with healthcare professionals in detecting cerebral aneurysms from operative videos. Secondly, we aimed to compare the neurosurgical team's ability to detect cerebral aneurysms with and without AI-assistance. Modern microscopic surgery enables the capture of operative video data on an unforeseen scale. Advances in computer vision, a branch of artificial intelligence (AI), have enabled automated analysis of operative video. These advances are likely to benefit clinicians, healthcare systems, and patients alike, yet such benefits are yet to be realised. In a cross-sectional comparative study, neurosurgeons, anaesthetists, and operating room (OR) nurses, all at varying stages of training and experience, reviewed still frames of aneurysm clipping operations and labelled frames as "aneurysm not in frame" or "aneurysm in frame". Frames then underwent analysis by the AI platform. A second round of data collection was performed whereby the neurosurgical team had AI-assistance. Accuracy of aneurysm detection was calculated for human only, AI only, and AI-assisted human groups. 5,154 individual frame reviews were collated from 338 healthcare professionals. Healthcare professionals correctly labelled 70% of frames without AI assistance, compared to 78% with AI-assistance (OR 1.77, P<0.001). Neurosurgical Attendings showed the greatest improvement, from 77% to 92% correct predictions with AI-assistance (OR 4.24, P=0.003). AI-assisted human performance surpassed both human and AI alone. Notably, across healthcare professionals surveyed, frame accuracy improved across all subspecialties and experience levels, particularly among the most experienced healthcare professionals. These results challenge the prevailing notion that AI primarily benefits junior clinicians, highlighting its crucial role throughout the surgical hierarchy as an essential component of modern surgical practice.

A Novel Natural Language Processing Model Transfer Strategy Tailored for Deep Learning Platforms

As artificial intelligence technology continues to evolve, numerous advancements enable smoother communication and collaboration between humans and computers. Natural language processing (NLP), as a key direction in the field of computer and artificial intelligence, has been attracting much attention. It focuses on the language used by people in daily life, devoting itself to improving the ability of language generation and understanding, thus making communication between people and computers more fluent and natural, effectively breaking through the problem of human-computer interaction. With the advancement of deep learning technology, the migration of NLP models onto deep learning platforms has emerged as a key trend. Model transfer, a crucial aspect of deep learning, holds significant value in enhancing the performance and efficiency of NLP models. This paper begins by outlining the fundamentals of deep learning platforms and NLP model transfer, followed by a comprehensive examination of current research progress and challenges in this field. Subsequently, we introduce a novel NLP model transfer strategy tailored for deep learning platforms and validate its effectiveness through rigorous experiments. In conclusion, the paper highlights our noteworthy advancements, pointing toward promising future developments.

DFUCare: deep learning platform for diabetic foot ulcer detection, analysis, and monitoring.

Diabetic foot ulcers (DFUs) are a severe complication among diabetic patients, often leading to amputation or even death. Early detection of infection and ischemia is essential for improving healing outcomes, but current diagnostic methods are invasive, time-consuming, and costly. There is a need for non-invasive, efficient, and affordable solutions in diabetic foot care. We developed DFUCare, a platform that leverages computer vision and deep learning (DL) algorithms to localize, classify, and analyze DFUs non-invasively. The platform combines CIELAB and YCbCr color space segmentation with a pre-trained YOLOv5s algorithm for wound localization. Additionally, deep-learning models were implemented to classify infection and ischemia in DFUs. The preliminary performance of the platform was tested on wound images acquired using a cell phone. DFUCare achieved an F1-score of 0.80 and a mean Average Precision (mAP) of 0.861 for wound localization. For infection classification, we obtained a binary accuracy of 79.76%, while ischemic classification reached 94.81% on the validation set. The system successfully measured wound size and performed tissue color and textural analysis for a comparative assessment of macroscopic wound features. In clinical testing, DFUCare localized wounds and predicted infected and ischemic with an error rate of less than 10%, underscoring the strong performance of the platform. DFUCare presents an innovative approach to wound care, offering a cost-effective, remote, and convenient healthcare solution. By enabling non-invasive and accurate analysis of wounds using mobile devices, this platform has the potential to revolutionize diabetic foot care and improve clinical outcomes through early detection of infection and ischemia.

