Integration of smartphone technology with the patient call-bell system provides the opportunity to enhance patient safety by supporting nurses\' ability to communicate and prioritize care delivery directly. However, challenges are associated with achieving a balance between alarm support and alarm fatigue, including distracting nurses from patient care or desensitizing the nurse to other alarms and calls [1]. Our hospitals have quantitative and anecdotal reports of seriously high volumes of wireless alerts on the nurses\' smartphones. Nurses have complained that the phones are generating too much noise to consume or timely prioritize. Preliminary alarm inventory revealed the Bed Exit wireless alert as a leading contributor of signal volume across many units and hospitals. The lack of standard policies and workflow improvement processes has increased nuisance alarms, making these Health Information Technologies less useful and safe. Using system data, workflow observations, and nursing interviews, Singh and Sittig\'s HIT Safety framework [2] was applied to identify and prioritize sociotechnical factors and interventions that impact the end-to-end Bed Exit alarm workflow. This study reviews the application of sociotechnical models and frameworks to reduce wireless calls without introducing risk and impacting patient care.

OBJECTIVE: Even though workflow analysis in the operating room has come a long way, current systems are still limited to research. In the quest for a robust, universal setup, hardly any attention has been given to the dimension of audio despite its numerous advantages, such as low costs, location, and sight independence, or little required processing power.
METHODS: We present an approach for audio-based event detection that solely relies on two microphones capturing the sound in the operating room. Therefore, a new data set was created with over 63 h of audio recorded and annotated at the University Hospital rechts der Isar. Sound files were labeled, preprocessed, augmented, and subsequently converted to log-mel-spectrograms that served as a visual input for an event classification using pretrained convolutional neural networks.
RESULTS: Comparing multiple architectures, we were able to show that even lightweight models, such as MobileNet, can already provide promising results. Data augmentation additionally improved the classification of 11 defined classes, including inter alia different types of coagulation, operating table movements as well as an idle class. With the newly created audio data set, an overall accuracy of 90%, a precision of 91% and a F1-score of 91% were achieved, demonstrating the feasibility of an audio-based event recognition in the operating room.
CONCLUSIONS: With this first proof of concept, we demonstrated that audio events can serve as a meaningful source of information that goes beyond spoken language and can easily be integrated into future workflow recognition pipelines using computational inexpensive architectures.

Operation notes are a crucial component of patient care. However, writing them manually is prone to human error, particularly in high pressured clinical environments. Automatic generation of operation notes from video recordings can alleviate some of the administrative burdens, improve accuracy, and provide additional information. To achieve this for endoscopic pituitary surgery, 27-steps were identified via expert consensus. Then, for the 97-videos recorded for this study, a timestamp of each step was annotated by an expert surgeon. To automatically determine whether a step is present in a video, a three-stage architecture was created. Firstly, for each step, a convolution neural network was used for binary image classification on each frame of a video. Secondly, for each step, the binary frame classifications were passed to a discriminator for binary video classification. Thirdly, for each video, the binary video classifications were passed to an accumulator for multi-label step classification. The architecture was trained on 77-videos, and tested on 20-videos, where a 0.80 weighted-F1 score was achieved. The classifications were inputted into a clinically based predefined template, and further enriched with additional video analytics. This work therefore demonstrates automatic generation of operative notes from surgical videos is feasible, and can assist surgeons during documentation.

OBJECTIVE: Online Adaptive Radiation Therapy (oART) follows a different treatment paradigm than conventional radiotherapy, and because of this, the resources, implementation, and workflows needed are unique. The purpose of this report is to outline our institution\'s experience establishing, organizing, and implementing an oART program using the Ethos therapy system.
METHODS: We include resources used, operational models utilized, program creation timelines, and our institutional experiences with the implementation and operation of an oART program. Additionally, we provide a detailed summary of our first year\'s clinical experience where we delivered over 1000 daily adaptive fractions. For all treatments, the different stages of online adaption, primary patient set-up, initial kV-CBCT acquisition, contouring review and edit of influencer structures, target review and edits, plan evaluation and selection, Mobius3D 2nd check and adaptive QA, 2nd kV-CBCT for positional verification, treatment delivery, and patient leaving the room, were analyzed.
RESULTS: We retrospectively analyzed data from 97 patients treated from August 2021-August 2022. One thousand six hundred seventy seven individual fractions were treated and analyzed, 632(38%) were non-adaptive and 1045(62%) were adaptive. Seventy four of the 97 patients (76%) were treated with standard fractionation and 23 (24%) received stereotactic treatments. For the adaptive treatments, the generated adaptive plan was selected in 92% of treatments. On average(±std), adaptive sessions took 34.52 ± 11.42 min from start to finish. The entire adaptive process (from start of contour generation to verification CBCT), performed by the physicist (and physician on select days), was 19.84 ± 8.21 min.
CONCLUSIONS: We present our institution\'s experience commissioning an oART program using the Ethos therapy system. It took us 12 months from project inception to the treatment of our first patient and 12 months to treat 1000 adaptive fractions. Retrospective analysis of delivered fractions showed that the average overall treatment time was approximately 35 min and the average time for the adaptive component of treatment was approximately 20 min.

