Search

In this paper, we tackle the problem of reconstructing earlier tumour configurations starting from a single spatial measurement at a later time. We describe the tumour evolution through a diffuse interface model coupling a Cahn-Hilliard-type equation for the tumour phase field to a reaction-diffusion equation for a key nutrient proportion, also accounting for chemotaxis effects. We stress that the ability to reconstruct earlier tumour states is crucial for calibrating the model used to predict the tumour dynamics and also to identify the areas where the tumour initially began to develop. However, backward-in-time inverse problems are well-known to be severely ill-posed, even for linear parabolic equations. Moreover, we also face additional challenges due to the complexity of a non-linear fourth-order parabolic system. Nonetheless, we can establish uniqueness by using logarithmic convexity methods under suitable a priori assumptions. To further address the ill-posedness of the inverse problem, we propose a Tikhonov regularisation approach that approximates the solution through a family of constrained minimisation problems. For such problems, we analytically derive the first-order necessary optimality conditions. Finally, we develop a computationally efficient numerical approximation of the optimisation problems by employing standard $C^0$-conforming first-order finite elements. We conduct numerical experiments on several pertinent test cases and observe that the proposed algorithm consistently meets expectations, delivering accurate reconstructions of the original ground truth., Comment: 54 pages

In this paper we consider and generalize a model, recently proposed and analytically investigated in its quasi-stationary approximation by the authors, for visco-elasticity with large deformations and conditional compatibility, where the independent variables are the stretch and the rotation tensors. The model takes the form of a system of integro-differential coupled equations. Here, its derivation is generalized to consider mixed boundary conditions, which may represent a wider range of physical applications then the case with Dirichlet boundary conditions considered in our previous contribution. This also introduces nontrivial technical difficulties in the theoretical framework, related to the definition and the regularity of the solutions of elliptic operators with mixed boundary conditions. As a novel contribution, we develop the analysis of the fully non-stationary version of the system where we consider inertia. In this context, we prove the existence of a local in time weak solution in three space dimensions, employing techniques from PDEs and convex analysis., Comment: arXiv admin note: text overlap with arXiv:2307.02992

We introduce a multi-species diffuse interface model for tumor growth, characterized by its incorporation of essential features related to chemotaxis, angiogenesis and proliferation mechanisms. We establish the weak well-posedness of the system within an appropriate variational framework, accommodating various choices for the nonlinear potentials. One of the primary novelties of the work lies in the rigorous establishment of the existence of a weak solution through the introduction of delicate approximation schemes. To our knowledge, this represents a novel advancement for both the intricate Cahn-Hilliard-Keller-Segel system and the Keller-Segel subsystem with source terms. Moreover, when specific conditions are met, such as having more regular initial data, a smallness condition on the chemotactic constant with respect to the magnitude of initial conditions and potentially focusing solely on the two-dimensional case, we provide regularity results for the weak solutions. Finally, we derive a continuous dependence estimate, which, in turn, leads to the uniqueness of the smoothed solution as a natural consequence.

In this article, we introduce the concept of energy-variational solutions for a large class of systems of nonlinear evolutionary partial differential equations. Under certain convexity assumptions, the existence of such solutions can be shown constructively by an adapted minimizing movement scheme. Weak-strong uniqueness follows by a suitable relative energy inequality. Our main motivation is to apply the general framework to viscoelastic fluid models. Therefore, we give a short overview on different versions of such models and their derivation. The abstract result is applied to two of these viscoelastic fluid models in full detail. In the conclusion, we comment on further applications of the general theory and its possible impact.

In this paper we derive a new model for visco-elasticity with large deformations where the independent variables are the stretch and the rotation tensors which intervene with second gradients terms accounting for physical properties in the principle of virtual power. Another basic feature of our model is that there is conditional compatibility, entering the model as kinematic constraints and depending on the magnitude of an internal force associated to dislocations. Moreover, due to the kinematic constraints, the virtual velocities depend on the solutions of the problem. As a consequence, the variational formulation of the problem and the related mathematical analysis are neither standard nor straightforward. We adopt the strategy to invert the kinematic constraints through Green propagators, obtaining a system of integro-differential coupled equations. As a first mathematical step, we develop the analysis of the model in a simplified setting, i.e. considering the quasi-stationary version of the full system where we neglect inertia. In this context, we prove the existence of a global in time strong solution in three space dimensions for the system, employing techniques from PDEs and convex analysis, thus obtaining a novel breakthrough in the field of three-dimensional finite visco-elasticity described in terms of the stretch and rotation variables. We also study a limit problem, letting the magnitude of the internal force associated to dislocations tend to zero, in which case the deformation becomes incompatible and the equations takes the form of a coupled system of PDEs. For the limit problem we obtain global existence, uniqueness and continuous dependence from data in three space dimensions.

