Manual Annotation Research Articles

The early dodder gets the host: decoding the coiling patterns of Cuscuta campestris with automated image processing.

We developed an in-house Python-based image analysis pipeline to investigate the movement patterns of Cuscuta. Our analysis unveiled that the coiling and circumnutation movements of Cuscuta are regulated by its intrinsic circadian rhythm. Cuscuta spp., commonly known as dodders, are rootless and leafless stem parasitic plants. Upon germination, Cuscuta starts rotating immediately in a counterclockwise direction (circumnutation) to locate a host plant, creating a seamless vascular connection to steal water and nutrients from its host. In this study, our aim was to elucidate the dynamics of the coiling patterns of Cuscuta, which is an essential step for successful parasitism. Using time-lapse photography, we recorded the circumnutation and coiling movements of C. campestris at different inoculation times on non-living hosts. Subsequent image analyses were facilitated through an in-house Python-based image processing pipeline to detect coiling locations, angles, initiation and completion times, and duration of coiling stages in between. The study revealed that the coiling efficacy of C. campestris varied with the inoculation time of day, showing higher success and faster initiation in morning than in evening. These observations suggest that Cuscuta, despite lacking leaves and a developed chloroplast, can discern photoperiod changes, significantly determining its parasitic efficiency. The automated image analysis results confirmed the reliability of our Python pipeline by aligning closely with manual annotations. This study provides significant insights into the parasitic strategies of C. campestris and demonstrates the potential of integrating computational image analysis in plant biology for exploring complex plant behaviors. Furthermore, this method provides an efficient tool for investigating plant movement dynamics, laying the foundation for future studies on mitigating the economic impacts of parasitic plants.

A framework for automated supraglacial lake detection and depth retrieval in ICESat-2 photon data across the Greenland and Antarctic ice sheets

Abstract. Water depths of supraglacial lakes on the ice sheets are difficult to monitor continuously due the lakes' ephemeral nature and inaccessible locations. Supraglacial lakes have been linked to ice shelf collapse in Antarctica and accelerated flow of grounded ice in Greenland. However, the impact of supraglacial lakes on ice dynamics has not been quantified accurately enough to predict their contribution to future mass loss and sea level rise. This is largely because ice-sheet-wide assessments of meltwater volumes rely on models that are poorly constrained due to a lack of accurate depth measurements. Various recent case studies have demonstrated that accurate supraglacial lake depths can be obtained from NASA's Ice, Cloud and land Elevation Satellite (ICESat-2) ATL03 photon-level data product. ATL03 comprises hundreds of terabytes of unstructured point cloud data, which has made it challenging to use this bathymetric capability at scale. Here, we present two new algorithms – Flat Lake and Underlying Ice Detection (FLUID) and Surface Removal and Robust Fit (SuRRF) – which together provide a fully automated and scalable method for lake detection and along-track depth determination from ATL03 data and establish a framework for its large-scale implementation using distributed high-throughput computing. We report FLUID–SuRRF algorithm performance over two regions known to have significant surface melt – central West Greenland and the Amery Ice Shelf catchment in East Antarctica – during two melt seasons. FLUID–SuRRF reveals a total of 1249 ICESat-2 lake segments up to 25 m deep, with more water during higher-melt years. In the absence of ground-truth data, manual annotation of test data suggests that our method reliably detects melt lakes along ICESat-2's ground tracks whenever the lake bed is visible or partially visible and estimates water depths with a mean absolute error <0.27 m. These results imply that our proposed framework has the potential to generate a comprehensive data product of accurate meltwater depths across both ice sheets.

