86,638 research outputs found
Data reduction and data visualization for automatic diagnosis using gene expression and clinical data
No full textAccurate diagnoses of specific diseases require, in general, the review of the whole medical history of a patient. Currently, even though many advances have been made for disease monitoring, domain experts are still requested to perform direct analyses in order to get a precise classification, thus implying significant efforts and costs. In this work we present a framework for automated diagnosis based on high-dimensional gene expression and clinical data. Given that high-dimensional data can be difficult to analyze and computationally expensive to process, we first perform data reduction to transform high-dimensional representations of data into a lower dimensional space, yet keeping them meaningful for our purposes. We used then different data visualization techniques to embed complex pieces of information in 2-D images, that are in turn used to perform diagnosis relying on deep learning approaches. Experimental analyses show that the proposed method achieves good performance, featuring a prediction Recall value between 91% and 99%
Targeting the Blood–Brain Tumor Barrier with Tumor Necrosis Factor-α
No full textThe blood–brain tumor barrier represents a major obstacle for anticancer drug delivery to brain tumors. Thus, novel strategies aimed at targeting and breaching this structure are of great experimental and clinical interest. This review is primarily focused on the development and use of a derivative of tumor necrosis factor-α (TNF) that can target and alter the blood–brain-tumor-barrier. This drug, called NGR-TNF, consists of a TNF molecule fused to the Cys-Asn-Gly-Arg-Cys-Gly (CNGRCG) peptide (called NGR), a ligand of aminopeptidase N (CD13)-positive tumor blood vessels. Results of preclinical studies suggest that this peptide-cytokine fusion product represents a valuable strategy for delivering TNF to tumor vessels in an amount sufficient to break the biological barriers that restrict drug penetration in cancer lesions. Moreover, clinical studies performed in patients with primary central nervous system lymphoma, have shown that an extremely low dose of NGR-TNF (0.8 μg/m2) is sufficient to promote selective blood–brain-tumor-barrier alteration, increase the efficacy of R-CHOP (a chemo-immunotherapy regimen) and improve patient survival. Besides reviewing these findings, we discuss the potential problems related to the instability and molecular heterogeneity of NGR-TNF and review the various approaches so far developed to obtain more robust and homogeneous TNF derivatives, as well as the pharmacological properties of other peptide/antibody-TNF fusion products, muteins and nanoparticles that are potentially useful for targeting the blood–brain tumor barrier. Compared to other TNF-related drugs, the administration of extremely low-doses of NGR-TNF or its derivatives appear as promising non-immunogenic approaches to overcome TNF counter-regulatory mechanism and systemic toxicity, thereby enabling safe breaking of the BBTB
Understanding Automatic Diagnosis and Classification Processes with Data Visualization
No full textProviding accurate diagnosis of diseases generally requires complex analyses of many clinical, biological and pathological variables. In this context, solutions based on machine learning techniques achieved relevant results in specific disease detection and classification, and can hence provide significant clinical decision support. However, such approaches suffer from the lack of proper means for interpreting the choices made by the models, especially in case of deep-learning ones. In order to improve interpretability and explainability in the process of making qualified decisions, we designed a system that allows for a partial opening of this black box by means of proper investigations on the rationale behind the decisions; this can provide improved understandings into which pre-processing steps are crucial for better performance. We tested our approach over artificial neural networks trained for automatic medical diagnosis based on high-dimensional gene expression and clinical data. Our tool analyzed the internal processes performed by the networks during the classification tasks in order to identify the most important elements involved in the training process that influence the network's decisions.We report the results of an experimental analysis aimed at assessing the viability of the proposed approach
Radiomics-Based Machine Learning Model for Predicting Overall and Progression-Free Survival in Rare Cancer: A Case Study for Primary CNS Lymphoma Patients
