Perspectives on label-free microscopy of heterogeneous and dynamic biological systems.

Significance: Advancements in label-free microscopy could provide real-time, non-invasive imaging with unique sources of contrast and automated standardized analysis to characterize heterogeneous and dynamic biological processes. These tools would overcome challenges with widely used methods that are destructive (e.g., histology, flow cytometry) or lack cellular resolution (e.g., plate-based assays, whole animal bioluminescence imaging). Aim: This perspective aims to (1) justify the need for label-free microscopy to track heterogeneous cellular functions over time and space within unperturbed systems and (2) recommend improvements regarding instrumentation, image analysis, and image interpretation to address these needs. Approach: Three key research areas (cancer research, autoimmune disease, and tissue and cell engineering) are considered to support the need for label-free microscopy to characterize heterogeneity and dynamics within biological systems. Based on the strengths (e.g., multiple sources of molecular contrast, non-invasive monitoring) and weaknesses (e.g., imaging depth, image interpretation) of several label-free microscopy modalities, improvements for future imaging systems are recommended. Conclusion: Improvements in instrumentation including strategies that increase resolution and imaging speed, standardization and centralization of image analysis tools, and robust data validation and interpretation will expand the applications of label-free microscopy to study heterogeneous and dynamic biological systems.

Enhancing Cardiovascular Risk Stratification: Radiomics of Coronary Plaque and Perivascular Adipose Tissue - Current Insights and Future Perspectives.

Radiomics, the quantitative extraction and mining of features from radiological images, has recently emerged as a promising source of non-invasive image-based cardiovascular biomarkers, potentially revolutionizing diagnostics and risk assessment. This review explores its application within coronary plaques and pericoronary adipose tissue, particularly focusing on plaque characterization and cardiac events prediction. By shedding light on the current state-of-the-art, achievements, and prospective avenues, this review contributes to a deeper understanding of the evolving landscape of radiomics in the context of coronary arteries. Finally, open challenges and existing gaps are emphasized to underscore the need for future efforts aimed at ensuring the robustness and reliability of radiomics studies, facilitating their clinical translation.

PlasmidHunter: accurate and fast prediction of plasmid sequences using gene content profile and machine learning.

Plasmids are extrachromosomal DNA found in microorganisms. They often carry beneficial genes that help bacteria adapt to harsh conditions. Plasmids are also important tools in genetic engineering, gene therapy, and drug production. However, it can be difficult to identify plasmid sequences from chromosomal sequences in genomic and metagenomic data. Here, we have developed a new tool called PlasmidHunter, which uses machine learning to predict plasmid sequences based on gene content profile. PlasmidHunter can achieve high accuracies (up to 97.6%) and high speeds in benchmark tests including both simulated contigs and real metagenomic plasmidome data, outperforming other existing tools.

AttABseq: an attention-based deep learning prediction method for antigen-antibody binding affinity changes based on protein sequences.

The optimization of therapeutic antibodies through traditional techniques, such as candidate screening via hybridoma or phage display, is resource-intensive and time-consuming. In recent years, computational and artificial intelligence-based methods have been actively developed to accelerate and improve the development of therapeutic antibodies. In this study, we developed an end-to-end sequence-based deep learning model, termed AttABseq, for the predictions of the antigen-antibody binding affinity changes connected with antibody mutations. AttABseq is a highly efficient and generic attention-based model by utilizing diverse antigen-antibody complex sequences as the input to predict the binding affinity changes of residue mutations. The assessment on the three benchmark datasets illustrates that AttABseq is 120% more accurate than other sequence-based models in terms of the Pearson correlation coefficient between the predicted and experimental binding affinity changes. Moreover, AttABseq also either outperforms or competes favorably with the structure-based approaches. Furthermore, AttABseq consistently demonstrates robust predictive capabilities across a diverse array of conditions, underscoring its remarkable capacity for generalization across a wide spectrum of antigen-antibody complexes. It imposes no constraints on the quantity of altered residues, rendering it particularly applicable in scenarios where crystallographic structures remain unavailable. The attention-based interpretability analysis indicates that the causal effects of point mutations on antibody-antigen binding affinity changes can be visualized at the residue level, which might assist automated antibody sequence optimization. We believe that AttABseq provides a fiercely competitive answer to therapeutic antibody optimization.