Computer Allied Intelligence in the Education Resource-Sharing Based inContract Deep Learning

A resource-sharing platform is an online space where individuals or organizations can exchange or access various resources, such as knowledge, skills, tools, or assets. These platforms facilitate collaboration and cooperation among users by providing a centralized hub for sharing resources efficiently. In university higher education, a resource-sharing platform serves as a vital tool for students, faculty, and staff to optimize access to a wide array of educational resources. These platforms enable the sharing of academic materials, such as textbooks, lecture notes, and research papers, fostering collaborative learning environments. This paper proposed the proposed method of Distributed Consensus Blockchain Deep Learning (dCBDL) for a resource-sharing deep learning platform in higher education aims to revolutionize collaborative learning and resource utilization. Initially, the dCBDL model uses a blockchain network that serves as the underlying infrastructure for the platform. This blockchain will store transaction records, verify the authenticity of shared resources, and ensure data integrity and immutability. The proposed model uses smart contracts and decentralized storage with a consensus mechanism for data processing in higher education. Finally, the proposed dCBDL model uses the deep learning model for the classification and assessment of student performance in higher education. The proposed model achieves a classification accuracy of 98% which is significantly higher than the conventional techniques such as blockchain architecture and SVM and LSTM classifiers.

Cataract and glaucoma detection based on Transfer Learning using MobileNet

A serious eye condition called cataracts can cause blindness. Early and accurate cataract detection is the most effective method for reducing risk and averting blindness. The optic nerve head is harmed by the neurodegenerative condition known as glaucoma. Machine learning and deep learning systems for glaucoma and cataract detection have recently received much attention in research. The automatic detection of these diseases also depends on deep learning transfer learning platforms like VeggNet, ResNet, and MobilNet. The authors proposed MobileNetV1 and MobileNetV2 based on an optimized architecture building lightweight deep neural networks using depth-wise separable convolutions. The experiments used publicly available data sets with both cataract & normal and glaucoma & normal images, and the results showed that the proposed model had the highest accuracy compared to the other models.

MalariaFlow: A comprehensive deep learning platform for multistage phenotypic antimalarial drug discovery

Malaria remains a significant global health challenge due to the growing drug resistance of Plasmodium parasites and the failure to block transmission within human host. While machine learning (ML) and deep learning (DL) methods have shown promise in accelerating antimalarial drug discovery, the performance of deep learning models based on molecular graph and other co-representation approaches warrants further exploration. Current research has overlooked mutant strains of the malaria parasite with varying degrees of sensitivity or resistance, and has not covered the prediction of inhibitory activities across the three major life cycle stages (liver, asexual blood, and gametocyte) within the human host, which is crucial for both treatment and transmission blocking. In this study, we manually curated a benchmark antimalarial activity dataset comprising 407,404 unique compounds and 410,654 bioactivity data points across ten Plasmodium phenotypes and three stages. The performance was systematically compared among two fingerprint-based ML models (RF::Morgan and XGBoost:Morgan), four graph-based DL models (GCN, GAT, MPNN, and Attentive FP), and three co-representations DL models (FP-GNN, HiGNN, and FG-BERT), which reveal that: 1) The FP-GNN model achieved the best predictive performance, outperforming the other methods in distinguishing active and inactive compounds across balanced, more positive, and more negative datasets, with an overall AUROC of 0.900; 2) Fingerprint-based ML models outperformed graph-based DL models on large datasets (>1000 compounds), but the three co-representations DL models were able to incorporate domain-specific chemical knowledge to bridge this gap, achieving better predictive performance. These findings provide valuable guidance for selecting appropriate ML and DL methods for antimalarial activity prediction tasks. The interpretability analysis of the FP-GNN model revealed its ability to accurately capture the key structural features responsible for the liver- and blood-stage activities of the known antimalarial drug atovaquone. Finally, we developed a web server, MalariaFlow, incorporating these high-quality models for antimalarial activity prediction, virtual screening, and similarity search, successfully predicting novel triple-stage antimalarial hits validated through experimental testing, demonstrating its effectiveness and value in discovering potential multistage antimalarial drug candidates.

