Survival Dynamics in Advanced Ovarian Cancer: R2 Resection Versus No-Surgery Paths Explored.

BACKGROUND: Cytoreductive surgery is critical for optimal tumor clearance in advanced epithelial ovarian cancer (EOC). Despite best efforts, some patients may experience R2 (>1 cm) resection, while others may not undergo surgery at all. We aimed to compare outcomes between advanced EOC patients undergoing R2 resection and those who had no surgery. METHODS: Retrospective data from 51 patients with R2 resection were compared to 122 patients with no surgery between January 2015 and December 2019 at a UK tertiary referral centre. Progression-free survival (PFS) and overall survival (OS) were the study endpoints. Principal Component Analysis and Term Frequency - Inverse Document Frequency scores were utilized for data discrimination and prediction of R>2 cm from computed tomography pre-operative reports, respectively. RESULTS: No statistical significance was observed, except for age (73 vs 67 years in the no- surgery vs R2 group, P: .001). Principal Components explained 34% of data variances. Reasons for no surgery included age, co-morbidities, patient preference, refractory disease, patient deterioration or disease progression, and absence of measurable intra- abdominal disease). The median PFS and OS were 12 and 14 months for no-surgery, vs 14 and 26 months for R2 (P: .138 and P: .001, respectively). Serous histology and performance status independently predicted PFS in both no-surgery and R2 cohorts. In the no-surgery cohort, serous histology independently predicted OS, while in the R2 cohorts, both serous histology and adjuvant chemotherapy were independent prognostic features for OS. The bi-grams "abdominopelvic ascites" and "solid omental" were amongst those best discriminating between R>2 cm and R1-2 cm. CONCLUSIONS: R2 resection and no-surgery cohorts displayed unfavourable prognosis with a notable degree of uniformity. When cytoreduction results in suboptimal results, the survival benefit may still be higher compared to those who underwent no surgery.

The study examined outcomes in advanced epithelial ovarian cancer (EOC) patients who underwent either R2 (suboptimal) surgical resection or received no surgery at all at a UK tertiary referral center. Sophisticated machine learning methodolgies were used to analyze data patterns and predict the extent of resection (>2 cm) from pre-operative CT reports. Reasons for not undergoing surgery included older age, presence of other medical conditions, patient preference, progressive disease, patient decline, or lack of detectable intra-abdominal disease. Factors like serous histology and performance status iinfluenced the risk of recurrence in both groups, while serous histology and adjuvant chemotherapy predicted the risk of death in the R2 group. Word sequences like "omental disease" and "reduced bulk" helped differentiate between R>2 cm and less extensive resections (R1-2 cm). In summary, both R2 resection and no-surgery groups had poor outcomes, but patients who underwent R2 resection generally had better survival compared to those who received no surgery, even when complete tumor removal was not achieved.

Performance of Publicly Available Large Language Models on Internal Medicine Board-style Questions.

Ongoing research attempts to benchmark large language models (LLM) against physicians' fund of knowledge by assessing LLM performance on medical examinations. No prior study has assessed LLM performance on internal medicine (IM) board examination questions. Limited data exists on how knowledge supplied to the models, derived from medical texts improves LLM performance. The performance of GPT-3.5, GPT-4.0, LaMDA and Llama 2, with and without additional model input augmentation, was assessed on 240 randomly selected IM board-style questions. Questions were sourced from the Medical Knowledge Self-Assessment Program released by the American College of Physicians with each question serving as part of the LLM prompt. When available, LLMs were accessed both through their application programming interface (API) and their corresponding chatbot. Mode inputs were augmented with Harrison's Principles of Internal Medicine using the method of Retrieval Augmented Generation. LLM-generated explanations to 25 correctly answered questions were presented in a blinded fashion alongside the MKSAP explanation to an IM board-certified physician tasked with selecting the human generated response. GPT-4.0, accessed either through Bing Chat or its API, scored 77.5-80.7% outperforming GPT-3.5, human respondents, LaMDA and Llama 2 in that order. GPT-4.0 outperformed human MKSAP users on every tested IM subject with its highest and lowest percentile scores in Infectious Disease (80th) and Rheumatology (99.7th), respectively. There is a 3.2-5.3% decrease in performance of both GPT-3.5 and GPT-4.0 when accessing the LLM through its API instead of its online chatbot. There is 4.5-7.5% increase in performance of both GPT-3.5 and GPT-4.0 accessed through their APIs after additional input augmentation. The blinded reviewer correctly identified the human generated MKSAP response in 72% of the 25-question sample set. GPT-4.0 performed best on IM board-style questions outperforming human respondents. Augmenting with domain-specific information improved performance rendering Retrieval Augmented Generation a possible technique for improving accuracy in medical examination LLM responses.