HPB SO33 - Transcriptomic characterisation of the diabetogenic pancreatic ductal adenocarcinoma phenotype

Abstract Background New-onset diabetes mellitus (NOD) can manifest as a consequence of pancreatic ductal adenocarcinoma (PDAC), although not all PDAC patients develop NOD. For many patients, hyperglycaemia begins 3 years prior to PDAC diagnosis, with 30% of PDAC patients experiencing NOD by the time their cancer is diagnosed. The mechanisms underlying the onset of diabetes mellitus (DM) in PDAC and its consequences for disease development, detection and treatment are not well understood. The aim of this study was to characterise the transcriptomic differences between PDAC tumours that present with NOD compared to those that do not. Method Eight tumour samples from patients undergoing surgery for PDAC (4 without DM (HbA1c <35), 4 with NOD (HbA1c >65, within 3 years of cancer diagnosis) underwent single cell-RNA sequencing (scRNAseq) after taking intra-operative biopsies using Chromium 10X Genomics technology. Unique cell clusters were identified using CellRanger and cell types annotated using scATOMIC and singleR datasets. Manual annotation was undertaken when results were not agreeable after consulting the literature. Differential expression between groups was performed using DESeq2. Significant genes were identified as having a log2 fold-change >1.5 and p<0.05. Gene set enrichment was performed using Reactome. Analysis was conducted in RStudio. Results Single cell suspensions were created with a mean cell count of 1.2 million cells and a mean viability of 90%. PDAC cells and the full complement of tumour microenvironmental cells were recovered. Significant differential expression was seen in the PDAC cells themselves, CD4+ cells, macrophages, mast cells and fibroblasts. Cancer associated fibroblasts (CAFs) seemed not to differ. PDAC cells with NOD seemed to be enriched in MYC oncogenes whereas those without DM seemed to be enriched in KRAS mutations. There was a mixture of “basal” and “classical” transcriptional subtypes between the two groups. Conclusion Differences between diabetogenic PDAC and non-diabetogenic PDAC have been observed at the transcriptional level. Diabetogenic tumours seem to be promoted by MYC oncogenes and be more inflammatory than non-diabetogenic tumours. These findings require validation which will be done on independent samples using spatial transcriptomics to allow endocrine-exocrine crosstalk to be determined. Understanding the molecular pathways in diabetogenic pancreatic tumours could enhance future strategies to detect PDAC earlier and to target diabetes-associated vulnerabilities for therapeutic purposes.

Application of deep learning algorithms for identifying deterioration in the ushnisha (Head Bun) of the Leshan Giant Buddha

AbstractThe Leshan Giant Buddha’s ushnisha (Head Bun) has suffered from loss of lime plaster, cracks, and biological damage, compromising its structural integrity and reducing the effectiveness of the drainage system in the Buddha's head. The infiltration of moisture has led to water damage within the statue, significantly accelerating its weathering. This situation urgently requires protection and reinforcement measures. Detecting deterioration in the ushnisha is a crucial step in the preservation process. In this study, we utilized two deep learning models for pixel-level semantic segmentation of the damage. Due to the small size of the cracks, a weighted loss function was applied to improve both the training speed of the model and the efficiency of crack identification. This weighting strategy proved effective for both models. The weighted K-Net model achieved a mean accuracy (mAcc) of 90.23% and a mean intersection-over-union (mIoU) of 69.55%, with a damage segmentation speed of 7 images per second, which is 1309 times faster than manual segmentation. By applying the trained deep learning models to re-examine the ushnisha, we successfully identified damage that had been overlooked during manual annotation. Using the model’s enhanced results, we conducted a comprehensive quantification of the damage across all ushnisha and identified the most severely affected areas. Additionally, we performed a model interpretability analysis to explain the decision-making process and principles of the deep learning models. This research provides significant practical value for detecting and quantifying damage in the Leshan Giant Buddha.

