Search

Abstract Artificial intelligence (AI) is revolutionizing the field of medical imaging, holding the potential to shift medicine from a reactive “sick-care” approach to a proactive focus on healthcare and prevention. The successful development of AI in this domain relies on access to large, comprehensive, and standardized real-world datasets that accurately represent diverse populations and diseases. However, images and data are sensitive, and as such, before using them in any way the data needs to be modified to protect the privacy of the patients. This paper explores the approaches in the domain of five EU projects working on the creation of ethically compliant and GDPR-regulated European medical imaging platforms, focused on cancer-related data. It presents the individual approaches to the de-identification of imaging data, and describes the problems and the solutions adopted in each case. Further, lessons learned are provided, enabling future projects to optimally handle the problem of data de-identification. Critical relevance statement This paper presents key approaches from five flagship EU projects for the de-identification of imaging and clinical data offering valuable insights and guidelines in the domain. Key Points ΑΙ models for health imaging require access to large amounts of data. Access to large imaging datasets requires an appropriate de-identification process. This paper provides de-identification guidelines from the AI for health imaging (AI4HI) projects.

Abstract Background Developing trustworthy artificial intelligence (AI) models for clinical applications requires access to clinical and imaging data cohorts. Reusing of publicly available datasets has the potential to fill this gap. Specifically in the domain of breast cancer, a large archive of publicly accessible medical images along with the corresponding clinical data is available at The Cancer Imaging Archive (TCIA). However, existing datasets cannot be directly used as they are heterogeneous and cannot be effectively filtered for selecting specific image types required to develop AI models. This work focuses on the development of a homogenized dataset in the domain of breast cancer including clinical and imaging data. Methods Five datasets were acquired from the TCIA and were harmonized. For the clinical data harmonization, a common data model was developed and a repeatable, documented “extract-transform-load” process was defined and executed for their homogenization. Further, Digital Imaging and COmmunications in Medicine (DICOM) information was extracted from magnetic resonance imaging (MRI) data and made accessible and searchable. Results The resulting harmonized dataset includes information about 2,035 subjects with breast cancer. Further, a platform named RV-Cherry-Picker enables search over both the clinical and diagnostic imaging datasets, providing unified access, facilitating the downloading of all study imaging that correspond to specific series’ characteristics (e.g., dynamic contrast-enhanced series), and reducing the burden of acquiring the appropriate set of images for the respective AI model scenario. Conclusions RV-Cherry-Picker provides access to the largest, publicly available, homogenized, imaging/clinical dataset for breast cancer to develop AI models on top. Relevance statement We present a solution for creating merged public datasets supporting AI model development, using as an example the breast cancer domain and magnetic resonance imaging images. Key points • The proposed platform allows unified access to the largest, homogenized public imaging dataset for breast cancer. • A methodology for the semantically enriched homogenization of public clinical data is presented. • The platform is able to make a detailed selection of breast MRI data for the development of AI models. Graphical Abstract

The exponential growth of Knowledge Graphs necessitates effective and efficient methods for their exploration and understanding. Frequently Asked Questions (FAQ) is a service that typically presents a list of questions and answers related to a specific topic, and which is intended to help people understand that topic. Although FAQ has already shown its value on large websites and is widely used, to the best of our knowledge it has not yet been exploited for Knowledge Graphs. In this paper, we present ULYSSES, the first system for automatically constructing FAQ lists for large Knowledge Graphs. Our method consists of three key steps. First, we select the most frequent queries by exploiting the available query logs. Next, we answer the selected queries, using the original graph. Finally, we construct textual descriptions of both the queries and the corresponding answers, exploring state-of-the-art transformer models, i.e., ChatGPT 3.5 and Gemini 1.5 Pro. We evaluate the results of each model, using a human-constructed FAQ list, contributing a unique dataset to the domain and showing the benefits of our approach.