Continuity Unveiled: Evaluating Cytoreduction Outcomes for Advanced Ovarian Cancer Amidst the COVID-19 Era at an ESGO Designated Centre of Excellence.

BACKGROUND/AIM: The COVID-19 pandemic brought unprecedented global changes, necessitating adjustments to address public health challenges. The impact on advanced epithelial ovarian cancer (EOC) surgery, marked by increased perioperative risks, and changes in management plans was explored in this study based on promptly published British Gynaecologic Cancer Society (BGCS) and European Society of Gynaecologic Oncology (ESGO) guidelines. PATIENTS AND METHODS: Retrospective data from 332 patients with advanced EOC who underwent cytoreductive surgery at a UK tertiary center were analyzed, and the outcomes were compared between pre-COVID-19 (2018-2019) (n=189) and COVID-19 era (2020-2021) (n=143) cohorts, covering the same timeframe (March to December). Primary outcomes included residual disease (RD) and progression-free survival (PFS), while secondary outcomes were the ESGO quality indicators (QIs) for advanced EOC surgery. Kaplan-Meier curves were produced to illustrate PFS. RESULTS: Complete cytoreduction rates remained comparable at 74.07% and 72.03% for pre-COVID-19 and COVID-19 groups, respectively. Differences were observed in ECOG performance status (p=0.015), Intensive Care Unit (ICU) admissions (p=0.039) with less interval debulking surgeries (p=0.03), lower surgical complexity scores (p=0.02), and longer operative times in the COVID-19 group (p=0.01) compared to the pre-COVID-19 group. The median PFS rates were 37 months and 34 months in the pre-COVID-19 and COVID-19 groups, respectively (p=0.08). The surgical QIs 1-3 remained uncompromised during the COVID-19 era. CONCLUSION: Management modifications prompted by the COVID-19 pandemic did not adversely impact cytoreduction rates or PFS.

RoBERTa-Assisted Outcome Prediction in Ovarian Cancer Cytoreductive Surgery Using Operative Notes.

INTRODUCTION: Contemporary efforts to predict surgical outcomes focus on the associations between traditional discrete surgical risk factors. We aimed to determine whether natural language processing (NLP) of unstructured operative notes improves the prediction of residual disease in women with advanced epithelial ovarian cancer (EOC) following cytoreductive surgery. METHODS: Electronic Health Records were queried to identify women with advanced EOC including their operative notes. The Term Frequency - Inverse Document Frequency (TF-IDF) score was used to quantify the discrimination capacity of sequences of words (n-grams) regarding the existence of residual disease. We employed the state-of-the-art RoBERTa-based classifier to process unstructured surgical notes. Discrimination was measured using standard performance metrics. An XGBoost model was then trained on the same dataset using both discrete and engineered clinical features along with the probabilities outputted by the RoBERTa classifier. RESULTS: The cohort consisted of 555 cases of EOC cytoreduction performed by eight surgeons between January 2014 and December 2019. Discrete word clouds weighted by n-gram TF-IDF score difference between R0 and non-R0 resection were identified. The words 'adherent' and 'miliary disease' best discriminated between the two groups. The RoBERTa model reached high evaluation metrics (AUROC .86; AUPRC .87, precision, recall, and F1 score of .77 and accuracy of .81). Equally, it outperformed models that used discrete clinical and engineered features and outplayed the performance of other state-of-the-art NLP tools. When the probabilities from the RoBERTa classifier were combined with commonly used predictors in the XGBoost model, a marginal improvement in the overall model's performance was observed (AUROC and AUPRC of .91, with all other metrics the same). CONCLUSION/IMPLICATIONS: We applied a sui generis approach to extract information from the abundant textual surgical data and demonstrated how it can be effectively used for classification prediction, outperforming models relying on conventional structured data. State-of-art NLP applications in biomedical texts can improve modern EOC care.