Abstract 4142799: Central role of LIM domain binding 3 (LDB3) protein in Tetralogy of Fallot assessed by single-nuclei multiome profiling of paediatric right ventricular samples

Background and aims: Tetralogy of Fallot (TOF; EC: 1565/2019) is the most common cyanotic congenital heart defect with a world-wide incidence of 0.02–0.04%. The pathophysiological mechanisms driven by variable genetic background are poorly understood. We applied the single-nuclei multiome profiling, composed of ATAC and RNA-seq from the same nuclei, of right ventricular samples of children with TOF. The final aim was to define cell types and characterize gene regulation states in different cell clusters including activation and maturation. Methods: Right ventricular samples were prospectively gathered during routinely performed cardiac surgery for TOF correction (EC: 1565/2019). In total, six paediatric patient samples were processed, and on average ~1,500 nuclei were sequenced per sample. Data was analysed using Scanpy to cluster and annotate cells, pycisTopic to identify cell stats and cis-regulatory topics, and SCENIC+ for enhancer-driven gene Regulatory network (eRegulon) construction. Results: Through manual annotation, using published markers, we identified cell types including cardiomyocyte, fibroblast, endothelial, pericyte, immune and neuronal cells. From this, we observed cell type specific expression, with overexpression of decorin in fibroblasts, PTPRC in immune cells, and neurexin genes in neuronal cells found. Differential expression analysis highlighted patient specific patterns of expression, of which gene set enrichment showed enrichment for genes associated with actomyosin structure organisation, dilated cardiomyopathy and cardiac muscle contraction. Integration of the gene expression data with the ATAC-seq data, across all patients, allowed for the construction of eRegulons networks. In detail, in the cardiomyocyte-annotated clusters, we identified enrichment of MEF2, MEIS2 and EPAS1 family enhancers with a central role of LIM domain binding 3 protein (LDB3, associated with cytoskeletal actin and muscle alpha-actin binding) (Fig. 1). Conclusion: Single nuclei multiome profiling of paediatric right ventricular TOF samples identified several patient specific patterns of expression and regulatory networks, related to the development of cyanotic congenital heart disease, and sheds light to the potential drivers of TOF.

Public opinion mining in social media about Ethiopian broadcasts using deep learning.

Now adays people express and share their opinions on various events on the internet thanks to social media. Opinion mining is the process of interpreting user-generated opinion data on social media. Aside from its lack of resources in opinion-mining tasks, Amharic presents numerous difficulties because of its complex structure and variety of dialects. Analyzing every comment written in Amharic is a challenging task. Significant advancements in opinion mining have been achieved using deep learning. An opinion-mining model was used in this study to classify user comments written in Amharic as positive or negative. The domains that we focus on in this study are YouTube and Facebook. From the Ethiopian broadcasts YouTube and Facebook official pages, we gathered 11,872 unstructured data for this study using www.exportcomment.com , and Facebook page tools. Text preprocessing and feature extraction techniques were used, in addition to manual annotation by linguistic specialists. The dataset was prepared for the experiment after annotation, preprocessing, and representation. LSTM, GRU, BiGRU, BiLSTM, and a hybrid of CNN with BiLSTM classifiers from the TensorFlow Keras deep learning library were used to train the model using the dataset, which was split using the 80/20 train-test method, which proved effective for classification problems. Finally, we achieved of 94.27%, 95.20%, 95.49%, 95.62%, and 96.08% using GRU, BiGRU, LSTM, BiLSTM, and CNN with BiLSTM, respectively, in word2vec embedding model.

Abstract Su806: Enhancing Cardiac Arrest Registries and Research – A Framework for Handling Defibrillator Data