Reasoning with large language models for medical question answering.

OBJECTIVES: To investigate approaches of reasoning with large language models (LLMs) and to propose a new prompting approach, ensemble reasoning, to improve medical question answering performance with refined reasoning and reduced inconsistency. MATERIALS AND METHODS: We used multiple choice questions from the USMLE Sample Exam question files on 2 closed-source commercial and 1 open-source clinical LLM to evaluate our proposed approach ensemble reasoning. RESULTS: On GPT-3.5 turbo and Med42-70B, our proposed ensemble reasoning approach outperformed zero-shot chain-of-thought with self-consistency on Steps 1, 2, and 3 questions (+3.44%, +4.00%, and +2.54%) and (2.3%, 5.00%, and 4.15%), respectively. With GPT-4 turbo, there were mixed results with ensemble reasoning again outperforming zero-shot chain-of-thought with self-consistency on Step 1 questions (+1.15%). In all cases, the results demonstrated improved consistency of responses with our approach. A qualitative analysis of the reasoning from the model demonstrated that the ensemble reasoning approach produces correct and helpful reasoning. CONCLUSION: The proposed iterative ensemble reasoning has the potential to improve the performance of LLMs in medical question answering tasks, particularly with the less powerful LLMs like GPT-3.5 turbo and Med42-70B, which may suggest that this is a promising approach for LLMs with lower capabilities. Additionally, the findings show that our approach helps to refine the reasoning generated by the LLM and thereby improve consistency even with the more powerful GPT-4 turbo. We also identify the potential and need for human-artificial intelligence teaming to improve the reasoning beyond the limits of the model.

Public comfort with the use of ChatGPT and expectations for healthcare.

OBJECTIVES: To examine whether comfort with the use of ChatGPT in society differs from comfort with other uses of AI in society and to identify whether this comfort and other patient characteristics such as trust, privacy concerns, respect, and tech-savviness are associated with expected benefit of the use of ChatGPT for improving health. MATERIALS AND METHODS: We analyzed an original survey of U.S. adults using the NORC AmeriSpeak Panel (n = 1787). We conducted paired t-tests to assess differences in comfort with AI applications. We conducted weighted univariable regression and 2 weighted logistic regression models to identify predictors of expected benefit with and without accounting for trust in the health system. RESULTS: Comfort with the use of ChatGPT in society is relatively low and different from other, common uses of AI. Comfort was highly associated with expecting benefit. Other statistically significant factors in multivariable analysis (not including system trust) included feeling respected and low privacy concerns. Females, younger adults, and those with higher levels of education were less likely to expect benefits in models with and without system trust, which was positively associated with expecting benefits (P = 1.6 × 10-11). Tech-savviness was not associated with the outcome. DISCUSSION: Understanding the impact of large language models (LLMs) from the patient perspective is critical to ensuring that expectations align with performance as a form of calibrated trust that acknowledges the dynamic nature of trust. CONCLUSION: Including measures of system trust in evaluating LLMs could capture a range of issues critical for ensuring patient acceptance of this technological innovation.

Precision medicine: Externally validated explainable AI support tool for predicting sustainability of infliximab and vedolizumab in ulcerative colitis.