Exploring the Potential Role of Upper Abdominal Peritonectomy in Advanced Ovarian Cancer Cytoreductive Surgery Using Explainable Artificial Intelligence.

The Surgical Complexity Score (SCS) has been widely used to describe the surgical effort during advanced stage epithelial ovarian cancer (EOC) cytoreduction. Referring to a variety of multi-visceral resections, it best combines the numbers with the complexity of the sub-procedures. Nevertheless, not all potential surgical procedures are described by this score. Lately, the European Society for Gynaecological Oncology (ESGO) has established standard outcome quality indicators pertinent to achieving complete cytoreduction (CC0). There is a need to define what weight all these surgical sub-procedures comprising CC0 would be given. Prospectively collected data from 560 surgically cytoreduced advanced stage EOC patients were analysed at a UK tertiary referral centre.We adapted the structured ESGO ovarian cancer report template. We employed the eXtreme Gradient Boosting (XGBoost) algorithm to model a long list of surgical sub-procedures. We applied the Shapley Additive explanations (SHAP) framework to provide global (cohort) explainability. We used Cox regression for survival analysis and constructed Kaplan-Meier curves. The XGBoost model predicted CC0 with an acceptable accuracy (area under curve [AUC] = 0.70; 95% confidence interval [CI] = 0.63-0.76). Visual quantification of the feature importance for the prediction of CC0 identified upper abdominal peritonectomy (UAP) as the most important feature, followed by regional lymphadenectomies. The UAP best correlated with bladder peritonectomy and diaphragmatic stripping (Pearson's correlations > 0.5). Clear inflection points were shown by pelvic and para-aortic lymph node dissection and ileocecal resection/right hemicolectomy, which increased the probability for CC0. When UAP was solely added to a composite model comprising of engineered features, it substantially enhanced its predictive value (AUC = 0.80, CI = 0.75-0.84). The UAP was predictive of poorer progression-free survival (HR = 1.76, CI 1.14-2.70, P: 0.01) but not overall survival (HR = 1.06, CI 0.56-1.99, P: 0.86). The SCS did not have significant survival impact. Machine Learning allows for operational feature selection by weighting the relative importance of those surgical sub-procedures that appear to be more predictive of CC0. Our study identifies UAP as the most important procedural predictor of CC0 in surgically cytoreduced advanced-stage EOC women. The classification model presented here can potentially be trained with a larger number of samples to generate a robust digital surgical reference in high output tertiary centres. The upper abdominal quadrants should be thoroughly inspected to ensure that CC0 is achievable.

The Future of AI in Ovarian Cancer Research: The Large Language Models Perspective.

Conversational large language model (LLM)-based chatbots utilize neural networks to process natural language. By generating highly sophisticated outputs from contextual input text, they revolutionize the access to further learning, leading to the development of new skills and personalized interactions. Although they are not developed to provide healthcare, their potential to address biomedical issues is rather unexplored. Healthcare digitalization and documentation of electronic health records is now developing into a standard practice. Developing tools to facilitate clinical review of unstructured data such as LLMs can derive clinical meaningful insights for ovarian cancer, a heterogeneous but devastating disease. Compared to standard approaches, they can host capacity to condense results and optimize analysis time. To help accelerate research in biomedical language processing and improve the validity of scientific writing, task-specific and domain-specific language models may be required. In turn, we propose a bespoke, proprietary ovarian cancer-specific natural language using solely in-domain text, whereas transfer learning drifts away from the pretrained language models to fine-tune task-specific models for all possible downstream applications. This venture will be fueled by the abundance of unstructured text information in the electronic health records resulting in ovarian cancer research ultimately reaching its linguistic home.

Explainable SHAP-XGBoost models for in-hospital mortality after myocardial infarction.