Efficient and Accurate 3D Thickness Measurement in Vessel Wall Imaging: Overcoming Limitations of 2D Approaches Using the Laplacian Method.

The clinical significance of measuring vessel wall thickness is widely acknowledged. Recent advancements have enabled high-resolution 3D scans of arteries and precise segmentation of their lumens and outer walls; however, most existing methods for assessing vessel wall thickness are 2D. Despite being valuable, reproducibility and accuracy of 2D techniques depend on the extracted 2D slices. Additionally, these methods fail to fully account for variations in wall thickness in all dimensions. Furthermore, most existing approaches are difficult to be extended into 3D and their measurements lack spatial localization and are primarily confined to lumen boundaries. We advocate for a shift in perspective towards recognizing vessel wall thickness measurement as inherently a 3D challenge and propose adapting the Laplacian method as an outstanding alternative. The Laplacian method is implemented using convolutions, ensuring its efficient and rapid execution on deep learning platforms. Experiments using digital phantoms and vessel wall imaging data are conducted to showcase the accuracy, reproducibility, and localization capabilities of the proposed approach. The proposed method produce consistent outcomes that remain independent of centerlines and 2D slices. Notably, this approach is applicable in both 2D and 3D scenarios. It allows for voxel-wise quantification of wall thickness, enabling precise identification of wall volumes exhibiting abnormal wall thickness. Our research highlights the urgency of transitioning to 3D methodologies for vessel wall thickness measurement. Such a transition not only acknowledges the intricate spatial variations of vessel walls, but also opens doors to more accurate, localized, and insightful diagnostic insights.

Understanding patient-derived tumor organoid growth through an integrated imaging and mathematical modeling framework.

Patient-derived tumor organoids (PDTOs) are novel cellular models that maintain the genetic, phenotypic and structural features of patient tumor tissue and are useful for studying tumorigenesis and drug response. When integrated with advanced 3D imaging and analysis techniques, PDTOs can be used to establish physiologically relevant high-throughput and high-content drug screening platforms that support the development of patient-specific treatment strategies. However, in order to effectively leverage high-throughput PDTO observations for clinical predictions, it is critical to establish a quantitative understanding of the basic properties and variability of organoid growth dynamics. In this work, we introduced an innovative workflow for analyzing and understanding PDTO growth dynamics, by integrating a high-throughput imaging deep learning platform with mathematical modeling, incorporating flexible growth laws and variable dormancy times. We applied the workflow to colon cancer organoids and demonstrated that organoid growth is well-described by the Gompertz model of growth. Our analysis showed significant intrapatient heterogeneity in PDTO growth dynamics, with the initial exponential growth rate of an organoid following a lognormal distribution within each dataset. The level of intrapatient heterogeneity varied between patients, as did organoid growth rates and dormancy times of single seeded cells. Our work contributes to an emerging understanding of the basic growth characteristics of PDTOs, and it highlights the heterogeneity in organoid growth both within and between patients. These results pave the way for further modeling efforts aimed at predicting treatment response dynamics and drug resistance timing.

Deep Learning Artificial Intelligence Predicts Homologous Recombination Deficiency and Platinum Response From Histologic Slides.