Background: Cardiac arrest is the most time-sensitive medical condition, and survival probability is highly dependent on early, high-quality cardiopulmonary resuscitation (CPR) measures. Resuscitation registries are established tools for continuous quality improvement and serve as a source for research. Yet they lack comprehensive information on the quality of CPR measures themselves—such as the depth, rate, and release of chest compressions—and are limited in the temporal resolution of these measures. However, data on both are contained in the recordings of defibrillators. Aims: To develop a framework for healthcare providers to gain immediate case-specific feedback, reduce documentation efforts and make fine-grained information accessible for research. Approach: An open-source Python module named CPRDAT (Kern et al. Zenodo, https://doi.org/10.5281/zenodo.10057964) was designed. This module is capable of importing and extracting data from various defibrillators as well as data recorded by VitalRecorder ( VitalDB , Department of Anesthesiology and Pain Medicine, Seoul National University College of Medicine, Seoul, Korea) into a uniform data object. Methods were implemented to automatically process, analyze, and aggregate these time-series of physiological signals, with particular attention to CPR. To reduce manual annotation workload, classification algorithms were developed. Plotting routines were developed to display the signals interactively in web applications, allowing healthcare providers to interact with the data through zooming, scaling, shifting, concatenating, merging and labelling. Results: The module was implemented into the German Resuscitation Registry, offering users a manufacturer-agnostic, data-driven, case-specific debriefing tool. Extracted information on CPR measures is stored in the registry. The framework, able to manage multiple time-series data sources, was further implemented in other settings, such as the analysis of laboratory data of a CPR swine model. Conclusions: The new module dramatically enhances the accessibility of physiologic time-series recordings. More fine-grained data will provide deeper insights into processes and CPR physiology. Finally, the open-source Python module abolishes barriers in research.

EcoSound-web: an open-source, online platform for ecoacoustics.

Passive acoustic monitoring of soundscapes and biodiversity produces vast amounts of audio recordings, but the management and analyses of these raw data present technical challenges. A multitude of software solutions exist, but none can fulfil all purposes required for the management, processing, navigation, and analysis of acoustic data. The field of ecoacoustics needs a software tool that is free, evolving, and accessible. We take a step in that direction and present ecoSound-web: an open-source, online platform for ecoacoustics designed and built by ecologists and software engineers. ecoSound-web can be used for storing, re-sampling, organising, analysing, and sharing soundscape recording or metadata projects. Specifically, it allows manual annotation of soniferous animals and soundscape components, automatic annotation with deep-learning models for all birds and for UK bat species, peer-reviewing annotations, analysing audio in time and frequency dimensions, computing alpha acoustic indices, and providing reference sound libraries for different taxa. We present ecoSound-web's structure and features, and describe its operation for typical use cases such as sampling bird and bat communities, using a primate call library, and the analysis of soundscape components and acoustic indices. ecoSound-web is available from: https://github.com/ecomontec/ecoSound-web.

Automated acute pain prediction in domestic goats using deep learning-based models on video-recordings

Facial expressions are essential in animal communication, and facial expression-based pain scales have been developed for different species. Automated pain recognition offers a valid alternative to manual annotation with growing evidence across species. This study applied machine learning (ML) methods, using a pre-trained VGG-16 base and a Support Vector Machine classifier to automate pain recognition in caprine patients in hospital settings, evaluating different frame extraction rates and validation techniques. The study included goats of different breed, age, sex, and varying medical conditions presented to the University of Florida’s Large Animal Hospital. Painful status was determined using the UNESP-Botucatu Goat Acute Pain Scale. The final dataset comprised images from 40 goats (20 painful, 20 non-painful), with 2,253 ‘non-painful’ and 3,154 ‘painful’ images at 1 frame per second (FPS) extraction rate and 7,630 ‘non-painful’ and 9,071 ‘painful’ images at 3 FPS. Images were used to train deep learning-based models with different approaches. The model input was raw images, and pain presence was the target attribute (model output). For the single train-test split and 5-fold cross-validation, the models achieved approximately 80% accuracy, while the subject-wise 10-fold cross-validation showed mean accuracies above 60%. These findings suggest ML’s potential in goat pain assessment.

A Streamlined Workflow for Microscopy-Driven MALDI Imaging Mass Spectrometry Data Collection.

