Biology Methods and Protocols

IntroductionSurgical success hinges on two core factors: technical execution and cognitive planning. While the former can be trained and potentially augmented through robotics, the latter -- developing an accurate "mental roadmap" of a certain operation -- remains complex, deeply individualized and resistant to standardization. In neurosurgery, where minute anatomical distinctions can dictate outcomes, enhancing intraoperative guidance could reduce variability among surgeons and improve global standards. Recent developments in machine vision offer a promising avenue. Previous studies demonstrated that deep learning models could successfully identify anatomical landmarks in highly standardized procedures such as trans-sphenoidal surgery (TSS). However, the applicability of such techniques in more variable and multidimensional intracranial procedures remains unproven. This study investigates whether a deep learning model can recognize key anatomical structures during the more complex pterional trans-sylvian (PTS) approach. Materials and MethodsWe developed a deep learning object detection model (YOLOv7x) trained on 5.307 labeled frames from 78 surgical videos of 76 patients undergoing PTS. Surgical steps were standardized, and key anatomical targets--frontal/temporal dura, inferior frontal/superior temporal gyri, optic and olfactory nerves and internal carotid artery (ICA) -- were annotated by specifically trained neurosurgical residents and verified by the operating surgeon. Bounding boxes derived from segmentation masks served as training inputs. Performance was evaluated using five-fold cross-validation. ResultsThe model achieved promising detection performance for deep structures, particularly the optic nerve (AP50: 0.73) and ICA (AP50: 0.67). Superficial structures, like the dura and the cortical gyri, had lower precision (AP50 range: 0.25-0.45), likely due to morphological similarity and optical variability. Performance variability across classes reflects the complexity of the anatomical setting along with data limitations. ConclusionThis study shows the feasibility of applying machine vision techniques for anatomical detection in a complex and variable neurosurgical setting. While challenges remain in detecting less distinctive structures, the high accuracy achieved for deep anatomical landmarks validates this approach. These findings mark an essential step towards a machine vision surgical guidance system. Future applications could include real-time anatomical recognition, integration with neuronavigation and the development of AI-supported "surgical roadmaps" to improve intraoperative orientation and global neurosurgical practice.

BackgroundRisk stratification models estimate the probabilities of adverse outcomes after cardiac surgical procedures, which helps clinicians and patients make informed decisions. ObjectivesWe used the 12-lead electrocardiogram (ECG) and/or Society for Thoracic Surgeons (STS) variables to predict postoperative outcomes using deep learning methods that can incorporate diverse data types. MethodsWe developed a deep convolutional neural network ("ECGNet") that predicts operative mortality and other adverse outcomes using preoperative 12-lead ECGs (n=30,877) from 12,933 patients who underwent 13,299 cardiac surgical procedures. We also developed a deep neural network applied to preoperative STS variables ("STSNet"). STSNet and ECGNet are multi-task neural networks that utilize secondary outcomes to augment prediction of mortality using the same neural network. ResultsECGNet achieved a mean area under the receiver operating characteristic curve (AUC) of 0.85 for predicting operative mortality for all procedures and 0.93 for valve procedures. STSNet achieved a mean AUC of 0.85 for all procedures, with statistically similar performance as ECGNet for all procedures. Combining ECGNet and STSNet achieved a mean AUC of 0.90 for predicting operative mortality after all procedures, which is significantly higher than either ECGNet or STSNet alone. ConclusionsA deep neural network trained on STS features has higher predictive performance than previously reported for existing conventional models and is not limited to certain types of cardiac surgical procedures. A model trained on ECG alone can predict operative mortality with similar performance as STS features and adding ECG to STS features in a neural network can improve performance. These findings demonstrate the potential in leveraging deep learning on multidimensional data sources to predict outcomes after cardiac surgery. Condensed abstractIn this study, deep learning (DL) is applied to electrocardiograms and clinical features used in the standard STS risk prediction tools to generate new high-performing risk calculators for cardiac surgical procedures. Preoperative voltage waveforms contain information about cardiovascular risk and cardiac function and are passed as inputs to the deep learning model. These risk models apply to all cardiac procedures including those procedures that do not have standard STS risk calculators and provide improved performance. DL models enable the incorporation of additional modalities of data to improve risk prediction in cardiac surgery.