Cancers with homologous recombination deficiency (HRD) can benefit from platinum salts and poly(ADP-ribose) polymerase inhibitors. Standard diagnostic tests for detecting HRD require molecular profiling, which is not universally available. We trained DeepHRD, a deep learning platform for predicting HRD from hematoxylin and eosin (H&E)-stained histopathological slides, using primary breast (n = 1,008) and ovarian (n = 459) cancers from The Cancer Genome Atlas (TCGA). DeepHRD was compared with four standard HRD molecular tests using breast (n = 349) and ovarian (n = 141) cancers from multiple independent data sets, including platinum-treated clinical cohorts with RECIST progression-free survival (PFS), complete response (CR), and overall survival (OS) endpoints. DeepHRD predicted HRD from held-out H&E-stained breast cancer slides in TCGA with an AUC of 0.81 (95% CI, 0.77 to 0.85). This performance was confirmed in two independent primary breast cancer cohorts (AUC, 0.76 [95% CI, 0.71 to 0.82]). In an external platinum-treated metastatic breast cancer cohort, samples predicted as HRD had higher complete CR (AUC, 0.76 [95% CI, 0.54 to 0.93]) with 3.7-fold increase in median PFS (14.4 v 3.9 months; P = .0019) and hazard ratio (HR) of 0.45 (P = .0047). There were no significant differences in nonplatinum treatment outcome by predicted HRD status in three breast cancer cohorts, including CR (AUC, 0.39) and PFS (HR, 0.98, P = .95) in taxane-treated metastatic breast cancer. Through transfer learning to high-grade serous ovarian cancer, DeepHRD-predicted HRD samples had better OS after first-line (HR, 0.46; P = .030) and neoadjuvant (HR, 0.49; P = .015) platinum therapy in two cohorts. DeepHRD can predict HRD in breast and ovarian cancers directly from routine H&E slides across multiple external cohorts, slide scanners, and tissue fixation variables. When compared with molecular testing, DeepHRD classified 1.8- to 3.1-fold more patients with HRD, which exhibited better OS in high-grade serous ovarian cancer and platinum-specific PFS in metastatic breast cancer.

AMIKOMNET: Novel Structure for a Deep Learning Model to Enhance COVID-19 Classification Task Performance

Since early 2020, coronavirus has spread extensively throughout the globe. It was first detected in Wuhan, a province in China. Many researchers have proposed various models to solve problems related to COVID-19 detection. As traditional medical approaches take a lot of time to detect the virus and require specific laboratory tests, the adoption of artificial intelligence (AI), including machine learning, might play an important role in handling the problem. A great deal of research has seen the adoption of AI succeed in the early detection of COVID-19 using X-ray images. Unfortunately, the majority of deep learning adoption for COVID-19 detection has the shortcomings of high error detection and high computation costs. In this study, we employed a hybrid model using an auto-encoder (AE) and a convolutional neural network (CNN) (named AMIKOMNET) with a small number of layers and parameters. We implemented an ensemble learning mechanism in the AMIKOMNET model using Adaboost with the aim of reducing error detection in COVID-19 classification tasks. The experimental results for the binary class show that our model achieved high effectiveness, with 96.90% accuracy, 95.06% recall, 94.67% F1-score, and 96.03% precision. The experimental result for the multiclass achieved 95.13% accuracy, 94.93% recall, 95.75% F1-score, and 96.19% precision. The adoption of Adaboost in AMIKOMNET for the binary class increased the effectiveness of the model to 98.45% accuracy, 96.16% recall, 95.70% F1-score, and 96.87% precision. The adoption of Adaboost in AMIKOMNET in the multiclass classification task also saw an increase in performance, with an accuracy of 96.65%, a recall of 94.93%, an F1-score of 95.76%, and a precision of 96.19%. The implementation of AE to handle image feature extraction combined with a CNN used to handle dimensional image feature reduction achieved outstanding performance when compared to previous work using a deep learning platform. Exploiting Adaboost also increased the effectiveness of the AMIKOMNET model in detecting COVID-19.

PrivDNN: A Secure Multi-Party Computation Framework for Deep Learning using Partial DNN Encryption

In the past decade, we have witnessed an exponential growth of deep learning models, platforms, and applications. While existing DL applications and Machine Learning as a service (MLaaS) frameworks assume fully trusted models, the need for privacy-preserving DNN evaluation arises. In a secure multi-party computation scenario, both the model and the data are considered proprietary, i.e., the model owner does not want to reveal the highly valuable DL model to the user, while the user does not wish to disclose their private data samples either. Conventional privacy-preserving deep learning solutions ask the users to send encrypted samples to the model owners, who must handle the heavy lifting of ciphertext-domain computation with homomorphic encryption. In this paper, we present a novel solution, namely, PrivDNN, which (1) offloads the computation to the user side by sharing an encrypted deep learning model with them, (2) significantly improves the efficiency of DNN evaluation using partial DNN encryption, (3) ensures model accuracy and model privacy using a core neuron selection and encryption scheme. Experimental results show that PrivDNN reduces privacy-preserving DNN inference time and memory requirement by up to 97% while maintaining model performance and privacy. Codes can be found at https://github.com/LiangqinRen/PrivDNN