In this paper we study a non-local Cahn-Hilliard equation with singular single-well potential and degenerate mobility. This results as a particular case of a more general model derived for a binary, saturated, closed and incompressible mixture, composed by a tumor phase and a healthy phase, evolving in a bounded domain. The general system couples a Darcy-type evolution for the average velocity field with a convective reaction-diffusion type evolution for the nutrient concentration and a non-local convective Cahn-Hilliard equation for the tumor phase. The main mathematical difficulties are related to the proof of the separation property for the tumor phase in the Cahn-Hilliard equation: up to our knowledge, such problem is indeed open in the literature. For this reason, in the present contribution we restrict the analytical study to the Cahn-Hilliard equation only. For the non-local Cahn- Hilliard equation with singular single-well potential and degenerate mobility, we study the existence and uniqueness of weak solutions for spatial dimensions $d\leq 3$. After showing existence, we prove the strict separation property in three spatial dimensions, implying the same property also for lower spatial dimensions, which opens the way to the proof of uniqueness of solutions. Finally, we propose a well posed and gradient stable continuous finite element approximation of the model for $d\leq 3$, which preserves the physical properties of the continuos solution and which is computationally efficient, and we show simulation results in two spatial dimensions which prove the consistency of the proposed scheme and which describe the phase ordering dynamics associated to the system.

Semantic segmentation is a crucial step to extract quantitative information from medical (and, specifically, radiological) images to aid the diagnostic process, clinical follow-up. and to generate biomarkers for clinical research. In recent years, machine learning algorithms have become the primary tool for this task. However, its real-world performance is heavily reliant on the comprehensiveness of training data. Dafne is the first decentralized, collaborative solution that implements continuously evolving deep learning models exploiting the collective knowledge of the users of the system. In the Dafne workflow, the result of each automated segmentation is refined by the user through an integrated interface, so that the new information is used to continuously expand the training pool via federated incremental learning. The models deployed through Dafne are able to improve their performance over time and to generalize to data types not seen in the training sets, thus becoming a viable and practical solution for real-life medical segmentation tasks., Comment: 10 pages (main body), 5 figures. Work partially presented at the 2021 RSNA conference and at the 2023 ISMRM conference In this new version: added author and change in the acknowledgment

We propose a new Cahn-Hilliard phase field model coupled to incompressible viscoelasticity at large strains, obtained from a diffuse interface mixture model and formulated in the Eulerian configuration. A new kind of diffusive regularization, of Allen-Cahn type, is introduced in the transport equation for the deformation gradient, together with a regularizing interface term depending on the gradient of the deformation gradient in the free energy of the system. We study the global existence of a weak solution for the model. While standard diffusive regularizations of the transport equation for the deformation gradient presented in literature allows the existence study only for simplified cases, i.e. in two space dimensions and for convex elastic free energy densities of Neo-Hookean type which are independent from the phase field variable, the present regularization allows to study more general cases. In particular, we obtain the global existence of a weak solution in three space dimensions and for generic nonlinear elastic energy densities with polynomial growth. Our analysis considers elastic free energy densities which depend on the phase field variable and which can possibly degenerate for some values of the phase field variable. By means of an iterative argument based on elliptic regularity bootstrap steps, we find the maximum allowed polynomial growths of the Cahn-Hilliard potential and the elastic energy density which guarantee the existence of a solution in three space dimensions. We propose two unconditionally energy stable finite element approximations of the model, based on convex splitting ideas and on the use of a scalar auxiliary variable, proving the existence and stability of discrete solutions. We finally report numerical results for different test cases with shape memory alloy type free energy with pure phases characterized by different elastic properties.

Young children and their families benefit from early care and education (ECE) settings implementing high-quality practices that support children's well-being, health, and safety. Activities to support quality are a focal point of federal and state investments, including over $1 billion of the Child Care and Development Fund and nearly $240 million in Head Start Training and Technical Assistance. The Culture of Continuous Learning (CCL) project tested the "Breakthrough Series Collaborative" (BSC), a quality improvement methodology developed by the Institute for Healthcare Improvement. As part of the CCL project, ECE centers participating in the BSC learned new strategies for improving evidence-based practices. Participating centers focused on improving teaching practices associated with children's social and emotional learning (SEL). The purpose of this report is to share findings from the BSC feasibility study in ECE centers. The feasibility study examined how ECE centers (i.e., child care centers and Head Start) participated in BSC quality improvement activities and the extent to which the BSC promoted changes in individual practices and organizational dynamics. Findings from the CCL feasibility study can inform recommendations for implementing and evaluating a BSC at a larger scale by including more ECE centers of different types, and across diverse geographic areas. [This report was prepared with University of Massachusetts Boston, Institute for Early Education Leadership and Innovation and JRA Consulting Ltd. For "Promoting a Culture of Continuous Learning in Early Care and Education Settings: A Summary for ECE Leaders. OPRE Report 2021-207," see ED616364.]