Matrix-assisted laser desorption/ionization imaging mass spectrometry (MALDI IMS) is a rapidly advancing technology for biomedical research. As spatial resolution increases, however, so do acquisition time, file size, and experimental cost, which increases the need to perform precise sampling of targeted tissue regions to optimize the biological information gleaned from an experiment and minimize wasted resources. The ability to define instrument measurement regions based on key tissue features and automatically sample these specific regions of interest (ROIs) addresses this challenge. Herein, we demonstrate a workflow using standard software that allows for direct sampling of microscopy-defined regions by MALDI IMS. Three case studies are included, highlighting different methods for defining features from common sample types─manual annotation of vasculature in human brain tissue, automated segmentation of renal functional tissue units across whole slide images using custom segmentation algorithms, and automated segmentation of dispersed HeLa cells using open-source software. Each case minimizes data acquisition from unnecessary sample regions and dramatically increases throughput while uncovering molecular heterogeneity within targeted ROIs. This workflow provides an approachable method for spatially targeted MALDI IMS driven by microscopy as part of multimodal molecular imaging studies.

Deep learning for surgical instrument recognition and segmentation in robotic-assisted surgeries: a systematic review

Applying deep learning (DL) for annotating surgical instruments in robot-assisted minimally invasive surgeries (MIS) represents a significant advancement in surgical technology. This systematic review examines 48 studies that utilize advanced DL methods and architectures. These sophisticated DL models have shown notable improvements in the precision and efficiency of detecting and segmenting surgical tools. The enhanced capabilities of these models support various clinical applications, including real-time intraoperative guidance, comprehensive postoperative evaluations, and objective assessments of surgical skills. By accurately identifying and segmenting surgical instruments in video data, DL models provide detailed feedback to surgeons, thereby improving surgical outcomes and reducing complication risks. Furthermore, the application of DL in surgical education is transformative. The review underscores the significant impact of DL on improving the accuracy of skill assessments and the overall quality of surgical training programs. However, implementing DL in surgical tool detection and segmentation faces challenges, such as the need for large, accurately annotated datasets to train these models effectively. The manual annotation process is labor-intensive and time-consuming, posing a significant bottleneck. Future research should focus on automating the detection and segmentation process and enhancing the robustness of DL models against environmental variations. Expanding the application of DL models across various surgical specialties will be essential to fully realize this technology’s potential. Integrating DL with other emerging technologies, such as augmented reality (AR), also offers promising opportunities to further enhance the precision and efficacy of surgical procedures.

SAFENet: Semantic-Aware Feature Enhancement Network for unsupervised cross-domain road scene segmentation

Unsupervised cross-domain road scene segmentation has attracted substantial interest because of its capability to perform segmentation on new and unlabeled domains, thereby reducing the dependence on expensive manual annotations. This is achieved by leveraging networks trained on labeled source domains to classify images on unlabeled target domains. Conventional techniques usually use adversarial networks to align inputs from the source and the target in either of their domains. However, these approaches often fall short in effectively integrating information from both domains due to Alignment in each space usually leads to bias problems during feature learning. To overcome these limitations and enhance cross-domain interaction while mitigating overfitting to the source domain, we introduce a novel framework called Semantic-Aware Feature Enhancement Network (SAFENet) for Unsupervised Cross-domain Road Scene Segmentation. SAFENet incorporates the Semantic-Aware Enhancement (SAE) module to amplify the importance of class information in segmentation tasks and uses the semantic space as a new domain to guide the alignment of the source and target domains. Additionally, we integrate Adaptive Instance Normalization with Momentum (AdaIN-M) techniques, which convert the source domain image style to the target domain image style, thereby reducing the adverse effects of source domain overfitting on target domain segmentation performance. Moreover, SAFENet employs a Knowledge Transfer (KT) module to optimize network architecture, enhancing computational efficiency during testing while maintaining the robust inference capabilities developed during training. To further improve the segmentation performance, we further employ Curriculum Learning, a self-training mechanism that uses pseudo-labels derived from the target domain to iteratively refine the network. Comprehensive experiments on three well-known datasets, “Synthia→Cityscapes” and “GTA5→Cityscapes”, demonstrate the superior performance of our method. In-depth examinations and ablation studies verify the efficacy of each module within the proposed method.