Background. Despite excellent prognosis of early breast cancer, the patients face problems related to decreased quality of life and mental health. There is a need for easily available interventions targeting modifiable factors related to these problems. The aim of this study was to test the use of a new digital supportive intervention platform for early breast cancer patients. Material and Methods. Ninety-seven early breast cancer patients answered questions on wellbeing, exercise, and sociodemographic factors before systemic adjuvant treatment at the Helsinki University Hospital. Based on these answers and predictive algorithms for anxiety and depression, they were guided onto one or several digital intervention paths. Patients under 56 years of age were guided onto a nutrition path, those who exercised less than the current guideline recommendations onto an exercise path, and those at risk of mental health deterioration onto an empowerment path. Information on compliance was collected at 3 months on the amount of exercise and quality of life using EORTC-C30 scale, anxiety and depression using HADS scale at baseline and 12 months, and log-in information at 3 and 12 months. Results. Thirty-two patients followed the empowerment path, 43 the nutrition path, and 75 the exercise path. On a scale of 1–5, most of the participants (mean = 3.4; SD 0.815) found the interventions helpful and would have recommended testing and supportive interventions to their peers (mean = 3.70; SD 0.961). During the 10-week intervention period, the mean number of log-ins to the empowerment path was 3.69 (SD = 4.24); the nutrition path, 4.32 (SD = 2.891); and the exercise path, 8.33 (SD = 6.293). The higher number of log-ins to the empowerment (rho = 0.531, P=0.008, and n = 24) and exercise paths (rho = 0.330, P=0.01, and n = 59) was related to better global quality of life at one year. The number of log-ins correlated to the weekly amount of exercise in the exercise path (cc 0.740, P value

Abstract Artificial intelligence (AI) is transforming the field of medical imaging and has the potential to bring medicine from the era of ‘sick-care’ to the era of healthcare and prevention. The development of AI requires access to large, complete, and harmonized real-world datasets, representative of the population, and disease diversity. However, to date, efforts are fragmented, based on single–institution, size-limited, and annotation-limited datasets. Available public datasets (e.g., The Cancer Imaging Archive, TCIA, USA) are limited in scope, making model generalizability really difficult. In this direction, five European Union projects are currently working on the development of big data infrastructures that will enable European, ethically and General Data Protection Regulation-compliant, quality-controlled, cancer-related, medical imaging platforms, in which both large-scale data and AI algorithms will coexist. The vision is to create sustainable AI cloud-based platforms for the development, implementation, verification, and validation of trustable, usable, and reliable AI models for addressing specific unmet needs regarding cancer care provision. In this paper, we present an overview of the development efforts highlighting challenges and approaches selected providing valuable feedback to future attempts in the area. Key points • Artificial intelligence models for health imaging require access to large amounts of harmonized imaging data and metadata. • Main infrastructures adopted either collect centrally anonymized data or enable access to pseudonymized distributed data. • Developing a common data model for storing all relevant information is a challenge. • Trust of data providers in data sharing initiatives is essential. • An online European Union meta-tool-repository is a necessity minimizing effort duplication for the various projects in the area.

Abstract Identifying individual patient characteristics that contribute to long-term mental health deterioration following diagnosis of breast cancer (BC) is critical in clinical practice. The present study employed a supervised machine learning pipeline to address this issue in a subset of data from a prospective, multinational cohort of women diagnosed with stage I–III BC with a curative treatment intention. Patients were classified as displaying stable HADS scores (Stable Group; n = 328) or reporting a significant increase in symptomatology between BC diagnosis and 12 months later (Deteriorated Group; n = 50). Sociodemographic, life-style, psychosocial, and medical variables collected on the first visit to their oncologist and three months later served as potential predictors of patient risk stratification. The flexible and comprehensive machine learning (ML) pipeline used entailed feature selection, model training, validation and testing. Model-agnostic analyses aided interpretation of model results at the variable- and patient-level. The two groups were discriminated with a high degree of accuracy (Area Under the Curve = 0.864) and a fair balance of sensitivity (0.85) and specificity (0.87). Both psychological (negative affect, certain coping with cancer reactions, lack of sense of control/positive expectations, and difficulties in regulating negative emotions) and biological variables (baseline percentage of neutrophils, thrombocyte count) emerged as important predictors of mental health deterioration in the long run. Personalized break-down profiles revealed the relative impact of specific variables toward successful model predictions for each patient. Identifying key risk factors for mental health deterioration is an essential first step toward prevention. Supervised ML models may guide clinical recommendations toward successful illness adaptation.

