Exploring Nutritional Influence on Blood Glucose Forecasting for Type 1 Diabetes Using Explainable AI.

Type 1 diabetes mellitus (T1DM) is characterized by insulin deficiency and blood sugar control issues. The state-of-the-art solution is the artificial pancreas (AP), which integrates basal insulin delivery and glucose monitoring. However, APs are unable to manage postprandial glucose response (PGR) due to limited knowledge of its determinants, requiring additional information for accurate bolus delivery, such as estimated carbohydrate intake. This study aims to quantify the influence of various meal-related factors on predicting postprandial blood glucose levels (BGLs) at different time intervals (15 min, 60 min, and 120 min) after meals by using deep neural network (DNN) models. The prediction models incorporate preprandial blood glucose values, insulin dosage, and various meal-related nutritional factors such as intake of energy, carbohydrates, proteins, lipids, fatty acids, fibers, glycemic index, and glycemic load as input variables. The impact of input features was assessed by exploiting eXplainable Artificial Intelligence (XAI) methodologies, specifically SHapley Additive exPlanations (SHAP), which provide insights into each feature's contribution to the model predictions. By leveraging XAI methodologies, this study aims to enhance the interpretability and transparency of BGL prediction models and validate clinical literature hypotheses. The findings can aid in the development of decision-support tools for individuals with T1DM, facilitating PGR management and reducing the risks of adverse events. The improved understanding of PGR determinants may lead to advancements in AP technology and improve the overall quality of life for T1DM patients.

Assessment of blood perfusion quality in laparoscopic colorectal surgery by means of Machine Learning.

An innovative algorithm to automatically assess blood perfusion quality of the intestinal sector in laparoscopic colorectal surgery is proposed. Traditionally, the uniformity of the brightness in indocyanine green-based fluorescence consists only in a qualitative, empirical evaluation, which heavily relies on the surgeon's subjective assessment. As such, this leads to assessments that are strongly experience-dependent. To overcome this limitation, the proposed algorithm assesses the level and uniformity of indocyanine green used during laparoscopic surgery. The algorithm adopts a Feed Forward Neural Network receiving as input a feature vector based on the histogram of the green band of the input image. It is used to (i) acquire information related to perfusion during laparoscopic colorectal surgery, and (ii) support the surgeon in assessing objectively the outcome of the procedure. In particular, the algorithm provides an output that classifies the perfusion as adequate or inadequate. The algorithm was validated on videos captured during surgical procedures carried out at the University Hospital Federico II in Naples, Italy. The obtained results show a classification accuracy equal to [Formula: see text], with a repeatability of [Formula: see text]. Finally, the real-time operation of the proposed algorithm was tested by analyzing the video streaming captured directly from an endoscope available in the OR.

A survey on modern trainable activation functions.

In neural networks literature, there is a strong interest in identifying and defining activation functions which can improve neural network performance. In recent years there has been a renovated interest in the scientific community in investigating activation functions which can be trained during the learning process, usually referred to as trainable, learnable or adaptable activation functions. They appear to lead to better network performance. Diverse and heterogeneous models of trainable activation function have been proposed in the literature. In this paper, we present a survey of these models. Starting from a discussion on the use of the term "activation function" in literature, we propose a taxonomy of trainable activation functions, highlight common and distinctive proprieties of recent and past models, and discuss main advantages and limitations of this type of approach. We show that many of the proposed approaches are equivalent to adding neuron layers which use fixed (non-trainable) activation functions and some simple local rule that constrains the corresponding weight layers.

A framework to identify structured behavioral patterns within rodent spatial trajectories.

Animal behavior is highly structured. Yet, structured behavioral patterns-or "statistical ethograms"-are not immediately apparent from the full spatiotemporal data that behavioral scientists usually collect. Here, we introduce a framework to quantitatively characterize rodent behavior during spatial (e.g., maze) navigation, in terms of movement building blocks or motor primitives. The hypothesis that we pursue is that rodent behavior is characterized by a small number of motor primitives, which are combined over time to produce open-ended movements. We assume motor primitives to be organized in terms of two sparsity principles: each movement is controlled using a limited subset of motor primitives (sparse superposition) and each primitive is active only for time-limited, time-contiguous portions of movements (sparse activity). We formalize this hypothesis using a sparse dictionary learning method, which we use to extract motor primitives from rodent position and velocity data collected during spatial navigation, and successively to reconstruct past trajectories and predict novel ones. Three main results validate our approach. First, rodent behavioral trajectories are robustly reconstructed from incomplete data, performing better than approaches based on standard dimensionality reduction methods, such as principal component analysis, or single sparsity. Second, the motor primitives extracted during one experimental session generalize and afford the accurate reconstruction of rodent behavior across successive experimental sessions in the same or in modified mazes. Third, in our approach the number of motor primitives associated with each maze correlates with independent measures of maze complexity, hence showing that our formalism is sensitive to essential aspects of task structure. The framework introduced here can be used by behavioral scientists and neuroscientists as an aid for behavioral and neural data analysis. Indeed, the extracted motor primitives enable the quantitative characterization of the complexity and similarity between different mazes and behavioral patterns across multiple trials (i.e., habit formation). We provide example uses of this computational framework, showing how it can be used to identify behavioural effects of maze complexity, analyze stereotyped behavior, classify behavioral choices and predict place and grid cell displacement in novel environments.

Evidence for sparse synergies in grasping actions.