OBJECTIVE: Drug sustainability (DS), a surrogate marker for drug efficacy, is important, especially when aiming for precision medicine. However, it lacks reliable prediction methods. AIMS: To develop and externally validate a web-based artificial intelligence(AI)-derived tool for predicting DS of infliximab and vedolizumab in patients with moderate-to-severe Ulcerative Colitis (UC). METHODS: Data from three Israeli centers included infliximab or vedolizumab patients treated for >54 weeks. Sustainability meant no corticosteroids, hospitalizations or surgeries. Machine learning techniques predicted >54-week and overall DS using baseline clinical data. RESULTS: The model was developed using data from 246 patients from Rabin Medical Center and externally validated on 67 patients from Rambam Health Care Campus and Sheba Medical Center. No significant difference in DS was observed across the datasets. Most patients were biologic-naïve and primarily treated with vedolizumab. The model performed well, with an area under the ROC curve of 0.86, and showed good accuracy (65.5 %-76.9 %) across the test sets. CONCLUSIONS: The study introduces a novel, AI-based tool for predicting >54-week DS of infliximab and vedolizumab in moderate-to-severe UC, using baseline parameters. This can aid clinical decision-making in the framework of precision medicine, promising to optimize disease management while maintaining physician autonomy.

Real-World Application of Artificial Intelligence for Detecting Pathologic Gastric Atypia and Neoplastic Lesions.

PURPOSE: Results of initial endoscopic biopsy of gastric lesions often differ from those of the final pathological diagnosis. We evaluated whether an artificial intelligence-based gastric lesion detection and diagnostic system, ENdoscopy as AI-powered Device Computer Aided Diagnosis for Gastroscopy (ENAD CAD-G), could reduce this discrepancy. MATERIALS AND METHODS: We retrospectively collected 24,948 endoscopic images of early gastric cancers (EGCs), dysplasia, and benign lesions from 9,892 patients who underwent esophagogastroduodenoscopy between 2011 and 2021. The diagnostic performance of ENAD CAD-G was evaluated using the following real-world datasets: patients referred from community clinics with initial biopsy results of atypia (n=154), participants who underwent endoscopic resection for neoplasms (Internal video set, n=140), and participants who underwent endoscopy for screening or suspicion of gastric neoplasm referred from community clinics (External video set, n=296). RESULTS: ENAD CAD-G classified the referred gastric lesions of atypia into EGC (accuracy, 82.47%; 95% confidence interval [CI], 76.46%-88.47%), dysplasia (88.31%; 83.24%-93.39%), and benign lesions (83.12%; 77.20%-89.03%). In the Internal video set, ENAD CAD-G identified dysplasia and EGC with diagnostic accuracies of 88.57% (95% CI, 83.30%-93.84%) and 91.43% (86.79%-96.07%), respectively, compared with an accuracy of 60.71% (52.62%-68.80%) for the initial biopsy results (P<0.001). In the External video set, ENAD CAD-G classified EGC, dysplasia, and benign lesions with diagnostic accuracies of 87.50% (83.73%-91.27%), 90.54% (87.21%-93.87%), and 88.85% (85.27%-92.44%), respectively. CONCLUSIONS: ENAD CAD-G is superior to initial biopsy for the detection and diagnosis of gastric lesions that require endoscopic resection. ENAD CAD-G can assist community endoscopists in identifying gastric lesions that require endoscopic resection.

Image quality of whole-body diffusion MR images comparing deep-learning accelerated and conventional sequences.