Despite the ubiquity of data science, we are far from rigorously understanding how coding in data science is performed. Even though the scientific literature has hinted at the iterative and explorative nature of data science coding, we need further empirical evidence to understand this practice and its workflows in detail. Such understanding is critical to recognise the needs of data scientists and, for instance, inform tooling support. To obtain a deeper understanding of the iterative and explorative nature of data science coding, we analysed 470 Jupyter notebooks publicly available in GitHub repositories. We focused on the extent to which data scientists transition between different types of data science activities, or steps (such as data preprocessing and modelling), as well as the frequency and co-occurrence of such transitions. For our analysis, we developed a dataset with the help of five data science experts, who manually annotated the data science steps for each code cell within the aforementioned 470 notebooks. Using the first-order Markov chain model, we extracted the transitions and analysed the transition probabilities between the different steps. In addition to providing deeper insights into the implementation practices of data science coding, our results provide evidence that the steps in a data science workflow are indeed iterative and reveal specific patterns. We also evaluated the use of the annotated dataset to train machine-learning classifiers to predict the data science step(s) of a given code cell. We investigate the representativeness of the classification by comparing the workflow analysis applied to (a) the predicted data set and (b) the data set labelled by experts, finding an F1-score of about 71% for the 10-class data science step prediction problem.

UNASSIGNED: Surgical efficiency and variability are critical contributors to optimal outcomes, patient experience, care team experience, and total cost to treat per disease episode. Opportunities remain to develop scalable, objective methods to quantify surgical behaviors that maximize efficiency and reduce variability. Such objective measures can then be used to provide surgeons with timely and user-specific feedbacks to monitor performances and facilitate training and learning. In this study, we used objective task-level analysis to identify dominant contributors toward surgical efficiency and variability across the procedural steps of robotic-assisted sleeve gastrectomy (RSG) over a five-year period for a single surgeon. These results enable actionable insights that can both complement those from population level analyses and be tailored to an individual surgeon\'s practice and experience.
UNASSIGNED: Intraoperative video recordings of 77 RSG procedures performed by a single surgeon from 2015 to 2019 were reviewed and segmented into surgical tasks. Surgeon-initiated events when controlling the robotic-assisted surgical system were used to compute objective metrics. A series of multi-staged regression analysis were used to determine: if any specific tasks or patient body mass index (BMI) statistically impacted procedure duration; which objective metrics impacted critical task efficiency; and which task(s) statistically contributed to procedure variability.
UNASSIGNED: Stomach dissection was found to be the most significant contributor to procedure duration (β = 0.344, p< 0.001; R = 0.81, p< 0.001) followed by surgical inactivity and stomach stapling. Patient BMI was not found to be statistically significantly correlated with procedure duration (R = -0.01, p = 0.90). Energy activation rate, a robotic system event-based metric, was identified as a dominant feature in predicting stomach dissection duration and differentiating earlier and later case groups. Reduction of procedure variability was observed between earlier (2015-2016) and later (2017-2019) groups (IQR = 14.20 min vs. 6.79 min). Stomach dissection was found to contribute most to procedure variability (β = 0.74, p < 0.001).
UNASSIGNED: A surgical task-based objective analysis was used to identify major contributors to surgical efficiency and variability. We believe this data-driven method will enable clinical teams to quantify surgeon-specific performance and identify actionable opportunities focused on the dominant surgical tasks impacting overall procedure efficiency and consistency.