Background: A lack of explainability in published machine learning (ML) models limits clinicians' understanding of how predictions are made, in turn undermining uptake of the models into clinical practice. Objective: The purpose of this study was to develop explainable ML models to predict in-hospital mortality in patients hospitalized for myocardial infarction (MI). Methods: Adult patients hospitalized for an MI were identified in the National Inpatient Sample between January 1, 2012, and September 30, 2015. The resulting cohort comprised 457,096 patients described by 64 predictor variables relating to demographic/comorbidity characteristics and in-hospital complications. The gradient boosting algorithm eXtreme Gradient Boosting (XGBoost) was used to develop explainable models for in-hospital mortality prediction in the overall cohort and patient subgroups based on MI type and/or sex. Results: The resulting models exhibited an area under the receiver operating characteristic curve (AUC) ranging from 0.876 to 0.942, specificity 82% to 87%, and sensitivity 75% to 87%. All models exhibited high negative predictive value ≥0.974. The SHapley Additive exPlanation (SHAP) framework was applied to explain the models. The top predictor variables of increasing and decreasing mortality were age and undergoing percutaneous coronary intervention, respectively. Other notable findings included a decreased mortality risk associated with certain patient subpopulations with hyperlipidemia and a comparatively greater risk of death among women below age 55 years. Conclusion: The literature lacks explainable ML models predicting in-hospital mortality after an MI. In a national registry, explainable ML models performed best in ruling out in-hospital death post-MI, and their explanation illustrated their potential for guiding hypothesis generation and future study design.

Development of a Novel Intra-Operative Score to Record Diseases' Anatomic Fingerprints (ANAFI Score) for the Prediction of Complete Cytoreduction in Advanced-Stage Ovarian Cancer by Using Machine Learning and Explainable Artificial Intelligence.

BACKGROUND: The Peritoneal Carcinomatosis Index (PCI) and the Intra-operative Mapping for Ovarian Cancer (IMO), to a lesser extent, have been universally validated in advanced-stage epithelial ovarian cancer (EOC) to describe the extent of peritoneal dissemination and are proven to be powerful predictors of the surgical outcome with an added sensitivity of assessment at laparotomy of around 70%. This leaves room for improvement because the two-dimensional anatomic scoring model fails to reflect the patient's real anatomy, as seen by a surgeon. We hypothesized that tumor dissemination in specific anatomic locations can be more predictive of complete cytoreduction (CC0) and survival than PCI and IMO tools in EOC patients. (2) Methods: We analyzed prospectively data collected from 508 patients with FIGO-stage IIIB-IVB EOC who underwent cytoreductive surgery between January 2014 and December 2019 at a UK tertiary center. We adapted the structured ESGO ovarian cancer report to provide detailed information on the patterns of tumor dissemination (cancer anatomic fingerprints). We employed the extreme gradient boost (XGBoost) to model only the variables referring to the EOC disseminated patterns, to create an intra-operative score and judge the predictive power of the score alone for complete cytoreduction (CC0). Receiver operating characteristic (ROC) curves were then used for performance comparison between the new score and the existing PCI and IMO tools. We applied the Shapley additive explanations (SHAP) framework to support the feature selection of the narrated cancer fingerprints and provide global and local explainability. Survival analysis was performed using Kaplan-Meier curves and Cox regression. (3) Results: An intra-operative disease score was developed based on specific weights assigned to the cancer anatomic fingerprints. The scores range from 0 to 24. The XGBoost predicted CC0 resection (area under curve (AUC) = 0.88 CI = 0.854-0.913) with high accuracy. Organ-specific dissemination on the small bowel mesentery, large bowel serosa, and diaphragmatic peritoneum were the most crucial features globally. When added to the composite model, the novel score slightly enhanced its predictive value (AUC = 0.91, CI = 0.849-0.963). We identified a "turning point", ≤5, that increased the probability of CC0. Using conventional logistic regression, the new score was superior to the PCI and IMO scores for the prediction of CC0 (AUC = 0.81 vs. 0.73 and 0.67, respectively). In multivariate Cox analysis, a 1-point increase in the new intra-operative score was associated with poorer progression-free (HR: 1.06; 95% CI: 1.03-1.09, p < 0.005) and overall survival (HR: 1.04; 95% CI: 1.01-1.07), by 4% and 6%, respectively. (4) Conclusions: The presence of cancer disseminated in specific anatomical sites, including small bowel mesentery, large bowel serosa, and diaphragmatic peritoneum, can be more predictive of CC0 and survival than the entire PCI and IMO scores. Early intra-operative assessment of these areas only may reveal whether CC0 is achievable. In contrast to the PCI and IMO scores, the novel score remains predictive of adverse survival outcomes.