Young children and their families benefit from early care and education (ECE) settings that implement high-quality practices that support children's well-being, health, and safety. Activities to support quality are a focal point of federal and state investments, including more than $1 billion of the Child Care and Development Fund and nearly $240 million in Head Start Training and Technical Assistance. The work of quality improvement in ECE is challenging using existing methods, and sustained changes in quality across ECE settings have not been widespread. ECE leaders and staff are seeking new approaches to quality improvement that can promote meaningful and sustained changes in practice. The purpose of this brief is to share findings from the Culture of Continuous Learning (CCL) project, a study of the Breakthrough Series Collaborative (BSC) in ECE programs. The BSC is a quality improvement methodology tested in ECE programs to support practice change by addressing individual and organizational factors associated with sustaining high-quality practices. This brief will provide information for decision makers at the federal, state, and local levels who fund quality improvement strategies. [This brief was prepared with University of Massachusetts Boston, Institute for Early Education Leadership and Innovation and JRA Consulting Ltd. For "Promoting a Culture of Continuous Learning in Early Care and Education Settings. OPRE Report 2021-208," see ED616411.]

We propose a new class of phase field models coupled to viscoelasticity with large deformations, obtained from a diffuse interface mixture model composed by a phase with elastic properties and a liquid phase. The model is formulated in the Eulerian configuration and it is derived by imposing the mass balance for the mixture components and the momentum balance that comes from a generalized form of the principle of virtual powers. The latter considers the presence of a system of microforces and microstresses associated to the microscopic interactions between the mixture's constituents together with a system of macroforces and macrostresses associated to their viscoelastic behavior, taking into account also the friction between the phases. The free energy density of the system is given as the sum of a Cahn-Hilliard term and an elastic polyconvex term, with a coupling between the phase field variable and the elastic deformation gradient in the elastic contribution. General constitutive assumptions complying with a mechanical version of the second law of thermodynamics in isothermal situations are taken. We study the global existence of a weak solution for a simplified and regularized version of the general model, which considers an incompressible elastic free energy of Neo-Hookean type with elastic coefficients depending on the phase field variable. The regularization is properly designed to deal with the coupling between the phase field variable and the elastic deformation gradient in the elastic energy density. The analysis is made both in two and three space dimensions.

We develop a new four-phase tumor growth model with angiogenesis, derived from a diffuse-interface mixture model composed by a viable, a necrotic, a liquid and an angiogenetic component, coupled with two massless chemicals representing a perfectly diluted nutrient and an angiogenetic factor. This model is derived from variational principles complying with the second law of thermodynamics in isothermal situations, starting from biological constitutive assumptions on the tumor cells adhesion properties and on the infiltrative mechanics of tumor-induced vasculature in the tumor tissues, and takes the form of a coupled degenerate Cahn-Hilliard Keller-Segel system for the mixture components with reaction diffusion equations for the chemicals. The model is informed by neuroimaging data, which give informations about the patient-specific brain geometry and tissues microstructure, the distribution of the different tumor components, the white matter fiber orientations and the vasculature density. We describe specific and robust preprocessing steps to extract quantitative informations from the neuroimaging data and to construct a computational platform to solve the model on a patient-specific basis. We introduce a finite element approximation of the model which preserve the qualitative properties of the continuous solutions. Finally, we show simulation results for the patient-specific tumor evolution of a patient affected by GlioBlastoma Multiforme, considering two different test cases before surgery, corresponding to situations with high or low nutrient supply inside the tumor, and a test case after surgery. We show that our model correctly predicts the overall extension of the tumor distribution and the intensity of the angiogenetic process, paving the way for assisting the clinicians in properly assessing the therapy outcomes and in designing optimal patient-specific therapeutic schedules.

