AUCO ResNet: an end-to-end network for Covid-19 pre-screening from cough and breath.

This study presents the Auditory Cortex ResNet (AUCO ResNet), it is a biologically inspired deep neural network especially designed for sound classification and more specifically for Covid-19 recognition from audio tracks of coughs and breaths. Differently from other approaches, it can be trained end-to-end thus optimizing (with gradient descent) all the modules of the learning algorithm: mel-like filter design, feature extraction, feature selection, dimensionality reduction and prediction. This neural network includes three attention mechanisms namely the squeeze and excitation mechanism, the convolutional block attention module, and the novel sinusoidal learnable attention. The attention mechanism is able to merge relevant information from activation maps at various levels of the network. The net takes as input raw audio files and it is able to fine tune also the features extraction phase. In fact, a Mel-like filter is designed during the training, thus adapting filter banks on important frequencies. AUCO ResNet has proved to provide state of art results on many datasets. Firstly, it has been tested on many datasets containing Covid-19 cough and breath. This choice is related to the fact that that cough and breath are language independent, allowing for cross dataset tests with generalization aims. These tests demonstrate that the approach can be adopted as a low cost, fast and remote Covid-19 pre-screening tool. The net has also been tested on the famous UrbanSound 8K dataset, achieving state of the art accuracy without any data preprocessing or data augmentation technique.

Vehicular Traffic Congestion Classification by Visual Features and Deep Learning Approaches: A Comparison.

Automatic traffic flow classification is useful to reveal road congestions and accidents. Nowadays, roads and highways are equipped with a huge amount of surveillance cameras, which can be used for real-time vehicle identification, and thus providing traffic flow estimation. This research provides a comparative analysis of state-of-the-art object detectors, visual features, and classification models useful to implement traffic state estimations. More specifically, three different object detectors are compared to identify vehicles. Four machine learning techniques are successively employed to explore five visual features for classification aims. These classic machine learning approaches are compared with the deep learning techniques. This research demonstrates that, when methods and resources are properly implemented and tested, results are very encouraging for both methods, but the deep learning method is the most accurately performing one reaching an accuracy of 99.9% for binary traffic state classification and 98.6% for multiclass classification.

Enhancing early Parkinson's disease detection through multimodal deep learning and explainable AI: insights from the PPMI database.

Parkinson's is the second most common neurodegenerative disease, affecting nearly 8.5M people and steadily increasing. In this research, Multimodal Deep Learning is investigated for the Prodromal stage detection of Parkinson's Disease (PD), combining different 3D architectures with the novel Excitation Network (EN) and supported by Explainable Artificial Intelligence (XAI) techniques. Utilizing data from the Parkinson's Progression Markers Initiative, this study introduces a joint co-learning approach for multimodal fusion, enabling end-to-end training of deep neural networks and facilitating the learning of complementary information from both imaging and clinical modalities. DenseNet with EN outperformed other models, showing a substantial increase in accuracy when supplemented with clinical data. XAI methods, such as Integrated Gradients for ResNet and DenseNet, and Attention Heatmaps for Vision Transformer (ViT), revealed that DenseNet focused on brain regions believed to be critical to prodromal pathophysiology, including the right temporal and left pre-frontal areas. Similarly, ViT highlighted the lateral ventricles associated with cognitive decline, indicating their potential in the Prodromal stage. These findings underscore the potential of these regions as early-stage PD biomarkers and showcase the proposed framework's efficacy in predicting subtypes of PD and aiding in early diagnosis, paving the way for innovative diagnostic tools and precision medicine.

An Interpretable Adaptive Multiscale Attention Deep Neural Network for Tabular Data.

Deep learning (DL) has been demonstrated to be a valuable tool for analyzing signals such as sounds and images, thanks to its capabilities of automatically extracting relevant patterns as well as its end-to-end training properties. When applied to tabular structured data, DL has exhibited some performance limitations compared to shallow learning techniques. This work presents a novel technique for tabular data called adaptive multiscale attention deep neural network architecture (also named excited attention). By exploiting parallel multilevel feature weighting, the adaptive multiscale attention can successfully learn the feature attention and thus achieve high levels of F1-score on seven different classification tasks (on small, medium, large, and very large datasets) and low mean absolute errors on four regression tasks of different size. In addition, adaptive multiscale attention provides four levels of explainability (i.e., comprehension of its learning process and therefore of its outcomes): 1) calculates attention weights to determine which layers are most important for given classes; 2) shows each feature's attention across all instances; 3) understands learned feature attention for each class to explore feature attention and behavior for specific classes; and 4) finds nonlinear correlations between co-behaving features to reduce dataset dimensionality and improve interpretability. These interpretability levels, in turn, allow for employing adaptive multiscale attention as a useful tool for feature ranking and feature selection.

Integrating human expertise & automated methods for a dynamic and multi-parametric evaluation of large language models' feasibility in clinical decision-making.