Integrated statistical indicators from Scottish linked open government data.

Open Government Data (OGD), including statistical data, such as economic, environmental and social indicators, are data published by the public sector for free reuse. These data have a huge potential when exploited using Machine Learning methods. Linked Data technologies facilitate retrieving integrated statistical indicators by defining and executing SPARQL queries. However, statistical indicators are available in different temporal and spatial granularity levels as well using different units of measurement. This data article describes the integrated statistical indicators that were retrieved from the official Scottish data portal in order to facilitate the exploitation of Machine Learning methods in OGD. Multiple SPARQL queries as well as manual search in the data portal were employed towards this end. The resulted dataset comprises the maximum number of compatible datasets, i.e., datasets with matching temporal and spatial characteristics. In particular, the data include 60 statistical indicators from seven categories such as health and social care, housing, and crime and justice. The indicators refer to the 6,976 "2011 data zones" of Scotland, while the year of reference is 2015. Data are ready to be used by the research community, students, policy makers, and journalists and give rise to plenty of social, business, and research scenarios that can be solved using Machine Learning technologies and methods.

Explaining the Elusive Nature of a Well-Defined Threshold for Blood Transfusion in Advanced Epithelial Ovarian Cancer Cytoreductive Surgery.

There is no well-defined threshold for intra-operative blood transfusion (BT) in advanced epithelial ovarian cancer (EOC) surgery. To address this, we devised a Machine Learning (ML)-driven prediction algorithm aimed at prompting and elucidating a communication alert for BT based on anticipated peri-operative events independent of existing BT policies. We analyzed data from 403 EOC patients who underwent cytoreductive surgery between 2014 and 2019. The estimated blood volume (EBV), calculated using the formula EBV = weight × 80, served for setting a 10% EBV threshold for individual intervention. Based on known estimated blood loss (EBL), we identified two distinct groups. The Receiver operating characteristic (ROC) curves revealed satisfactory results for predicting events above the established threshold (AUC 0.823, 95% CI 0.76-0.88). Operative time (OT) was the most significant factor influencing predictions. Intra-operative blood loss exceeding 10% EBV was associated with OT > 250 min, primary surgery, serous histology, performance status 0, R2 resection and surgical complexity score > 4. Certain sub-procedures including large bowel resection, stoma formation, ileocecal resection/right hemicolectomy, mesenteric resection, bladder and upper abdominal peritonectomy demonstrated clear associations with an elevated interventional risk. Our findings emphasize the importance of obtaining a rough estimate of OT in advance for precise prediction of blood requirements.

Exploring the Quality of Dynamic Open Government Data Using Statistical and Machine Learning Methods.

Dynamic data (including environmental, traffic, and sensor data) were recently recognized as an important part of Open Government Data (OGD). Although these data are of vital importance in the development of data intelligence applications, such as business applications that exploit traffic data to predict traffic demand, they are prone to data quality errors produced by, e.g., failures of sensors and network faults. This paper explores the quality of Dynamic Open Government Data. To that end, a single case is studied using traffic data from the official Greek OGD portal. The portal uses an Application Programming Interface (API), which is essential for effective dynamic data dissemination. Our research approach includes assessing data quality using statistical and machine learning methods to detect missing values and anomalies. Traffic flow-speed correlation analysis, seasonal-trend decomposition, and unsupervised isolation Forest (iForest) are used to detect anomalies. iForest anomalies are classified as sensor faults and unusual traffic conditions. The iForest algorithm is also trained on additional features, and the model is explained using explainable artificial intelligence. There are 20.16% missing traffic observations, and 50% of the sensors have 15.5% to 33.43% missing values. The average percent of anomalies per sensor is 71.1%, with only a few sensors having less than 10% anomalies. Seasonal-trend decomposition detected 12.6% anomalies in the data of these sensors, and iForest 11.6%, with very few overlaps. To the authors' knowledge, this is the first time a study has explored the quality of dynamic OGD.