Evaluating Uncertainty Quantification in Medical Image Segmentation: A Multi-Dataset, Multi-Algorithm Study

Deep learning is revolutionizing various scientific fields, with medical applications at the forefront. One key focus is automating image segmentation, a process crucial in many clinical services. However, medical images are often ambiguous and challenging even for experts. To address this, reliable models need to quantify their uncertainty, allowing physicians to understand the model’s confidence in its segmentation. This paper explores how the performance and uncertainty of a model are influenced by the number of annotations per input sample. We examine the effects of both single and multiple manual annotations on various deep learning architectures. To tackle this question, we employ three widely recognized deep learning architectures and evaluate them across four publicly available datasets. Furthermore, we explore the effects of dropout rates on Monte Carlo models by examining uncertainty models with dropout rates of 20%, 40%, 60%, and 80%. Subsequently, we evaluate the models using various measurement metrics. The findings reveal that the influence of multiple annotations varies significantly depending on the datasets. Additionally, we observe that the dropout rate has minimal or no impact on the model’s performance unless there is a substantial loss of training data, primarily evident in the 80% dropout rate scenario.

Superpixel-based principal feature clustering annotation method for dual-phase microstructure segmentation

Metallographic analysis is one of the most commonly used techniques by materials scientists for studying metal materials. The deep learning methods, which have been widely applied in metallographic images analysis, demonstrate excellent performance in this task. However, the optimization of deep learning models often relies on a substantial amount of accurately labeled samples for effective supervision. To address this issue, this paper proposes a novel automatic annotation method based on brightness and spatial distribution which is suitable for deep learning-based segmentation of optical dual-phase metallographic microstructure. The proposed automatic annotated method includes the superpixel segmentation, principal feature extraction, and clustering algorithm therefore it is referred as the superpixel-based principal feature clustering annotation (SPFCA) method. SPFCA employs discriminative criteria similar to those used by metallurgists to differentiate between metallographic structures. Furthermore, it can mitigate the occasional errors inherent in manual annotation, leading to improved performance compared to models trained with expert annotations. Experimental validation was conducted using four self-built datasets with different image qualities to test the performance of models from different perspective. Initially, hyperparameter optimization for the SPFCA method tailored to our dataset was performed. Subsequently, SPFCA was utilized to guide the optimization of the convolutional neural network employed for segmentation. The results demonstrate that the segmentation model optimized with SPFCA guidance achieved an F1 score of 0.9226 in the single dataset without the need for manual labeling, surpassing the segmentation models optimized with expert annotations.

Positional encoding-guided transformer-based multiple instance learning for histopathology whole slide images classification

Background and objectives:Whole slide image (WSI) classification is of great clinical significance in computer-aided pathological diagnosis. Due to the high cost of manual annotation, weakly supervised WSI classification methods have gained more attention. As the most representative, multiple instance learning (MIL) generally aggregates the predictions or features of the patches within a WSI to achieve the slide-level classification under the weak supervision of WSI labels. However, most existing MIL methods ignore spatial position relationships of the patches, which is likely to strengthen the discriminative ability of WSI-level features. Methods:In this paper, we propose a novel positional encoding-guided transformer-based multiple instance learning (PEGTB-MIL) method for histopathology WSI classification. It aims to encode the spatial positional property of the patch into its corresponding semantic features and explore the potential correlation among the patches for improving the WSI classification performance. Concretely, the deep features of the patches in WSI are first extracted and simultaneously a position encoder is used to encode the spatial 2D positional information of the patches into the spatial-aware features. After incorporating the semantic features and spatial embeddings, multi-head self-attention (MHSA) is applied to explore the contextual and spatial dependencies of the fused features. Particularly, we introduce an auxiliary reconstruction task to enhance the spatial-semantic consistency and generalization ability of features. Results:The proposed method is evaluated on two public benchmark TCGA datasets (TCGA-LUNG and TCGA-BRCA) and two in-house clinical datasets (USTC-EGFR and USTC-GIST). Experimental results validate it is effective in the tasks of cancer subtyping and gene mutation status prediction. In the test stage, the proposed PEGTB-MIL outperforms the other state-of-the-art methods and respectively achieves 97.13±0.34%, 86.74±2.64%, 83.25±1.65%, and 72.52±1.63% of the area under the receiver operating characteristic (ROC) curve (AUC). Conclusion:PEGTB-MIL utilizes positional encoding to effectively guide and reinforce MIL, leading to enhanced performance on downstream WSI classification tasks. Specifically, the introduced auxiliary reconstruction module adeptly preserves the spatial-semantic consistency of patch features. More significantly, this study investigates the relationship between position information and disease diagnosis and presents a promising avenue for further research.