OBJECTIVES: To compare the image quality of deep learning accelerated whole-body (WB) with conventional diffusion sequences. METHODS: Fifty consecutive patients with bone marrow cancer underwent WB-MRI. Two experts compared axial b900 s/mm2 and the corresponding maximum intensity projections (MIP) of deep resolve boost (DRB) accelerated diffusion-weighted imaging (DWI) sequences (time of acquisition: 6:42 min) against conventional sequences (time of acquisition: 14 min). Readers assessed paired images for noise, artefacts, signal fat suppression, and lesion conspicuity using Likert scales, also expressing their overall subjective preference. Signal-to-noise and contrast-to-noise ratios (SNR and CNR) and the apparent diffusion coefficient (ADC) values of normal tissues and cancer lesions were statistically compared. RESULTS: Overall, radiologists preferred either axial DRB b900 and/or corresponding MIP images in almost 80% of the patients, particularly in patients with a high body-mass index (BMI > 25 kg/m2). In qualitative assessments, axial DRB images were preferred (preferred/strongly preferred) in 56-100% of cases, whereas DRB MIP images were favoured in 52-96% of cases. DRB-SNR/CNR was higher in all normal tissues (p < 0.05). For cancer lesions, the DRB-SNR was higher (p < 0.001), but the CNR was not different. DRB-ADC values were significantly higher for the brain and psoas muscles, but not for cancer lesions (mean difference: + 53 µm2/s). Inter-class correlation coefficient analysis showed good to excellent agreement (95% CI 0.75-0.93). CONCLUSION: DRB sequences produce higher-quality axial DWI, resulting in improved MIPs and significantly reduced acquisition times. However, differences in the ADC values of normal tissues need to be considered. CLINICAL RELEVANCE STATEMENT: Deep learning accelerated diffusion sequences produce high-quality axial images and MIP at reduced acquisition times. This advancement could enable the increased adoption of Whole Body-MRI for the evaluation of patients with bone marrow cancer. KEY POINTS: Deep learning reconstruction enables a more than 50% reduction in acquisition time for WB diffusion sequences. DRB images were preferred by radiologists in almost 80% of cases due to fewer artefacts, improved background signal suppression, higher signal-to-noise ratio, and increased lesion conspicuity in patients with higher body mass index. Cancer lesion diffusivity from DRB images was not different from conventional sequences.

Validation of Artificial Intelligence in the Classification of Adolescent Idiopathic Scoliosis and the Compairment to Clinical Manual Handling.

OBJECTIVE: The accurate measurement of Cobb angles is crucial for the effective clinical management of patients with adolescent idiopathic scoliosis (AIS). The Lenke classification system plays a pivotal role in determining the appropriate fusion levels for treatment planning. However, the presence of interobserver variability and time-intensive procedures presents challenges for clinicians. The purpose of this study is to compare the measurement accuracy of our developed artificial intelligence measurement system for Cobb angles and Lenke classification in AIS patients with manual measurements to validate its feasibility. METHODS: An artificial intelligence (AI) system measured the Cobb angle of AIS patients using convolutional neural networks, which identified the vertebral boundaries and sequences, recognized the upper and lower end vertebras, and estimated the Cobb angles of the proximal thoracic, main thoracic, and thoracolumbar/lumbar curves sequentially. Accordingly, the Lenke classifications of scoliosis were divided by oscillogram and defined by the AI system. Furthermore, a man-machine comparison (n = 300) was conducted for senior spine surgeons (n = 2), junior spine surgeons (n = 2), and the AI system for the image measurements of proximal thoracic (PT), main thoracic (MT), thoracolumbar/lumbar (TL/L), thoracic sagittal profile T5-T12, bending views PT, bending views MT, bending views TL/L, the Lenke classification system, the lumbar modifier, and sagittal thoracic alignment. RESULTS: In the AI system, the calculation time for each patient's data was 0.2 s, while the measurement time for each surgeon was 23.6 min. The AI system showed high accuracy in the recognition of the Lenke classification and had high reliability compared to senior doctors (ICC 0.962). CONCLUSION: The AI system has high reliability for the Lenke classification and is a potential auxiliary tool for spinal surgeons.

Artificial intelligence for detection of prostate cancer in biopsies during active surveillance.