Stratification of Length of Stay Prediction following Surgical Cytoreduction in Advanced High-Grade Serous Ovarian Cancer Patients Using Artificial Intelligence; the Leeds L-AI-OS Score.

(1) Background: Length of stay (LOS) has been suggested as a marker of the effectiveness of short-term care. Artificial Intelligence (AI) technologies could help monitor hospital stays. We developed an AI-based novel predictive LOS score for advanced-stage high-grade serous ovarian cancer (HGSOC) patients following cytoreductive surgery and refined factors significantly affecting LOS. (2) Methods: Machine learning and deep learning methods using artificial neural networks (ANN) were used together with conventional logistic regression to predict continuous and binary LOS outcomes for HGSOC patients. The models were evaluated in a post-hoc internal validation set and a Graphical User Interface (GUI) was developed to demonstrate the clinical feasibility of sophisticated LOS predictions. (3) Results: For binary LOS predictions at differential time points, the accuracy ranged between 70-98%. Feature selection identified surgical complexity, pre-surgery albumin, blood loss, operative time, bowel resection with stoma formation, and severe postoperative complications (CD3-5) as independent LOS predictors. For the GUI numerical LOS score, the ANN model was a good estimator for the standard deviation of the LOS distribution by ± two days. (4) Conclusions: We demonstrated the development and application of both quantitative and qualitative AI models to predict LOS in advanced-stage EOC patients following their cytoreduction. Accurate identification of potentially modifiable factors delaying hospital discharge can further inform services performing root cause analysis of LOS.

Metabolism and Bioavailability of Olive Bioactive Constituents Based on In Vitro, In Vivo and Human Studies.

Consumption of olive products has been established as a health-promoting dietary pattern due to their high content in compounds with eminent pharmacological properties and well-described bioactivities. However, their metabolism has not yet been fully described. The present critical review aimed to gather all scientific data of the past two decades regarding the absorption and metabolism of the foremost olive compounds, specifically of the phenylalcohols hydroxytyrosol (HTyr) and tyrosol (Tyr) and the secoiridoids oleacein (Olea), oleocanthal (Oleo) and oleuropein (Oleu). A meticulous record of the in vitro assays and in vivo (animals and humans) studies of the characteristic olive compounds was cited, and a critical discussion on their bioavailability and metabolism was performed taking into account data from their gut microbial metabolism. The existing critical review summarizes the existing knowledge regarding the bioavailability and metabolism of olive-characteristic phenylalchohols and secoiridoids and spotlights the lack of data for specific chemical groups and compounds. Critical observations and conclusions were derived from correlating structure with bioavailability data, while results from in vitro, animal and human studies were compared and discussed, giving significant insight to the future design of research approaches for the total bioavailability and metabolism exploration thereof.

Factors Predicting Surgical Effort Using Explainable Artificial Intelligence in Advanced Stage Epithelial Ovarian Cancer.

(1) Background: Surgical cytoreduction for epithelial ovarian cancer (EOC) is a complex procedure. Encompassed within the performance skills to achieve surgical precision, intra-operative surgical decision-making remains a core feature. The use of eXplainable Artificial Intelligence (XAI) could potentially interpret the influence of human factors on the surgical effort for the cytoreductive outcome in question; (2) Methods: The retrospective cohort study evaluated 560 consecutive EOC patients who underwent cytoreductive surgery between January 2014 and December 2019 in a single public institution. The eXtreme Gradient Boosting (XGBoost) and Deep Neural Network (DNN) algorithms were employed to develop the predictive model, including patient- and operation-specific features, and novel features reflecting human factors in surgical heuristics. The precision, recall, F1 score, and area under curve (AUC) were compared between both training algorithms. The SHapley Additive exPlanations (SHAP) framework was used to provide global and local explainability for the predictive model; (3) Results: A surgical complexity score (SCS) cut-off value of five was calculated using a Receiver Operator Characteristic (ROC) curve, above which the probability of incomplete cytoreduction was more likely (area under the curve [AUC] = 0.644; 95% confidence interval [CI] = 0.598−0.69; sensitivity and specificity 34.1%, 86.5%, respectively; p = 0.000). The XGBoost outperformed the DNN assessment for the prediction of the above threshold surgical effort outcome (AUC = 0.77; 95% [CI] 0.69−0.85; p < 0.05 vs. AUC 0.739; 95% [CI] 0.655−0.823; p < 0.95). We identified "turning points" that demonstrated a clear preference towards above the given cut-off level of surgical effort; in consultant surgeons with <12 years of experience, age <53 years old, who, when attempting primary cytoreductive surgery, recorded the presence of ascites, an Intraoperative Mapping of Ovarian Cancer score >4, and a Peritoneal Carcinomatosis Index >7, in a surgical environment with the optimization of infrastructural support. (4) Conclusions: Using XAI, we explain how intra-operative decisions may consider human factors during EOC cytoreduction alongside factual knowledge, to maximize the magnitude of the selected trade-off in effort. XAI techniques are critical for a better understanding of Artificial Intelligence frameworks, and to enhance their incorporation in medical applications.