Considering its relevance in the field of cryptography, integer factorization is a prominent application where Quantum computers are expected to have a substantial impact. Thanks to Shor's algorithm this peculiar problem can be solved in polynomial time. However, both the number of qubits and applied gates detrimentally affect the ability to run a particular quantum circuit on the near term Quantum hardware. In this work, we help addressing both these problems by introducing a reduced version of Shor's algorithm that proposes a step forward in increasing the range of numbers that can be factorized on noisy Quantum devices. The implementation presented in this work is general and does not use any assumptions on the number to factor. In particular, we have found noteworthy results in most cases, often being able to factor the given number with only one iteration of the proposed algorithm. Finally, comparing the original quantum algorithm with our version on simulator, the outcomes are identical for some of the numbers considered.

The automatic assignment of a severity score to the CT scans of patients affected by COVID-19 pneumonia could reduce the workload in radiology departments. This study aims at exploiting Artificial intelligence (AI) for the identification, segmentation and quantification of COVID-19 pulmonary lesions. We investigated the effects of using multiple datasets, heterogeneously populated and annotated according to different criteria. We developed an automated analysis pipeline, the LungQuant system, based on a cascade of two U-nets. The first one (U-net_1) is devoted to the identification of the lung parenchyma, the second one (U-net_2) acts on a bounding box enclosing the segmented lungs to identify the areas affected by COVID-19 lesions. Different public datasets were used to train the U-nets and to evaluate their segmentation performances, which have been quantified in terms of the Dice index. The accuracy in predicting the CT-Severity Score (CT-SS) of the LungQuant system has been also evaluated. Both Dice and accuracy showed a dependency on the quality of annotations of the available data samples. On an independent and publicly available benchmark dataset, the Dice values measured between the masks predicted by LungQuant system and the reference ones were 0.95$\pm$0.01 and 0.66$\pm$0.13 for the segmentation of lungs and COVID-19 lesions, respectively. The accuracy of 90% in the identification of the CT-SS on this benchmark dataset was achieved. We analysed the impact of using data samples with different annotation criteria in training an AI-based quantification system for pulmonary involvement in COVID-19 pneumonia. In terms of the Dice index, the U-net segmentation quality strongly depends on the quality of the lesion annotations. Nevertheless, the CT-SS can be accurately predicted on independent validation sets, demonstrating the satisfactory generalization ability of the LungQuant.

Parameters in mathematical models for glioblastoma multiforme (GBM) tumour growth are highly patient specific. Here we aim to estimate parameters in a Cahn-Hilliard type diffuse interface model in an optimised way using model order reduction (MOR) based on proper orthogonal decomposition (POD). Based on snapshots derived from finite element simulations for the full order model (FOM) we use POD for dimension reduction and solve the parameter estimation for the reduced order model (ROM). Neuroimaging data are used to define the highly inhomogeneous diffusion tensors as well as to define a target functional in a patient specific manner. The reduced order model heavily relies on the discrete empirical interpolation method (DEIM) which has to be appropriately adapted in order to deal with the highly nonlinear and degenerate parabolic PDEs. A feature of the approach is that we iterate between full order solves with new parameters to compute a POD basis function and sensitivity based parameter estimation for the ROM problems. The algorithm is applied using neuroimaging data for two clinical test cases and we can demonstrate that the reduced order approach drastically decreases the computational effort.

Tissue self-organization into defined and well-controlled three-dimensional structures is essential during development for the generation of organs. A similar, but highly deranged process might also occur during the aberrant growth of cancers, which frequently display a loss of the orderly structures of the tissue of origin, but retain a multicellular organization in the form of spheroids, strands, and buds. The latter structures are often seen when tumors masses switch to an invasive behavior into surrounding tissues. However, the general physical principles governing the self-organized architectures of tumor cell populations remain by and large unclear. In this work, we perform in-vitro experiments to characterize the growth properties of glioblastoma budding emerging from monolayers. Using a theoretical model and numerical tools here we find that such a topological transition is a self-organised, non-equilibrium phenomenon driven by the trade--off of mechanical forces and physical interactions exerted at cell-cell and cell-substrate adhesions. Notably, the unstable disorder states of uncontrolled cellular proliferation macroscopically emerge as complex spatio--temporal patterns that evolve statistically correlated by a universal law.