Lung Tumor Classification Optimizer with Augment Input Images

This work focuses on the accurate classification of lung tumors in Computed Tomography (CT) images to determine tumor type and stage, essential for guiding treatment decisions. In this research, pretrained model VGG19 and Generative Adversarial Networks (GANs) models are used to classify the lung tumor in order to enhance the classification accuracy by increasing the diversity of the training dataset. The dataset comprises 1097 CT image samples, with 120 classified as benign, 561 as malignant, and 416 as normal cases. The coding for this work was developed using Python, with TensorFlow as the deep learning platform, and simulations were conducted on Google Colab. Performance metrics such as accuracy, sensitivity, specificity, and F1 score are evaluated to assess the effectiveness of the classification model. The datasets were obtained from the Iraq-Oncology Teaching Hospital.

Evaluating the deep learning software tools for large-scale enterprises using a novel TODIFFA-MCDM framework

Deep learning (DL) is one of the most promising technological developments emerging in the fourth industrial revolution era for businesses to improve processes, increase efficiency, and reduce errors. Accordingly, hierarchical learning software selection is one of the most critical decision-making problems in integrating neural network applications into business models. However, selecting appropriate reinforcement learning software for integrating deep learning applications into enterprises’ business models takes much work for decision-makers. There are several reasons for this: first, practitioners’ limited knowledge and experience of DL makes it difficult for decision-makers to adapt this technology into their enterprises’ business model and significantly increases complex uncertainties. Secondly, according to the authors’ knowledge, no study in the literature addresses deep structured learning solutions with the help of MCDM approaches. Consequently, making inferences concerning criteria that should be considered in an evaluation process is impossible by considering the studies in the relevant literature. Considering these gaps, this study presents a novel decision-making approach developed by the authors. It involves the combination of two new decision-making approaches, MAXC (MAXimum of Criterion) and TODIFFA (the total differential of alternative), which were developed to solve current decision-making problems. When the most important advantages of this model are considered, it associates objective and subjective approaches and eliminates some critical limitations of these methodologies. Besides, it has an easily followable algorithm without the need for advanced mathematical knowledge for practitioners and provides highly stable and reliable results in solving complex decision-making problems. Another novelty of the study is that the criteria are determined with a long-term negotiation process that is part of comprehensive fieldwork with specialists. When the conclusions obtained using this model are briefly reviewed, the C2 “Data Availability and Quality” criterion is the most influential in selecting deep learning software. The C7 “Time Constraints” criterion follows the most influential factor. Remarkably, prior research has overlooked the correlation between the performance of Deep Learning (DL) platforms and the quality and accessibility of data. The findings of this study underscore the necessity for DL platform developers to devise solutions to enable DL platforms to operate effectively, notwithstanding the availability of clean, high-quality, and adequate data. Finally, the robustness check carried out to test the validity of the proposed model confirms the accuracy and robustness of the results obtained by implementing the suggested model.

Integrated Extended Kalman Filter and Deep Learning Platform for Electric Vehicle Battery Health Prediction

As the demand for electric vehicles (EVs) rises globally, ensuring the safety and reliability of EV battery systems becomes paramount. Accurately predicting the state of health (SoH) and state of charge (SoC) of EV batteries is crucial for maintaining their safe and consistent operation. This paper introduces a novel approach leveraging deep learning methodologies to predict battery SoH, focusing on implementing a system prototype for real-world applications. The proposed system integrates an extended Kalman filter (EKF) with a deep learning framework, forming a system prototype known as FELL, aimed at EV battery diagnosis and prediction. We devise an algorithm utilizing the EKF to estimate the SoH of the battery. We present a detailed overview of the system architecture and implementation, showcasing its predictive capabilities. Experimental results demonstrate the effectiveness of the system in accurately estimating battery SoH with notable improvements in prediction accuracy. Additionally, the FELL system provides users with real-time predictions and comparative analysis across multiple prediction models, offering valuable insights for EV battery management.