OBJECTIVES: To evaluate a cancer detecting artificial intelligence (AI) algorithm on serial biopsies in patients with prostate cancer on active surveillance (AS). PATIENTS AND METHODS: A total of 180 patients in the Prostate Cancer Research International Active Surveillance (PRIAS) cohort were prospectively monitored using pre-defined criteria. Diagnostic and re-biopsy slides from 2011 to 2020 (n = 4744) were scanned and analysed by an in-house AI-based cancer detection algorithm. The algorithm was analysed for sensitivity, specificity, and for accuracy to predict need for active treatment. Prognostic properties of cancer size, prostate-specific antigen (PSA) level and PSA density at diagnosis were evaluated. RESULTS: The sensitivity and specificity of the AI algorithm was 0.96 and 0.73, respectively, for correct detection of cancer areas. Original pathology report diagnosis was used as the reference method. The area of cancer estimated by the pathologists correlated highly with the AI detected cancer size (r = 0.83). By using the AI algorithm, 63% of the slides would not need to be read by a pathologist as they were classed as benign, at the risk of missing 0.55% slides containing cancer. Biopsy cancer content and PSA density at diagnosis were found to be prognostic of whether the patient stayed on AS or was discontinued for active treatment. CONCLUSION: The AI-based biopsy cancer detection algorithm could be used to reduce the pathologists' workload in an AS cohort. The detected cancer amount correlated well with the cancer length measured by the pathologist and the algorithm performed well in finding even small areas of cancer. To our knowledge, this is the first report on an AI-based algorithm in digital pathology used to detect cancer in a cohort of patients on AS.

Leveraging Quantitative Systems Pharmacology and Artificial Intelligence to advance treatment of Chronic Kidney Disease Mineral Bone Disorder.

Chronic kidney disease mineral bone disorder (CKD-MBD) is a complex clinical syndrome responsible for the accelerated cardiovascular mortality seen in individuals afflicted with CKD. Current approaches to therapy have failed to improve clinical outcomes adequately, likely due to targeting surrogate biochemical parameters as articulated by the guideline developer, KDIGO (Kidney Disease: Improving Global Outcomes). We hypothesized that using a Systems Biology Approach combining machine learning with mathematical modeling, we could test a novel approach to therapy targeting the abnormal movement of mineral out of bone and into soft tissue that is characteristic of CKD-MBD. The mathematical model describes the movement of calcium and phosphate between body compartments in response to standard therapeutic agents. The machine learning technique we applied is Reinforcement Learning (RL). We compared calcium, phosphate, PTH, and mineral movement out of bone and into soft tissue under four scenarios: standard approach (KDIGO), achievement of KDIGO guidelines using RL (RLKDIGO), targeting abnormal mineral flux (RLFLUX), and combining achievement of KDIGO guidelines with minimization of abnormal mineral flux (RLKDIGOFLUX). We demonstrate through simulations that explicitly targeting abnormal mineral flux significantly decreases abnormal mineral movement compared to standard approach while achieving acceptable biochemical outcomes. These investigations highlight the limitations of current therapeutic targets, primarily secondary hyperparathyroidism, and emphasize the central role of deranged phosphate homeostasis in the genesis of the CKD-MBD syndrome.

[Application of artificial intelligence in glaucoma. Part 1. Neural networks and deep learning in glaucoma screening and diagnosis]. / Primenenie metodov iskusstvennogo intellekta pri glaukome. Chast' 1. Neiroseti i glubokoe obuchenie v skrininge i diagnostike glaukomy.

This article reviews literature on the use of artificial intelligence (AI) for screening, diagnosis, monitoring and treatment of glaucoma. The first part of the review provides information how AI methods improve the effectiveness of glaucoma screening, presents the technologies using deep learning, including neural networks, for the analysis of big data obtained by methods of ocular imaging (fundus imaging, optical coherence tomography of the anterior and posterior eye segments, digital gonioscopy, ultrasound biomicroscopy, etc.), including a multimodal approach. The results found in the reviewed literature are contradictory, indicating that improvement of the AI models requires further research and a standardized approach. The use of neural networks for timely detection of glaucoma based on multimodal imaging will reduce the risk of blindness associated with glaucoma.

Virtual Labs and Designer Bugs - Generative AI, Synthetic Biology and National Security.