BACKGROUND: Recent enhancements in Large Language Models (LLMs) such as ChatGPT have exponentially increased user adoption. These models are accessible on mobile devices and support multimodal interactions, including conversations, code generation, and patient image uploads, broadening their utility in providing healthcare professionals with real-time support for clinical decision-making. Nevertheless, many authors have highlighted serious risks that may arise from the adoption of LLMs, principally related to safety and alignment with ethical guidelines. OBJECTIVE: To address these challenges, we introduce a novel methodological approach designed to assess the specific feasibility of adopting LLMs within a healthcare area, with a focus on clinical nursing, evaluating their performance and thereby directing their choice. Emphasizing LLMs' adherence to scientific advancements, this approach prioritizes safety and care personalization, according to the "Organization for Economic Co-operation and Development" frameworks for responsible AI. Moreover, its dynamic nature is designed to adapt to future evolutions of LLMs. METHOD: Through integrating advanced multidisciplinary knowledge, including Nursing Informatics, and aided by a prospective literature review, seven key domains and specific evaluation items were identified as follows:A Peer Review by experts in Nursing and AI was performed, ensuring scientific rigor and breadth of insights for an essential, reproducible, and coherent methodological approach. By means of a 7-point Likert scale, thresholds are defined in order to classify LLMs as "unusable", "usable with high caution", and "recommended" categories. Nine state of the art LLMs were evaluated using this methodology in clinical oncology nursing decision-making, producing preliminary results. Gemini Advanced, Anthropic Claude 3 and ChatGPT 4 achieved the minimum score of the State of the Art Alignment & Safety domain for classification as "recommended", being also endorsed across all domains. LLAMA 3 70B and ChatGPT 3.5 were classified as "usable with high caution." Others were classified as unusable in this domain. CONCLUSION: The identification of a recommended LLM for a specific healthcare area, combined with its critical, prudent, and integrative use, can support healthcare professionals in decision-making processes.

Multi-speed transformer network for neurodegenerative disease assessment and activity recognition.

BACKGROUND AND OBJECTIVE: Neurodegenerative diseases are the most frequent age-related diseases. This type of disease, if not discovered in the initial stage, will compromise the quality of life of the affected subject. Thus, a timely diagnosis is of paramount importance. One of the most used tasks from neurologists to detect and determine the severity of the disease is analysing human gait. This work presents the dataset named "Beside Gait" containing timeseries of coordinates of extracted body joints of people with neurodegenerative diseases in various stages of the disease as well as control subjects. In addition, the novel Multi-Speed transformer technique will be presented and benchmarked against several other techniques making use of deep learning and Shallow Learning. The objective is to recognize subjects affected by some form of neurodegenerative disease in early stage using a computer vision technique making use of deep learning that can be integrated into a smartphone app for offline inference with the aim of promptly initiate investigations and treatment to improve the patient's quality of life. METHODS: The recorded videos were processed, and the skeleton of the person in the video was extracted using pose estimation. The raw time-series coordinates of the joints extracted by the pose estimation algorithm were tested against novel deep neural network architectures and Shallow Learning techniques. In this work, the proposed Multi-Speed Transformer is benchmarked against other deep neural networks such as Temporal Convolutional Neural Networks, Transformers, as well as Shallow Learning techniques making use of feature extraction and different classifiers such as Random Forests, K Nearest Neighbours, Ada Boost, Linear and RBF SVM. The proposed Multi-Speed Transformer architecture has been developed to learn short and long-term patterns to model the various pathological gaits. RESULTS: The Multi-Speed Transformer outperformed all other existing models reaching an accuracy of 96.9%, a sensitivity of 96.9%, a precision of 97.7%, and a specificity of 97.1% in binary classification. The accuracy in multi-class classification for detecting the presence of the disease in various stages is 71.6%, the sensitivity is 67.7%, and the specificity is 71.8%. In addition, tests have also been conducted against two other different activity recognition datasets, namely SHREC and JHMDB, in the exact same conditions. Multi-Speed Transformer has demonstrated to beat always all other tested techniques as well as the techniques reviewed in the state-of-the-art with respectively of accuracy 91.8% and 74%. Having those datasets more than two classes, specificity was not computed. CONCLUSIONS: The Multi-Speed Transformer is a valuable technique for neurodegenerative disease assessment through computer vision. In addition, the novel dataset "Beside Gait" here presented is an important starting point for future research work on automatic recognition of neurodegenerative diseases using gait analysis.

Human Gait Analysis in Neurodegenerative Diseases: A Review.

This paper reviews the recent literature on technologies and methodologies for quantitative human gait analysis in the context of neurodegenerative diseases. The use of technological instruments can be of great support in both clinical diagnosis and severity assessment of these pathologies. In this paper, sensors, features and processing methodologies have been reviewed in order to provide a highly consistent work that explores the issues related to gait analysis. First, the phases of the human gait cycle are briefly explained, along with some non-normal gait patterns (gait abnormalities) typical of some neurodegenerative diseases. Then the paper reports the most common processing techniques for both feature selection and extraction and for classification and clustering. Finally, a conclusive discussion on current open problems and future directions is outlined.