PurposePatient body composition is a major factor in patient management. Indeed, assessment of SMI as well as VFA and, to a lesser extent, SFA is a major factor in patient survival, particularly in surgery. However, to date, there is no simple, rapid, open-access assessment method. The aim of this work is to provide a simple, rapid and accurate tool for assessing patients body composition. Material and methodsA total of 343 patients underwent liver transplantation at the University Hospital of Rennes between January 1st, 2012 and December 31s, 2018. Image analysis was performed using the open source software ImageJ. Tissue distinction was based on Hounsfield density. The training dataset used 332 images (320 for training and 12 for validation). The model was evaluated on 11 patients. The complete software and video package is available at https://github.com/tpecot/MuViSS. ResultsIn total, the model was trained with 332 images and evaluated on 11 images. Model accuracy is 0.974 (SD 0.003), Jaccards index is 0.98 for visceral fat, 0.895 for muscle and 0.94 for subcutaneous fat. The Dice index is 0.958 (SD 0.003) for visceral fat, 0.944 (SD: 0.012) for muscle and 0.970 (SD: 0.013) for subcutaneous fat. Finally, the Normalized root mean square error is 0.007 for visceral fat, 0.0518 for muscle and 0.0124 for subcutaneous fat. ConclusionTo our knowledge, this is the first freely available model for assessing body composition. The model is fast, simple and accurate, based on Deep Learning. Statements and declarationsAll authors declare no conflict of interest

The Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2), also known as 2019 novel coronavirus (2019-nCoV), is a highly infectious RNA virus. A still-debated percentage of patients develop coronavirus disease 2019 (COVID-19) after infection, whose symptoms include fever, cough, shortness of breath and fatigue. Acute and life-threatening respiratory symptoms are experienced by 10-20% of symptomatic patients, particularly those with underlying medical conditions that includes diabetes, COPD and pregnancy. One of the main challenges in the containment of COVID-19 is the identification and isolation of asymptomatic/pre-symptomatic individuals. As communities re-open, large numbers of people will need to be tested and contact-tracing of positive patients will be required to prevent additional waves of infections and enable the continuous monitoring of the viral loads COVID-19 positive patients. A number of molecular assays are currently in clinical use to detect SARS-CoV-2. Many of them can accurately test hundreds or even thousands of patients every day. However, there are presently no testing platforms that enable more than 10,000 tests per day. Here, we describe the foundation for the REcombinase Mediated BaRcoding and AmplificatioN Diagnostic Tool (REMBRANDT), a high-throughput Next Generation Sequencing-based approach for the simultaneous screening of over 100,000 samples per day. The REMBRANDT protocol includes direct two-barcoded amplification of SARS-CoV-2 and control amplicons using an isothermal reaction, and the downstream library preparation for Illumina sequencing and bioinformatics analysis. This protocol represents a potentially powerful approach for community screening, a major bottleneck for testing samples from a large patient population for COVID-19.

Accurate and interpretable brain tumor classification remains a critical challenge due to the heterogeneity of tumor types and the complexity of MRI data. This paper presents a hybrid deep learning framework that synergizes Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) for multi-class brain tumor diagnosis. The model leverages CNNs for localized spatial feature extraction and ViTs for capturing long-range contextual information, followed by an attention-guided fusion mechanism. To enhance generalization and reduce feature redundancy, an Improved Aquila Optimizer (AQO) is employed for metaheuristic feature selection. The model is trained and evaluated on the Kaggle brain MRI dataset, comprising 3,264 T1-weighted contrast-enhanced axial slices categorized into four classes: glioma, meningioma, pituitary tumor, and no tumor. To ensure interpretability, SHAP and Grad-CAM are integrated to visualize both semantic and spatial relevance in predictions. The proposed method achieves a classification accuracy of 97.2%, F1-score of 0.96, and AUC-ROC of 0.98, outperforming baseline CNN and ViT models.