AI technologies can pose a major national security concern. AI programs could be used to develop chemical and biological agents which circumvent existing protective measures or medical treatments, or to design pathogens with capabilities they do not naturally possess (gain-of-function research). Although Australia has a strong legislative framework relating to research into genetically modified organisms, the framework requires the interaction of more than 10 different government departments, universities and funding agencies. Further, there are few guidelines about the responsible use of AI in biological research where existing laws and policies do not apply to research that is conducted "virtually", even where that research may have national security implications. This article explores these under-scrutinised concepts in Australia's biological security frameworks.

Innovative and practical tools for monitoring and assessing biodiversity status and impacts of multiple human pressures in marine systems.

Human activities at sea can produce pressures and cumulative effects on ecosystem components that need to be monitored and assessed in a cost-effective manner. Five Horizon European projects have joined forces to collaboratively increase our knowledge and skills to monitor and assess the ocean in an innovative way, assisting managers and policy-makers in taking decisions to maintain sustainable activities at sea. Here, we present and discuss the status of some methods revised during a summer school, aiming at better management of coasts and seas. We include novel methods to monitor the coastal and ocean waters (e.g. environmental DNA, drones, imaging and artificial intelligence, climate modelling and spatial planning) and innovative tools to assess the status (e.g. cumulative impacts assessment, multiple pressures, Nested Environmental status Assessment Tool (NEAT), ecosystem services assessment or a new unifying approach). As a concluding remark, some of the most important challenges ahead are assessing the pros and cons of novel methods, comparing them with benchmark technologies and integrating these into long-standing time series for data continuity. This requires transition periods and careful planning, which can be covered through an intense collaboration of current and future European projects on marine biodiversity and ecosystem health.

Invasive fractional-flow-reserve prediction by coronary CT angiography using artificial intelligence vs. computational fluid dynamics software in intermediate-grade stenosis.

Coronary computed angiography (CCTA) with non-invasive fractional flow reserve (FFR) calculates lesion-specific ischemia when compared with invasive FFR and can be considered for patients with stable chest pain and intermediate-grade stenoses according to recent guidelines. The objective of this study was to compare a new CCTA-based artificial-intelligence deep-learning model for FFR prediction (FFRAI) to computational fluid dynamics CT-derived FFR (FFRCT) in patients with intermediate-grade coronary stenoses with FFR as reference standard. The FFRAI model was trained with curved multiplanar-reconstruction CCTA images of 500 stenotic vessels in 413 patients, using FFR measurements as the ground truth. We included 37 patients with 39 intermediate-grade stenoses on CCTA and invasive coronary angiography, and with FFRCT and FFR measurements in this retrospective proof of concept study. FFRAI was compared with FFRCT regarding the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy for predicting FFR ≤ 0.80. Sensitivity, specificity, PPV, NPV, and diagnostic accuracy of FFRAI in predicting FFR ≤ 0.80 were 91% (10/11), 82% (23/28), 67% (10/15), 96% (23/24), and 85% (33/39), respectively. Corresponding values for FFRCT were 82% (9/11), 75% (21/28), 56% (9/16), 91% (21/23), and 77% (30/39), respectively. Diagnostic accuracy did not differ significantly between FFRAI and FFRCT (p = 0.12). FFRAI performed similarly to FFRCT for predicting intermediate-grade coronary stenoses with FFR ≤ 0.80. These findings suggest FFRAI as a potential non-invasive imaging tool for guiding therapeutic management in these stenoses.

Beyond black-box models: explainable AI for embryo ploidy prediction and patient-centric consultation.