The increase in the size and complexity of large knowledge graphs now available online has resulted in the emergence of many approaches focusing on enabling the quick exploration of the content of those data sources. Structural non-quotient semantic summaries have been proposed in this direction that involve first selecting the most important nodes and then linking them, trying to extract the most useful subgraph out of the original graph. However, the current state of the art systems use costly centrality measures for identifying the most important nodes, whereas even costlier procedures have been devised for linking the selected nodes. In this paper, we address both those deficiencies by first exploiting embeddings for node selection, and then by meticulously selecting approximate algorithms for node linking. Experiments performed over two real-world big KGs demonstrate that the summaries constructed using our method enjoy better quality. Specifically, the coverage scores obtained were 0.8, 0.81, and 0.81 for DBpedia v3.9 and 0.94 for Wikidata dump 2018, across 20%, 25%, and 30% summary sizes, respectively. Additionally, our method can compute orders of magnitude faster than the state of the art.

Abstract A huge amount of imaging data is becoming available worldwide and an incredible range of possible improvements can be provided by artificial intelligence algorithms in clinical care for diagnosis and decision support. In this context, it has become essential to properly manage and handle these medical images and to define which metadata have to be considered, in order for the images to provide their full potential. Metadata are additional data associated with the images, which provide a complete description of the image acquisition, curation, analysis, and of the relevant clinical variables associated with the images. Currently, several data models are available to describe one or more subcategories of metadata, but a unique, common, and standard data model capable of fully representing the heterogeneity of medical metadata has not been yet developed. This paper reports the state of the art on metadata models for medical imaging, the current limitations and further developments, and describes the strategy adopted by the Horizon 2020 “AI for Health Imaging” projects, which are all dedicated to the creation of imaging biobanks.

BackgroundHealth professionals are often faced with the need to identify women at risk of manifesting poor psychological resilience following the diagnosis and treatment of breast cancer. Machine learning algorithms are increasingly used to support clinical decision support (CDS) tools in helping health professionals identify women who are at risk of adverse well-being outcomes and plan customized psychological interventions for women at risk. Clinical flexibility, cross-validated performance accuracy, and model explainability permitting person-specific identification of risk factors are highly desirable features of such tools. ObjectiveThis study aimed to develop and cross-validate machine learning models designed to identify breast cancer survivors at risk of poor overall mental health and global quality of life and identify potential targets of personalized psychological interventions according to an extensive set of clinical recommendations. MethodsA set of 12 alternative models was developed to improve the clinical flexibility of the CDS tool. All models were validated using longitudinal data from a prospective, multicenter clinical pilot at 5 major oncology centers in 4 countries (Italy, Finland, Israel, and Portugal; the Predicting Effective Adaptation to Breast Cancer to Help Women to BOUNCE Back [BOUNCE] project). A total of 706 patients with highly treatable breast cancer were enrolled shortly after diagnosis and before the onset of oncological treatments and were followed up for 18 months. An extensive set of demographic, lifestyle, clinical, psychological, and biological variables measured within 3 months after enrollment served as predictors. Rigorous feature selection isolated key psychological resilience outcomes that could be incorporated into future clinical practice. ResultsBalanced random forest classifiers were successful at predicting well-being outcomes, with accuracies ranging between 78% and 82% (for 12-month end points after diagnosis) and between 74% and 83% (for 18-month end points after diagnosis). Explainability and interpretability analyses built on the best-performing models were used to identify potentially modifiable psychological and lifestyle characteristics that, if addressed systematically in the context of personalized psychological interventions, would be most likely to promote resilience for a given patient. ConclusionsOur results highlight the clinical utility of the BOUNCE modeling approach by focusing on resilience predictors that can be readily available to practicing clinicians at major oncology centers. The BOUNCE CDS tool paves the way for personalized risk assessment methods to identify patients at high risk of adverse well-being outcomes and direct valuable resources toward those most in need of specialized psychological interventions.

Nowadays, in the pursuit of personalized health and well-being, dietary choices are critical. This paper introduces a novel recommendation system designed to provide users with personalized meal plans, consisting of breakfast, lunch, snack, and dinner, in alignment with their health history and preferences from other similar users. More specifically, our system exploits collaborative filtering first to identify other users with similar dietary preferences and uses this information to propose suitable recipes to individuals. The whole process is enhanced by analyzing the individual’s health history, including dietary restrictions, nutritional needs, and specific diet plans, such as low-carb or vegetarian. This ensures that the generated meal plans are not only aligned with the user’s taste but also contribute to the overall wellness of the user. A distinctive feature of our system is its dynamic adaptation feature, which enables users to make real-time adjustments to their meal plans based on their personal constraints and preferences, directly impacting future recommendations. We evaluate the usability of the system through a series of experiments on a large real-world data set of recipes, showing that our system is able to provide highly personalized, dynamic, and accurate recommendations.