IntroductionPreoperative neurosurgical planning is a keen step to avoiding surgical complications, reducing morbidity, and improving patient safety. The incursion of machine learning (ML) in this domain has recently gained attention, given the notable advantages in processing large data sets and potentially generating efficient and accurate algorithms in patient care. ObjectiveTo evaluate the evolving applications of ML algorithms in the preoperative planning of brain and spine surgery. MethodsIn accordance with the Arksey and OMalley framework, a scoping review was conducted using three databases (Pubmed, Embase, and Web of Science). Articles that described the use of ML for preoperative planning in brain and spine surgery were included. Relevant data were collected regarding the neurosurgical field of application, patient baseline features, disease description, type of ML technology, studys aim, preoperative ML algorithm description, and advantages and limitations of ML algorithms. ResultsOur search strategy yielded 7,407 articles, of which 8 studies (5 retrospective, 2 prospective, and 1 experimental study) satisfied the inclusion criteria. Clinical information from 518 patients (62.7% female; mean age: 44.8 years) was used for generating ML algorithms, including convolutional neural network (14.3%), logistic regression (14.3%), random forest (14.3%), and other algorithms (Table 1). Neurosurgical fields of applications included functional neurosurgery (37.5%), tumor surgery (37.5%), and spine surgery (25%). The main advantages of ML included automated processing of clinical and imaging information, selection of an individualized patient surgical approach and data-driven support for treatment decision-making. All studies reported technical limitations, such as long processing time, algorithmic bias, limited generalizability, and the need for database updating and maintenance. O_TBL View this table: org.highwire.dtl.DTLVardef@cddb4corg.highwire.dtl.DTLVardef@f87cd7org.highwire.dtl.DTLVardef@1cc34e3org.highwire.dtl.DTLVardef@1a4346aorg.highwire.dtl.DTLVardef@16d4f47_HPS_FORMAT_FIGEXP M_TBL O_FLOATNOTable 1.C_FLOATNO O_TABLECAPTIONCharacteristics of included studies, demographics, and clinical information. AI: artificial intelligence; AIS: adolescent idiopathic scoliosis; CT: computed tomography; DBS: deep brain stimulation; DL: deep learning; ML: machine learning; MRI: magnetic resonance imaging; PC: principal components; VS: vestibular schwannoma; 3D: Tridimensional. C_TABLECAPTION C_TBL ConclusionML algorithms for preoperative neurosurgical planning are being developed for efficient, automated, and safe treatment decision-making. Enhancing the robustness, transparency, and understanding of ML applications will be crucial for their successful integration into neurosurgical practice.

IntroductionPancreaticoduodenectomy (PD) for patients with pancreatic ductal adenocarcinoma (PDAC) is associated with a high risk of postoperative complications (PoCs) and risk prediction of these is therefore critical for optimal treatment planning. We hypothesize that novel deep learning network approaches through transfer learning may be superior to legacy approaches for PoC risk prediction in the PDAC surgical setting. MethodsData from the US National Surgical Quality Improvement Program (NSQIP) 2002-2018 was used, with a total of 5,881,881 million patients, including 31,728 PD patients. Modelling approaches comprised of a model trained on a general surgery patient cohort and then tested on a PD specific cohort (general model), a transfer learning model trained on the general surgery patients with subsequent transfer and retraining on a PD-specific patient cohort (transfer learning model), a model trained and tested exclusively on the PD-specific patient cohort (direct model), and a benchmark random forest model trained on the PD patient cohort (RF model). The models were subsequently compared against the American College of Surgeons (ACS) surgical risk calculator (SRC) in terms of predicting mortality and morbidity risk. ResultsBoth the general model and transfer learning model outperformed the RF model in 14 and 16 out of 19 prediction tasks, respectively. Additionally, both models outperformed the direct model on 17 out of the 19 tasks. The transfer learning model also outperformed the general model on 11 out of the 19 prediction tasks. The transfer learning model outperformed the ACS-SRC regarding mortality and all the models outperformed the ACS-SRC regarding the morbidity prediction with the general model achieving the highest Receiver Operator Area Under the Curve (ROC AUC) of 0.668 compared to the 0.524 of the ACS SRC. ConclusionDNNs deployed using a transfer learning approach may be of value for PoC risk prediction in the PD setting.

Genomic surveillance has become a useful tool for better understanding virus pathogenicity, origin and spread. Obtaining accurately assembled, complete viral genomes directly from clinical samples is still a challenging. Here, we describe three protocols using a unique primer set designed to recover long reads of SARS-CoV-2 directly from total RNA extracted from clinical samples. This protocol is useful, accessible and adaptable to laboratories with varying resources and access to distinct sequencing methods: Nanopore, Illumina and/or Sanger.