Converging evidence shows that hand-actions are controlled at the level of synergies and not single muscles. One intriguing aspect of synergy-based action-representation is that it may be intrinsically sparse and the same synergies can be shared across several distinct types of hand-actions. Here, adopting a normative angle, we consider three hypotheses for hand-action optimal-control: sparse-combination hypothesis (SC) - sparsity in the mapping between synergies and actions - i.e., actions implemented using a sparse combination of synergies; sparse-elements hypothesis (SE) - sparsity in synergy representation - i.e., the mapping between degrees-of-freedom (DoF) and synergies is sparse; double-sparsity hypothesis (DS) - a novel view combining both SC and SE - i.e., both the mapping between DoF and synergies and between synergies and actions are sparse, each action implementing a sparse combination of synergies (as in SC), each using a limited set of DoFs (as in SE). We evaluate these hypotheses using hand kinematic data from six human subjects performing nine different types of reach-to-grasp actions. Our results support DS, suggesting that the best action representation is based on a relatively large set of synergies, each involving a reduced number of degrees-of-freedom, and that distinct sets of synergies may be involved in distinct tasks.

Neural networks with non-uniform embedding and explicit validation phase to assess Granger causality.

A challenging problem when studying a dynamical system is to find the interdependencies among its individual components. Several algorithms have been proposed to detect directed dynamical influences between time series. Two of the most used approaches are a model-free one (transfer entropy) and a model-based one (Granger causality). Several pitfalls are related to the presence or absence of assumptions in modeling the relevant features of the data. We tried to overcome those pitfalls using a neural network approach in which a model is built without any a priori assumptions. In this sense this method can be seen as a bridge between model-free and model-based approaches. The experiments performed will show that the method presented in this work can detect the correct dynamical information flows occurring in a system of time series. Additionally we adopt a non-uniform embedding framework according to which only the past states that actually help the prediction are entered into the model, improving the prediction and avoiding the risk of overfitting. This method also leads to a further improvement with respect to traditional Granger causality approaches when redundant variables (i.e. variables sharing the same information about the future of the system) are involved. Neural networks are also able to recognize dynamics in data sets completely different from the ones used during the training phase.

A Programmer-Interpreter Neural Network Architecture for Prefrontal Cognitive Control.

There is wide consensus that the prefrontal cortex (PFC) is able to exert cognitive control on behavior by biasing processing toward task-relevant information and by modulating response selection. This idea is typically framed in terms of top-down influences within a cortical control hierarchy, where prefrontal-basal ganglia loops gate multiple input-output channels, which in turn can activate or sequence motor primitives expressed in (pre-)motor cortices. Here we advance a new hypothesis, based on the notion of programmability and an interpreter-programmer computational scheme, on how the PFC can flexibly bias the selection of sensorimotor patterns depending on internal goal and task contexts. In this approach, multiple elementary behaviors representing motor primitives are expressed by a single multi-purpose neural network, which is seen as a reusable area of "recycled" neurons (interpreter). The PFC thus acts as a "programmer" that, without modifying the network connectivity, feeds the interpreter networks with specific input parameters encoding the programs (corresponding to network structures) to be interpreted by the (pre-)motor areas. Our architecture is validated in a standard test for executive function: the 1-2-AX task. Our results show that this computational framework provides a robust, scalable and flexible scheme that can be iterated at different hierarchical layers, supporting the realization of multiple goals. We discuss the plausibility of the "programmer-interpreter" scheme to explain the functioning of prefrontal-(pre)motor cortical hierarchies.

From motor to sensory processing in mirror neuron computational modelling.

Typical patterns of hand-joint covariation arising in the context of grasping actions enable one to provide simplified descriptions of these actions in terms of small sets of hand-joint parameters. The computational model of mirror mechanisms introduced here hypothesizes that mirror neurons are crucially involved in coding and making this simplified motor information available for both action recognition and control processes. In particular, grasping action recognition processes are modeled in terms of a visuo-motor loop enabling one to make iterated use of mirror-coded motor information. In simulation experiments concerning the classification of reach-to-grasp actions, mirror-coded information was found to simplify the processing of visual inputs and to improve action recognition results with respect to recognition procedures that are solely based on visual processing. The visuo-motor loop involved in action recognition is a distinctive feature of this model which is coherent with the direct matching hypothesis. Moreover, the visuo-motor loop sets the model introduced here apart from those computational models that identify mirror neuron activity in action observation with the final outcome of computational processes unidirectionally flowing from sensory (and usually visual) to motor systems.

A connectionist architecture for view-independent grip-aperture computation.

This paper addresses the problem of extracting view-invariant visual features for the recognition of object-directed actions and introduces a computational model of how these visual features are processed in the brain. In particular, in the test-bed setting of reach-to-grasp actions, grip aperture is identified as a good candidate for inclusion into a parsimonious set of hand high-level features describing overall hand movement during reach-to-grasp actions. The computational model NeGOI (neural network architecture for measuring grip aperture in an observer-independent way) for extracting grip aperture in a view-independent fashion was developed on the basis of functional hypotheses about cortical areas that are involved in visual processing. An assumption built into NeGOI is that grip aperture can be measured from the superposition of a small number of prototypical hand shapes corresponding to predefined grip-aperture sizes. The key idea underlying the NeGOI model is to introduce view-independent units (VIP units) that are selective for prototypical hand shapes, and to integrate the output of VIP units in order to compute grip aperture. The distinguishing traits of the NEGOI architecture are discussed together with results of tests concerning its view-independence and grip-aperture recognition properties. The overall functional organization of NEGOI model is shown to be coherent with current functional models of the ventral visual stream, up to and including temporal area STS. Finally, the functional role of the NeGOI model is examined from the perspective of a biologically plausible architecture which provides a parsimonious set of high-level and view-independent visual features as input to mirror systems.