Automated machine learning model for fundus image classification by health-care professionals with no coding experience

To assess the feasibility of code-free deep learning (CFDL) platforms in the prediction of binary outcomes from fundus images in ophthalmology, evaluating two distinct online-based platforms (Google Vertex and Amazon Rekognition), and two distinct datasets. Two publicly available datasets, Messidor-2 and BRSET, were utilized for model development. The Messidor-2 consists of fundus photographs from diabetic patients and the BRSET is a multi-label dataset. The CFDL platforms were used to create deep learning models, with no preprocessing of the images, by a single ophthalmologist without coding expertise. The performance metrics employed to evaluate the models were F1 score, area under curve (AUC), precision and recall. The performance metrics for referable diabetic retinopathy and macular edema were above 0.9 for both tasks and CFDL. The Google Vertex models demonstrated superior performance compared to the Amazon models, with the BRSET dataset achieving the highest accuracy (AUC of 0.994). Multi-classification tasks using only BRSET achieved similar overall performance between platforms, achieving AUC of 0.994 for laterality, 0.942 for age grouping, 0.779 for genetic sex identification, 0.857 for optic, and 0.837 for normality with Google Vertex. The study demonstrates the feasibility of using automated machine learning platforms for predicting binary outcomes from fundus images in ophthalmology. It highlights the high accuracy achieved by the models in some tasks and the potential of CFDL as an entry-friendly platform for ophthalmologists to familiarize themselves with machine learning concepts.

Analysis of 3D pathology samples using weakly supervised AI

Human tissue, which is inherently three-dimensional (3D), is traditionally examined through standard-of-care histopathology as limited two-dimensional (2D) cross-sections that can insufficiently represent the tissue due to sampling bias. To holistically characterize histomorphology, 3D imaging modalities have been developed, but clinical translation is hampered by complex manual evaluation and lack of computational platforms to distill clinical insights from large, high-resolution datasets. We present TriPath, a deep-learning platform for processing tissue volumes and efficiently predicting clinical outcomes based on 3D morphological features. Recurrence risk-stratification models were trained on prostate cancer specimens imaged with open-top light-sheet microscopy or microcomputed tomography. By comprehensively capturing 3D morphologies, 3D volume-based prognostication achieves superior performance to traditional 2D slice-based approaches, including clinical/histopathological baselines from six certified genitourinary pathologists. Incorporating greater tissue volume improves prognostic performance and mitigates risk prediction variability from sampling bias, further emphasizing the value of capturing larger extents of heterogeneous morphology.

Brain Tumor Recognition Using Artificial Intelligence Neural-Networks (BRAIN): A Cost-Effective Clean-Energy Platform

Brain tumors necessitate swift detection and classification for optimal patient outcomes. Deep learning has been extensively utilized to recognize complex tumor patterns in magnetic resonance imaging (MRI) images, aiding in tumor diagnosis, treatment, and prognostication. However, model complexity and limited generalizability with unfamiliar data hinder appropriate clinical integration. The objective of this study is to develop a clean-energy cloud-based deep learning platform to classify brain tumors. Three datasets of a total of 2611 axial MRI images were used to train our multi-layer convolutional neural network (CNN). Our platform automatically optimized every transfer learning and data augmentation feature combination to provide the highest predictive accuracy for our classification task. Our proposed system identified and classified brain tumors successfully and efficiently with an overall precision value of 96.8% [95% CI; 93.8–97.6]. Using clean energy supercomputing resources and cloud platforms cut our workflow to 103 min, $0 in total cost, and a negligible carbon footprint (0.0014 kg eq CO2). By leveraging automated optimized learning, we developed a cost-effective deep learning (DL) platform that accurately classified brain tumors from axial MRI images of different levels. Although studies have identified machine learning tools to overcome these obstacles, only some are cost-effective, generalizable, and usable regardless of experience.