Explainable Artificial Intelligence for Prediction of Complete Surgical Cytoreduction in Advanced-Stage Epithelial Ovarian Cancer.

Complete surgical cytoreduction (R0 resection) is the single most important prognosticator in epithelial ovarian cancer (EOC). Explainable Artificial Intelligence (XAI) could clarify the influence of static and real-time features in the R0 resection prediction. We aimed to develop an AI-based predictive model for the R0 resection outcome, apply a methodology to explain the prediction, and evaluate the interpretability by analysing feature interactions. The retrospective cohort finally assessed 571 consecutive advanced-stage EOC patients who underwent cytoreductive surgery. An eXtreme Gradient Boosting (XGBoost) algorithm was employed to develop the predictive model including mostly patient- and surgery-specific variables. The Shapley Additive explanations (SHAP) framework was used to provide global and local explainability for the predictive model. The XGBoost accurately predicted R0 resection (area under curve [AUC] = 0.866; 95% confidence interval [CI] = 0.8−0.93). We identified "turning points" that increased the probability of complete cytoreduction including Intraoperative Mapping of Ovarian Cancer Score and Peritoneal Carcinomatosis Index < 4 and <5, respectively, followed by Surgical Complexity Score > 4, patient's age < 60 years, and largest tumour bulk < 5 cm in a surgical environment of optimized infrastructural support. We demonstrated high model accuracy for the R0 resection prediction in EOC patients and provided novel global and local feature explainability that can be used for quality control and internal audit.

Marine litter on the beaches of the Adriatic and Ionian Seas: An assessment of their abundance, composition and sources.

The abundance, composition and sources of marine litter were determined on beaches located in the seven countries of the Adriatic-Ionian macroregion, namely Albania, Bosnia and Herzegovina, Croatia, Greece, Italy, Montenegro and Slovenia. A total of 70,581 marine litter items were classified and recorded through one-year long surveys carried out in 31 sites. The average litter density of 0.67 items/m2 found within this study is considered to be relatively high. The beaches investigated differed in terms of human-induced pressures; their majority is classified either as semi-urban or semi-rural, while very few beaches could be characterized as urban or remote/natural. The majority of litter items were made of artificial/anthropogenic polymer materials accounting for 91.1% of all litter. Litter from shoreline sources accounted for 33.4% of all litter collected. The amount of litter from sea-based sources ranged in the different countries from 1.54% to 14.84%, with an average of 6.30% at regional level.

The linked medical data access control framework.

The integration of medical data coming from multiple sources is important in clinical research. Amongst others, it enables the discovery of appropriate subjects in patient-oriented research and the identification of innovative results in epidemiological studies. At the same time, the integration of medical data faces significant ethical and legal challenges that impose access constraints. Some of these issues can be addressed by making available aggregated instead of raw record-level data. In many cases however, there is still a need for controlling access even to the resulting aggregated data, e.g., due to data provider's policies. In this paper we present the Linked Medical Data Access Control (LiMDAC) framework that capitalizes on Linked Data technologies to enable controlling access to medical data across distributed sources with diverse access constraints. The LiMDAC framework consists of three Linked Data models, namely the LiMDAC metadata model, the LiMDAC user profile model, and the LiMDAC access policy model. It also includes an architecture that exploits these models. Based on the framework, a proof-of-concept platform is developed and its performance and functionality are evaluated by employing two usage scenarios.