OBJECTIVE: We tackle the problem of online surgical phase recognition in laparoscopic procedures, which is key in developing context-aware supporting systems. We propose a novel approach to take temporal context in surgical videos into account by precise modeling of temporal neighborhoods.
METHODS: We propose a two-stage model to perform phase recognition. A CNN model is used as a feature extractor to project RGB frames into a high-dimensional feature space. We introduce a novel paradigm for surgical phase recognition which utilizes graph neural networks to incorporate temporal information. Unlike recurrent neural networks and temporal convolution networks, our graph-based approach offers a more generic and flexible way for modeling temporal relationships. Each frame is a node in the graph, and the edges in the graph are used to define temporal connections among the nodes. The flexible configuration of temporal neighborhood comes at the price of losing temporal order. To mitigate this, our approach takes temporal orders into account by encoding frame positions, which is important to reliably predict surgical phases.
RESULTS: Experiments are carried out on the public Cholec80 dataset that contains 80 annotated videos. The experimental results highlight the superior performance of the proposed approach compared to the state-of-the-art models on this dataset.
CONCLUSIONS: A novel approach for formulating video-based surgical phase recognition is presented. The results indicate that temporal information can be incorporated using graph-based models, and positional encoding is important to efficiently utilize temporal information. Graph networks open possibilities to use evidence theory for uncertainty analysis in surgical phase recognition.

The objective of this study was to clarify gaze information patterns of nurses gathering patient information using electronic health records. We recorded the electronic health record screen on which nurses\' gazes were presented using an eye tracker and analyzed the recorded images. The analysis revealed two types of gaze information patterns of nurses engaged in patient information gathering. However, no regularity was observed in the gaze information patterns of the nurses viewing the electronic health record sections after selecting a patient.

BACKGROUND: Intraoperative digital subtraction angiography (ioDSA) allows early treatment evaluation after neurovascular procedures. However, the value and efficiency of this procedure has been discussed controversially. We have evaluated the additional value of hybrid operating room equipped with an Artis Zeego robotic c-arm regarding cost, efficiency and workflow. Furthermore, we have performed a risk-benefit analysis and compared it with indocyanine green (ICG) angiography.
METHODS: For 3 consecutive years, we examined all neurovascular patients, treated in the hybrid operating theater in a risk-benefit analysis. After using microdoppler and ICG angiography for best operative result, every patient received an additional ioDSA to look for remnants or unfavorable clip placement which might lead to a change of operating strategy or results. Furthermore, a workflow-analysis reviewing operating steps, staff positioning, costs, technical errors or complications were conducted on randomly selected cases.
RESULTS: 54 patients were enrolled in the risk-benefit analysis, 22 in the workflow analysis. The average duration of a cerebrovascular operation was 4 h 58 min 2 min 35 s accounted for ICG angiography, 46 min 4 s for ioDSA. Adverse events occurred during one ioDSA. In risk-benefit analysis, ioDSA was able to detect a perfusion rest in 2 out of 43 cases (4,7%) of aneurysm surgery, which could not have been visualized by ICG angiography before. In arterio-venous-malformation (AVM) surgery, one of 11 examined patients (7,7%) showed a remnant in ioDSA and resulted in additional resection. The average cost of an ioDSA in Ulm University can be estimated with 1928,00€.
CONCLUSIONS: According to our results ioDSA associated complications are low. Relevant findings in ioDSA can potentially avoid additional intervention, however, due to the high costs and lower availability, the main advantage might lie in the treatment of selected patients with complexes neurovascular pathologies since ICG angiography is equally safe but associated with lower costs and better availability.

BACKGROUND: Artificial intelligence (AI) has recently achieved considerable success in different domains including medical applications. Although current advances are expected to impact surgery, up until now AI has not been able to leverage its full potential due to several challenges that are specific to that field.
CONCLUSIONS: This review summarizes data-driven methods and technologies needed as a prerequisite for different AI-based assistance functions in the operating room. Potential effects of AI usage in surgery will be highlighted, concluding with ongoing challenges to enabling AI for surgery.
CONCLUSIONS: AI-assisted surgery will enable data-driven decision-making via decision support systems and cognitive robotic assistance. The use of AI for workflow analysis will help provide appropriate assistance in the right context. The requirements for such assistance must be defined by surgeons in close cooperation with computer scientists and engineers. Once the existing challenges will have been solved, AI assistance has the potential to improve patient care by supporting the surgeon without replacing him or her.