Using Deep Learning Approaches to Quantify Landscape Preference of the Chinese Grand Canal: An Empirical Case Study of the Yangzhou Ancient Canal

Landscape preference emerges from the dynamic interaction between individuals and their environment and plays a pivotal role in the preservation and enhancement of the Chinese Grand Canal’s scenery. As a vast linear heritage, employing conventional methods for analyzing landscape preferences can be resource-intensive in terms of both time and labor. Amid the rapid advancement of Big Data and Artificial Intelligence (AI), a cognitive framework for understanding the Chinese Grand Canal’s landscape preferences has been developed, encompassing two primary aspects: the characteristic features of landscape preference and its spatial organization. Geotagged photographs from tourism media platforms focused on the Yangzhou Ancient Canal were utilized, and the EasyDL deep learning platform was employed to devise a model. This model assesses current landscape preferences through an analysis of photographic content, element composition patterns, and geospatial distribution, integrating social network and point density analyses. Our findings reveal that the fusion of Yangzhou Ancient Canal and classical gardens creates a sought-after ‘Canal and Watercraft Remains’ landscape. Tourists’ preferences for different landscape types are reflected in the way the elements are combined in the photographs. Overall, landscape preferences are dense in the north and sparse in the south. Differences in tourists’ perceptions of the value of and preferences for heritage sites lead to significant variations in tourist arrivals at different sites. This approach demonstrates efficiency and scalability in evaluating the Chinese Grand Canal landscape, offering valuable insights for its strategic planning and conservation efforts.

Creatinine-to-cystatin C ratio and body composition predict response to PD-1 inhibitors-based combination treatment in metastatic gastric cancer.

Creatinine-to-cystatin C ratio (CCR) and body composition (BC) parameters have emerged as significant prognostic factors in cancer patients. However, the potential effects of CCR in gastric cancer (GC) remains to be elucidated. This multi-center retrospective study explored the predictive and prognostic value of CCR and BC-parameters in patients with metastatic GC receiving PD-1 inhibitors-based combination therapy. One hundred and thirteen GC patients undergoing PD-1 inhibitors-based combination therapy were enrolled at three academic medical centers from January 2021 to July 2023.A deep-learning platform based on U-Net was developed to automatically segment skeletal muscle index (SMI), subcutaneous adipose tissue index (SATI) and visceral adipose tissue index (VATI). Patients were divided into two groups based on the median of CCR or the upper tertile of BC-parameters. Logistic and Cox regression analysis were used to determine the effect of CCR and BC-parameters in predicting response rates and survival rates. The CCR was positively correlated with SMI (r=0.43; P<0.001), but not with SATI or VATI (P>0.05). Multivariable logistic analysis identified that both low CCR (OR=0.423, P=0.066 for ORR; OR=0.026, P=0.005 for DCR) and low SATI (OR=0.270, P=0.020 for ORR; OR=0.149, P=0.056 for DCR) were independently associated with worse objective response rate (ORR) and disease control rate (DCR). Patients with low CCR or low SATI had significantly lower 8-month progression-free survival (PFS) rate and 16-month overall survival (OS) rate than those with high CCR (PFS rate, 37.6% vs. 55.1%, P=0.011; OS rate, 19.4% vs. 44.9%, P=0.002) or those with high SATI (PFS rate, 37.2% vs. 53.8%, P=0.035; OS rate, 8.0% vs. 36.0%, P<0.001). Multivariate Cox analysis showed that low CCR (HR=2.395, 95% CI: 1.234-4.648, P=0.010 for PFS rate; HR=2.528, 95% CI: 1.317-4.854, P=0.005 for OS rate) and low SATI (HR=2.188, 95% CI: 1.050-4.560, P=0.037 for PFS rate; HR=2.818, 95% CI: 1.381-5.752, P=0.004 for OS rate) were both independent prognostic factors of poor 8-month PFS rate and 16-month OS rate. A nomogram based on CCR and BC-parameters showed a good performance in predicting the 12- and 16-month OS, with a concordance index of 0.756 (95% CI, 0.722-0.789). Low pre-treatment CCR and SATI were independently associated with lower response rates and worse survival in patients with metastatic GC receiving PD-1 inhibitors-based combination therapy.