BackgroundTraditional surgical training relies on an apprenticeship model, which is subjective and threatened by human bias. Performance metric scales attempt to offer more objective feedback by providing a structured grading rubric, but these scores are still ultimately subjective. Leveraging computer vision and artificial intelligence to assess surgical performance has the potential to shift subjective measurements into automated and objective feedback for trainees. Materials and MethodsThis retrospective, multi-institutional study analyzed 319 laparoscopic cholecystectomy videos from IRB-approved deidentified datasets and segmented the videos into 2862 clips. Using an internally validated video-based assessment rubric, we annotated video clips across five metacompetency domains: tissue handling, psychomotor skills, efficiency, dissection quality, and exposure quality. Short video segments (<90s) were rated on a 5-point scale by expert raters. We trained a deep learning model (DINOv2) to classify composite high (4-5) vs low (1-3) metacompetency scores, representing yes/no binary feedback, a realistic comparison to the operating room. Model performance was evaluated via area under the receiver operating characteristic curve. ResultsAmong 2862 LC video clips, model performance was highest for dissection quality during the exposing gallbladder step (AUROC 91.5%, 95% confidence interval [CI], 84.5-96.5). Moderate performance was observed for efficiency (AUROC 72.6%, 95% CI 59.9-83.2) and exposure quality (AUROC 68.7%, 95% CI 55.2-81.8). Dissection and exposure quality scores during hepatocystic triangle dissection yielded AUROCs of 63.8% (95% CI 56.9-71.3) and 66.0% (95% CI 53.1-76.7), respectively. ConclusionWe demonstrate the feasibility of a purely vision-based deep learning model to grade surgical skill based on metacompetencies with excellent performance during simple steps. This technique represents an advance over prior whole-video approaches that rely heavily on tool- tracking and kinematic data, and may lead to greater model performance by using binary feedback on increasingly specific step segmentation.

In the field of pattern recognition, achieving high accuracy is essential. While training a model to recognize different complex images, it is vital to fine-tune the model to achieve the highest accuracy possible. One strategy for fine-tuning a model involves changing its activation function. Most pre-trained models use ReLU as their default activation function, but switching to a different activation function like Hard-Swish could be beneficial. This study evaluates the performance of models using ReLU, Swish and Hard-Swish activation functions across diverse image datasets. Our results show a 2.06% increase in accuracy for models on the CIFAR-10 dataset and a 0.30% increase in accuracy for models on the ATLAS dataset. Modifying the activation functions in architecture of pre-trained models lead to improved overall accuracy.

Telomere length dynamics are an established biomarker of health and aging in animals. The study of telomeres in numerous species has been facilitated by methods to measure telomere length by real-time quantitative PCR (qPCR). In this method, telomere length is determined by quantifying the amount of telomeric DNA repeats in a sample and normalizing this to the total amount of genomic DNA. This normalization requires the development of genomic reference primers suitable for qPCR, which remains challenging in non-model organism with genomes that have not been sequenced. Here we report reference primers that can be used in qPCR to measure telomere lengths in any vertebrate species. We designed primer pairs to amplify genetic elements that are highly conserved between evolutionarily distant taxa and tested them in species that span the vertebrate tree of life. We report five primer pairs that meet the specificity and reproducibility standards of qPCR. In addition, we demonstrate how to choose the best primers for a given species by testing the primers on multiple individuals within a species and applying an established computational tool. These reference primers can facilitate the application of qPCR-based telomere length measurements in any vertebrate species of ecological or economic interest.

BackgroundMicrosatellite stability status determination is important for prognostication and therapeutic decision making in colorectal cancer management, but the conventional methods for this assessment are not readily available, especially in low- and middle-income countries. Deep learning (DL) models have been proposed for addressing this problem; however, potential computational cost due to model complexity and inadequate explainability may limit their adoption in low-resource settings. This study explored the potential of explainable lightweight models for detection of microsatellite instability in colorectal cancer. MethodsDL models were trained using a public dataset of colorectal cancer histology images and then used to classify a set of test images into one of two classes: microsatellite instability or microsatellite stability. The models were compared for efficiency. Gradient-weighted class activation mapping (Grad-CAM) was used to interpret the models decision making. ResultsThe simpler convolutional neural network (CNN) trained from scratch had modest performance (accuracy=0.757, area under receiver-operating characteristic curve [AUROC]=0.840). With an attention mechanism added, these values increased, but specificity and sensitivity reduced. Pretrained models performed better than the ones trained from scratch, and EfficientNet_B0 had the best balance of high performance and low computational requirements (accuracy=0.936, AUROC=0.990, negative predictive value=0.923, specificity=0.953, 4,010,000 trainable parameters, 0.38 gigaFLOPs). However, a simple CNN model with attention mechanism had the best interpretability based on Grad-CAM. ConclusionThis study demonstrated that DL models that are lightweight when compared to previously proposed ones can be useful for colorectal cancer microsatellite instability screening in resource-limited settings while balancing performance and computational efficiency.