PURPOSE: To determine if an explainable artificial intelligence (XAI) model enhances the accuracy and transparency of predicting embryo ploidy status based on embryonic characteristics and clinical data. METHODS: This retrospective study utilized a dataset of 1908 blastocyst embryos. The dataset includes ploidy status, morphokinetic features, morphology grades, and 11 clinical variables. Six machine learning (ML) models including Random Forest (RF), Linear Discriminant Analysis (LDA), Logistic Regression (LR), Support Vector Machine (SVM), AdaBoost (ADA), and Light Gradient-Boosting Machine (LGBM) were trained to predict ploidy status probabilities across three distinct datasets: high-grade embryos (HGE, n = 1107), low-grade embryos (LGE, n = 364), and all-grade embryos (AGE, n = 1471). The model's performance was interpreted using XAI, including SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) techniques. RESULTS: The mean maternal age was 38.5 ± 3.85 years. The Random Forest (RF) model exhibited superior performance compared to the other five ML models, achieving an accuracy of 0.749 and an AUC of 0.808 for AGE. In the external test set, the RF model achieved an accuracy of 0.714 and an AUC of 0.750 (95% CI, 0.702-0.796). SHAP's feature impact analysis highlighted that maternal age, paternal age, time to blastocyst (tB), and day 5 morphology grade significantly impacted the predictive model. In addition, LIME offered specific case-ploidy prediction probabilities, revealing the model's assigned values for each variable within a finite range. CONCLUSION: The model highlights the potential of using XAI algorithms to enhance ploidy prediction, optimize embryo selection as patient-centric consultation, and provides reliability and transparent insights into the decision-making process.

Multidisciplinary Approach to Pulmonary Embolism and the Role of the Pulmonary Embolism Response Team.

PURPOSE OF REVIEW: Acute pulmonary embolism (PE) is a leading cause of cardiovascular death and morbidity, and presents a major burden to healthcare systems. The field has seen rapid growth with development of innovative clot reduction technologies, as well as ongoing multicenter trials that may completely revolutionize care of PE patients. However, current paucity of robust clinical trials and guidelines often leave individual physicians managing patients with acute PE in a dilemma. RECENT FINDINGS: The pulmonary embolism response team (PERT) was developed as a platform to rapidly engage multiple specialists to deliver evidence-based, organized and efficient care and help address some of the gaps in knowledge. Several centers investigating outcomes following implementation of PERT have demonstrated shorter hospital and intensive-care unit stays, lower use of inferior vena cava filters, and in some instances improved mortality. Since the advent of PERT, early findings demonstrate promise with improved outcomes after implementation of PERT. Incorporation of artificial intelligence (AI) into PERT has also shown promise with more streamlined care and reducing response times. Further clinical trials are needed to examine the impact of PERT model on care delivery and clinical outcomes.

Portraying on-road CO2 concentrations using street view panoramas and ensemble learning.

A significant reduction in carbon dioxide (CO2) emissions caused by transportation is essential for attaining sustainable urban development. Carbon concentrations from road traffic in urban areas exhibit complex spatial patterns due to the impact of street configurations, mobile sources, and human activities. However, a comprehensive understanding of these patterns, which involve complex interactions, is still lacking due to the human perspective of road interface characteristics has not been taken into account. In this study, a mobile travel platform was constructed to collect both on-road navigation Street View Panoramas (OSVPs) and the corresponding CO2 concentrations. >100 thousand sample pairs that matched "street view-CO2 concentration" were obtained, covering 675.8 km of roads in Shenzhen, China. In addition, four ensemble learning (EL) models were utilized to establish nonlinear connections between the semantic and object features of streetscapes and CO2 concentrations. After performing EL fusion modeling, the predictive R2 in the test set exceeded 90 %, and the mean absolute error (MAE) was <3.2 ppm. The model was applied to Baidu Street View Panoramas (BSVPs) in Shenzhen to generate a map of average on-road CO2 with a 100 m resolution, and the Local Indicator of Spatial Association (LISA) was then used to identify high CO2 intensity spatial clusters. Additionally, the Light Gradient Boost-SHapley Additive exPlanation (LGB-SHAP) analysis revealed that vertically planted trees can reduce CO2 emissions from on-road sources. Moreover, the factors that affect on-road CO2 exhibit interaction and threshold effects. Street View Panoramas (SVPs) and Artificial Intelligence (AI) were adopted here to enhance the spatial measurement of on-road CO2 concentrations and the understanding of driving factors. Our approach facilitates the assessment and design of low-emission transportation in urban areas, which is critical for promoting sustainable traffic development.