Enhancing emerging technology discovery in nanomedicine by integrating innovative sentences using BERT and NLDA

Abstract Purpose Nanomedicine has significant potential to revolutionize biomedicine and healthcare through innovations in diagnostics, therapeutics, and regenerative medicine. This study aims to develop a novel framework that integrates advanced natural language processing, noise-free topic modeling, and multidimensional bibliometrics to systematically identify emerging nanomedicine technology topics from scientific literature. Design/methodology/approach The framework involves collecting full-text articles from PubMed Central and nanomedicine-related metrics from the Web of Science for the period 2013–2023. A fine-tuned BERT model is employed to extract key informative sentences. Noiseless Latent Dirichlet Allocation (NLDA) is applied to model interpretable topics from the cleaned corpus. Additionally, we develop and apply metrics for novelty, innovation, growth, impact, and intensity to quantify the emergence of novel technological topics. Findings By applying this methodology to nanomedical publications, we identify an increasing emphasis on research aligned with global health priorities, particularly inflammation and biomaterial interactions in disease research. This methodology provides deeper insights through full-text analysis and leading to a more robust discovery of emerging technologies. Research limitations One limitation of this study is its reliance on the existing scientific literature, which may introduce publication biases and language constraints. Additionally, manual annotation of the dataset, while thorough, is subject to subjectivity and can be time-consuming. Future research could address these limitations by incorporating more diverse data sources, and automating the annotation process. Practical implications The methodology presented can be adapted to explore emerging technologies in other scientific domains. It allows for tailored assessment criteria based on specific contexts and objectives, enabling more precise analysis and decision-making in various fields. Originality/value This study offers a comprehensive framework for identifying emerging technologies in nanomedicine, combining theoretical insights and practical applications. Its potential for adaptation across scientific disciplines enhances its value for future research and decision-making in technology discovery.

Machine Learning Approach to Injury Monitoring in Children and Adolescents

Abstract Pediatric injuries are common and associated with individual suffering and costs for the healthcare systems and society. The incidence of injuries among children in Switzerland has so far been insufficiently monitored. Emergency department patient records can be a valuable database to assess medically relevant injuries and gain information for monitoring and evidence-based prevention. The talk will outline our stepwise development of a machine-learning program for a pediatric injury monitoring system and lessons learnt. In this feasibility study we draw on electronic case histories of children treated for injuries between 2018 and 2022 at the University Children’s Hospital Zürich emergency department (N= to 30’884). The ML- approach follows different steps: step 1. utility evaluation of electronic pediatric patient records regarding ML and injury monitoring and prevention requirements in a test-sample of 100 electronic pediatric case histories; step 2. choice/adaptation of coding tree; step 3. coder training and inter-coder reliability testing; step 4. manual annotation of a sub-sample (n 1000) of the electronic pediatric patient records, step 5. design and testing of the ML program in an iterative process, and step 6. performance testing in a random sample of pediatric patient records. First lessons learnt indicate that electronic case history data contain relevant data and fulfill requirements for monitoring and prevention purposes. The IDB coding system proved too complex and detailed for the pediatric data base, providing too few examples for the ML-learning process, and needed to be adapted. Initial data analyses indicate the majority of cases are classified as less severe (73%) and only few (<1%) as very severe but the later provide the most detailed data. ML is a promising approach and makes use of hospital data for a national monitoring of pediatric injuries in Switzerland.