This study presents a framework for processing Digital Imaging and Communications in Medicine (DICOM) medical imaging data by integrating synthetic objects for volumetric analysis and simulation for applications in assessment of computed tomography (CT) imaging used in thoracic surgery. Functions are designed to generate synthetic objects including geometric shapes such as spheres, cubes, rectangular prisms, cylinders, and blobs with known volumes. Validation is performed through test functions to ensure accuracy and consistency. Additionally, the use of UNet models for segmenting various chest pathologies, such as hemothorax and pneumothorax, as well as organs, is demonstrated. The created framework is used to generate synthetic data to address the scarcity of publicly available hemothorax CT imaging data. Models achieved high performance, assessed by various metrics. The framework and models provide a robust tool for data augmentation and analysis in medical imaging, potentially enhancing clinical decision-making and supporting research in thoracic surgery and related fields.

This paper documents the development of a novel method to predict the occurrence and exact locations of mitochondrial fission, fusion and depolarisation events in three dimensions. These occurrence and location of these events were successfully predicted with a three-dimensional version of the Pix2Pix generative adversarial network (GAN) as well as a three-dimensional adversarial segmentation network called the Vox2Vox GAN. The Pix2Pix GAN predicted the locations of mitochondrial fission, fusion and depolarisation events with accuracies of 35.9%, 33.2% and 4.90%, respectively. Similarly, the Vox2Vox GAN achieved accuracies of 37.1%, 37.3% and 7.43%. The accuracies achieved by the networks in this paper are too low for the immediate implementation of these tools in life science research. They do however indicate that the networks have modelled the mitochondrial dynamics to some degree of accuracy and may therefore still be helpful as an indication of where events might occur if time lapse sequences are not available. The prediction of these morphological mitochondrial events have, to our knowledge, never been achieved before in literature. The results from this paper can be used as a baseline for the results obtained by future work.

Quantitative Loop-mediated isothermal amplification (qLAMP) is a relatively new method that has gained popularity in recent years, particularly in disease identification, including during the recent SARS-CoV-2 pandemic. Unlike conventional quantitative PCR (qPCR), qLAMP features a linear amplification phase before the exponential phase. Determining cycle threshold (Ct) values through automatic thresholding may therefore produce inaccurate results, and the nature of these thresholds complicates comparability between studies and softwares. We introduce a new method for transforming sigmoidal amplification curves into inflection threshold curve (iCt) to address issues with auto thresholds and analysis of qLAMP. This method is implemented as a collection of R functions named LAMPrey, suitable for analysis of both qPCR and qLAMP reactions performed in the two most commonly used real-time thermocyclers. We simulate qLAMP amplification differences, demonstrate that iCt and Ct methods perform equivalently for conventional qPCR with an Illumina library quantitation kit, and show that iCt values outperform Ct values for quantifying qLAMP reactions in zebrafish embryos. All scripts developed for this paper are available at https://github.com/dodged13/LAMPrey Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=96 SRC="FIGDIR/small/651076v1_ufig1.gif" ALT="Figure 1"> View larger version (37K): org.highwire.dtl.DTLVardef@d5bbforg.highwire.dtl.DTLVardef@102713eorg.highwire.dtl.DTLVardef@158bcbcorg.highwire.dtl.DTLVardef@cf11b3_HPS_FORMAT_FIGEXP M_FIG C_FIG

The introduction of deep learning in both imaging and genomics has significantly advanced the analysis of biomedical data. For complex diseases such as cancer different data modalities may reveal different disease characteristics, and the integration of imaging with genomic data has the potential to unravel additional information then when using these data sources in isolation. Here, we propose a DL framework that by combining histopathology images with gene expression profiles can predict prognosis of brain tumors. Using two separate cohorts of 783 adult and 305 pediatric brain tumors, the developed multimodal data models achieved better prediction results compared to the single data models, but also leads to the identification of more relevant biological pathways. Importantly, when testing our adult models on a third independent brain tumor dataset, we show our multimodal framework is able to generalize and performs better on new data from different cohorts. Furthermore, leveraging the concept of transfer learning, we demonstrate how our multimodal models pre-trained on pediatric glioma can be used to predict prognosis for two more rare (less available samples) pediatric brain tumors, i.e. ependymoma and medulloblastoma. To summarize, our study illustrates that a multimodal data fusion approach can be successfully implemented and customized to model clinical outcome of adult and pediatric brain tumors.