In this paper, we investigate how semantic relations between concepts extracted from medical documents can be employed to improve the retrieval of medical literature. Semantic relations explicitly represent relatedness between concepts and carry high informative power that can be leveraged to improve the effectiveness of retrieval functionalities of clinical decision support systems. We present preliminary results and show how relations are able to provide a sizable increase of the precision for several topics, albeit having no impact on others. We then discuss some future directions to minimize the impact of negative results while maximizing the impact of good results., Comment: 4 pages, 1 figure, DTMBio-KMH 2018, in conjunction with ACM 27th Conference on Information and Knowledge Management (CIKM), October 22-26 2018, Lingotto, Turin, Italy

This literature review examines the evidence base of the Breakthrough Series Collaborative (BSC) to better understand its outputs and outcomes, as well as the contextual factors that may support success. The review's findings will be incorporated into the next phase of the "Culture of Continuous Learning (CCL) Project," which includes a feasibility study to examine the implementation of the BSC in both Head Start and child care settings. Literature included in this review included peer-reviewed research articles, policy briefs, research reports, and white papers. The literature cited in this review was published between 1998 and 2016. Sections I through V of this report include a mix of literature related to the BSC and other quality improvement methods relevant for early childhood, while Section V includes literature specific to the outcomes related to BSC implementation. The literature search specific to the BSC identified 49 articles, of which 34 included a focus on BSC outcomes. These articles largely focused on studies conducted in the health care field. Overall, the literature examining the implementation and outcomes indicates that the BSC methodology has consistent evidence of improving practices in healthcare, child welfare, and mental health service provision by facilitating the implementation of evidence-based practices (EBP) and building the infrastructure to support quality improvement efforts. The literature suggests that the CCL Project may need to: (1) Support buy-in and engagement in the BSC at multiple levels of the system; (2) Ensure supports for implementation, including training, supervision, and data systems; (3) Develop strategies for building capacity to engage in the improvement process at all levels; (4) Examine ongoing incentives for supporting participation in the initiative; and (5) Identify strategies for trouble-shooting common problems that arise.

A soft solid is said to be initially stressed if it is subjected to a state of internal stress in its unloaded reference configuration. Developing a sound mathematical framework to model initially stressed solids in nonlinear elasticity is key for many applications in engineering and biology. This work investigates the links between the existence of elastic minimizers and the constitutive restrictions for initially stressed materials subjected to finite deformations. In particular, we consider a subclass of constitutive responses in which the strain energy density is taken as a scalar valued function of both the deformation gradient and the initial stress tensor. The main advantage of this approach is that the initial stress tensor belongs to the group of the divergence-free symmetric tensors satisfying the boundary condition in any given reference configuration. However, it is still unclear which physical restrictions must be imposed for the well-posedness of this elastic problem. Assuming that the constitutive response depends on the choice of the reference configuration only through the initial stress tensor, under given conditions we prove the local existence of a relaxed state given by an implicit tensor function of the initial stress distribution. This tensor function is generally not unique, and can be transformed accordingly to the symmetry group of the material at fixed initial stresses. These results allow to extend Ball's existence theorem of elastic minimizers for the proposed constitutive choice of initially stressed materials.

Facebook News Feed personalization algorithm has a significant impact, on a daily basis, on the lifestyle, mood and opinion of millions of Internet users. Nonetheless, the behavior of such algorithm lacks transparency, motivating measurements, modeling and analysis in order to understand and improve its properties. In this paper, we propose a reproducible methodology encompassing measurements, an analytical model and a fairness-based News Feed design. The model leverages the versatility and analytical tractability of time-to-live (TTL) counters to capture the visibility and occupancy of publishers over a News Feed. Measurements are used to parameterize and to validate the expressive power of the proposed model. Then, we conduct a what-if analysis to assess the visibility and occupancy bias incurred by users against a baseline derived from the model. Our results indicate that a significant bias exists and it is more prominent at the top position of the News Feed. In addition, we find that the bias is non-negligible even for users that are deliberately set as neutral with respect to their political views, motivating the proposal of a novel and more transparent fairness-based News Feed design., Comment: IFIP Perfomance 2018 - 36th International Symposium on Computer Performance, Modeling, Measurements and Evaluation

Facebook News Feed personalization algorithm has a significant impact, on a daily basis, on the lifestyle, mood and opinion of millions of Internet users. Nonetheless, the behavior of such algorithms usually lacks transparency, motivating measurements, modeling and analysis in order to understand and improve its properties. In this paper, we propose a reproducible methodology encompassing measurements and an analytical model to capture the visibility of publishers over a News Feed. First, measurements are used to parameterize and to validate the expressive power of the proposed model. Then, we conduct a what-if analysis to assess the visibility bias incurred by the users against a baseline derived from the model. Our results indicate that a significant bias exists and it is more prominent at the top position of the News Feed. In addition, we found that the bias is non-negligible even for users that are deliberately set as neutral with respect to their political views., Comment: International Symposium on Foundations of Open Source Intelligence and Security Informatics (FOSINT-SI 2018). In conjunction with ASONAM 2018. Barcelona, Spain, August 28-31, 2018