A comparative study of LLMs and NLP approaches for supporting business process analysis

ABSTRACT This paper compares two approaches to support Business Process Analysis (BPA) for the construction of a Business Process Knowledge Base (BPKB). The methodology is based on the BPA Canvas metamodel, which starts from the preliminary output of the business analysis (an interview). We focus on the extraction of key elements from the text to build the BPKBcore. We experiment with approaches based on Natural Language Processing (NLP) and Large Language Models (LLMs), using a running example to compare the outcomes with manual annotations. The experiment shows that the LLM-based approach yields better performance, especially with enriched prompts.

CriteriaMapper: establishing the automatic identification of clinical trial cohorts from electronic health records by matching normalized eligibility criteria and patient clinical characteristics

The use of electronic health records (EHRs) holds the potential to enhance clinical trial activities. However, the identification of eligible patients within EHRs presents considerable challenges. We aimed to develop a CriteriaMapper system for phenotyping eligibility criteria, enabling the identification of patients from EHRs with clinical characteristics that match those criteria. We utilized clinical trial eligibility criteria and patient EHRs from the Mount Sinai Database. The CriteriaMapper system was developed to normalize the criteria using national standard terminologies and in-house databases, facilitating computability and queryability to bridge clinical trial criteria and EHRs. The system employed rule-based pattern recognition and manual annotation. Our system normalized 367 out of 640 unique eligibility criteria attributes, covering various medical conditions including non-small cell lung cancer, small cell lung cancer, prostate cancer, breast cancer, multiple myeloma, ulcerative colitis, Crohn’s disease, non-alcoholic steatohepatitis, and sickle cell anemia. About 174 criteria were encoded with standard terminologies and 193 were normalized using the in-house reference tables. The agreement between automated and manual normalization was high (Cohen’s Kappa = 0.82), and patient matching demonstrated a 0.94 F1 score. Our system has proven effective on EHRs from multiple institutions, showing broad applicability and promising improved clinical trial processes, leading to better patient selection, and enhanced clinical research outcomes.

Highlighted Diffusion Model as Plug-in Priors for Polyp Segmentation.

Automated polyp segmentation from colonoscopy images is crucial for colorectal cancer diagnosis. The accuracy of such segmentation, however, is challenged by two main factors. First, the variability in polyps' size, shape, and color, coupled with the scarcity of well-annotated data due to the need for specialized manual annotation, hampers the efficacy of existing deep learning methods. Second, concealed polyps often blend with adjacent intestinal tissues, leading to poor contrast that challenges segmentation models. Recently, diffusion models have been explored and adapted for polyp segmentation tasks. However, the significant domain gap between RGB-colonoscopy images and grayscale segmentation masks, along with the low efficiency of the diffusion generation process, hinders the practical implementation of these models. To mitigate these challenges, we introduce the Highlighted Diffusion Model Plus (HDM+), a two-stage polyp segmentation framework. This framework incorporates the Highlighted Diffusion Model (HDM) to provide explicit semantic guidance, thereby enhancing segmentation accuracy. In the initial stage, the HDM is trained using highlighted ground-truth data, which emphasizes polyp regions while suppressing the background in the images. This approach reduces the domain gap by focusing on the image itself rather than on the segmentation mask. In the subsequent second stage, we employ the highlighted features from the trained HDM's U-Net model as plug-in priors for polyp segmentation, rather than generating highlighted images, thereby increasing efficiency. Extensive experiments conducted on six polyp segmentation benchmarks demonstrate the effectiveness of our approach.