AO_SCPLOWBSTRACTC_SCPLOWCancer is a complex disease characterized by the uncontrolled growth of abnormal cells in the body but can be prevented and even cured when detected early. Advanced medical imaging has introduced Whole Slide Images (WSIs). When combined with deep learning techniques, it can be used to extract meaningful features. These features are useful for various tasks such as classification and segmentation. There have been numerous studies involving the use of WSIs for survival analysis. Hence, it is crucial to determine their effectiveness for specific use cases. In this paper, we compared three publicly available vision encoders-UNI, Phikon and ResNet18 which are trained on millions of histopathological images, to generate feature embedding for survival analysis. WSIs cannot be fed directly to a network due to their size. We have divided them into 256 x 256 pixels patches and used a vision encoder to get feature embeddings. These embeddings were passed into an aggregator function to get representation at the WSI level which was then passed to a Long Short Term Memory (LSTM) based risk prediction head for survival analysis. Using breast cancer data from The Cancer Genome Atlas Program (TCGA) and k-fold cross-validation, we demonstrated that transformer-based models are more effective in survival analysis and achieved better C-index on average than ResNet-based architecture. The code1 for this study will be made available.

Malignant brain tumors are generally classified to be extremely aggressive and often can be fatal when not met with immediate action. Glioblastoma Multiforme is the most common type of malignant tumor found in the brain and is extremely aggressive. For this reason, advanced detection of malignant brain tumors is necessary for optimal mitigation. Conversely, the classification of tumors during Medical Resonance Imaging can be difficult due to bodily movements resulting in the movement of the tumor. The movement of the tumor can disrupt targeted radiotherapy and can also, at times, result in treatments about radiotherapy damaging healthy areas of the brain rather than areas of the tumor. This study proposes a novel deep learning system that can identify tumors from MRI images; which can be helpful for the case of early detection, as well as being able to track tumors during active imaging; resulting in higher efficiency with targeted radiotherapy. This is done utilizing Convolutional Neural Networks (CNNs) created via deep learning frameworks. With the image identification of tumors; 97% accuracy was achieved with optimization. The tumor-classification deep learning system achieved an accuracy of 98%. Further testing is required for optimization; with this optimization, higher accuracy can be reached.

Background and ObjectivesWe report the first intraoperative deployment of a real-time machine vision system in neurosurgery, derived from our previous anatomical detection work, automatically identifying structures during endoscopic endonasal surgery. Existing systems demonstrate promising performance in offline anatomical recognition, yet so far none have been implemented during live operations. MethodsA real-time anatomy detection model was trained using the YOLOv8 architecture (Ultralytics). Following training completion in the PyTorch environment, the model was exported to ONNX format and further optimized using the NVIDIA TensorRT engine. Deployment was carried out using the NVIDIA Holoscan SDK, the system ran on an NVIDIA Clara AGX developer kit. We used the model for real-time recognition of intraoperative anatomical structures and compared it with the same video labelled manually as reference. Model performance was reported using the average precision at an intersection-over-union threshold of 0.5 (AP50). Furthermore, end-to-end delay from frame acquisition to the display of the annotated output was measured. ResultsA mean AP50 of 0.56 was achieved. The model demonstrated reliable detection of the most relevant landmarks in the transsphenoidal corridor. The mean end-to-end latency of the model was 47.81 ms (median 46.57 ms). ConclusionFor the first time, we demonstrate that clinical-grade, real-time machine-vision assistance during neurosurgery is feasible and can provide continuous, automated anatomical guidance from the surgical field. This approach may enhance intraoperative orientation, reduce cognitive load, and offer a powerful tool for surgical training. These findings represent an initial step toward integrating real-time AI support into routine neurosurgical workflows.

I.AO_SCPLOWBSTRACTC_SCPLOWUnderstanding the temporal evolution of cells poses a significant challenge in developmental biology. This study embarks on a comparative analysis of various machine-learning techniques to classify sequences of cell colony images, thereby aiming to capture dynamic transitions of cellular states. Utilizing transfer learning with advanced classification networks, we achieved high accuracy in single-timestamp image categorization. We introduce temporal models--LSTM, R-Transformer, and ViViT--to explore the effectiveness of integrating temporal features in classification, comparing their performance against non-temporal models. This research benchmarks various machine learning approaches in understanding cellular dynamics, setting a foundation for future studies to enhance our understanding of cellular developments with computational methods, contributing significantly to biological research advancements.