No full textPrimary Central Nervous System Lymphoma (PCNSL) is an aggressive neoplasm with a poor prognosis. Although therapeutic progresses have significantly improved Overall Survival (OS), a number of patients do not respond to HD–MTX-based chemotherapy (15–25%) or experience relapse (25–50%) after an initial response. The reasons underlying this poor response to therapy are unknown. Thus, there is an urgent need to develop improved predictive models for PCNSL. In this study, we investigated whether radiomics features can improve outcome prediction in patients with PCNSL. A total of 80 patients diagnosed with PCNSL were enrolled. A patient sub-group, with complete Magnetic Resonance Imaging (MRI) series, were selected for the stratification analysis. Following radiomics feature extraction and selection, different Machine Learning (ML) models were tested for OS and Progression-free Survival (PFS) prediction. To assess the stability of the selected features, images from 23 patients scanned at three different time points were used to compute the Interclass Correlation Coefficient (ICC) and to evaluate the reproducibility of each feature for both original and normalized images. Features extracted from Z-score normalized images were significantly more stable than those extracted from non-normalized images with an improvement of about 38% on average (p-value < (Formula presented.)). The area under the ROC curve (AUC) showed that radiomics-based prediction overcame prediction based on current clinical prognostic factors with an improvement of 23% for OS and 50% for PFS, respectively. These results indicate that radiomics features extracted from normalized MR images can improve prognosis stratification of PCNSL patients and pave the way for further study on its potential role to drive treatment choice
Structural Positional Encoding for knowledge integration in transformer-based medical process monitoring
No full textLogic-based machine learning for transparent ethical agents
No full textAutonomous intelligent agents are increasingly engaging in human communities. Thus, they must be expected to follow social and ethical norms of the community in which they are deployed in. In this work we present an approach for developing such ethical agents which are able to develop ethical decision making and judgment capabilities by learning from interactions with the users. Our approach is a logic-based approach and the resulting ethical agents are transparent by design.Autonomous intelligent agents are increasingly engaging in human communities. Thus, they must be expected to follow social and ethical norms of the community in which they are deployed in. In this work we present an approach for developing such ethical agents which are able to develop ethical decision making and judgment capabilities by learning from interactions with the users. Our approach is a logic-based approach and the resulting ethical agents are transparent by design.Autonomous intelligent agents are increasingly engaging in human communities. Thus, they must be expected to follow social and ethical norms of the community in which they are deployed in. In this work we present an approach for developing such ethical agents which are able to develop ethical decision making and judgment capabilities by learning from interactions with the users. Our approach is a logic-based approach and the resulting ethical agents are transparent by design
Efficient compliance checking of RDF data
No full textAutomated compliance checking, i.e. the task of automatically assessing whether states of affairs comply with normative systems, has recently received a lot of attention from the scientific community, also as a consequence of the increasing investments in Artificial Intelligence technologies for the legal domain (LegalTech). The authors of this paper deem as crucial the research and implementation of compliance checkers that can directly process data in RDF format, as nowadays more and more (big) data in this format are becoming available worldwide, across a multitude of different domains. Among the automated technologies that have been used in recent literature, to the best of our knowledge, only two of them have been evaluated with input states of affairs encoded in RDF format. This paper formalizes a selected use case in these two technologies and compares the implementations, also in terms of simulations with respect to shared synthetic datasets
Reasoning on Information Term Semantics with ASP for Constructive ELꓕ
No full textConstructive description logics represent different re-interpretations of description logics (DLs) under constructive semantics. Constructive description logics have been mostly studied for their formal properties, while limited practical approaches have been shown for their use in Knowledge Representation languages and tools (which, on the other hand, constitute the distinctive applications of description logics). To address this aspect, we recently studied the relation of constructive DLs based on Information Term semantics with Answer Set semantics in the context of the positive logic EL. In this paper we continue this study in the direction of more expressive DLs by considering the introduction of negative information, leading to a constructive interpretation for the DL EL⊥. We show that formal results linking the constructive semantics to answer set semantics can be extended to the case of negative information in EL⊥
Ensuring trustworthy and ethical behavior in intelligent logical agents?
No full textAutonomous Intelligent Agents are employed in many important autonomous applications upon which the life and welfare of living beings and vital social functions may depend. Therefore, agents should be trustworthy. Apriori certification techniques can be useful, but are not sufficient for agents that evolve, and thus modify their epistemic and belief state. In this paper we propose/ refine/extend techniques for run-time assurance, based upon introspective self-monitoring and checking. The aim is to build a 'toolkit' to allow an agent designer/developer to ensure trustworthy and ethical behavior
- …
