196,013 research outputs found
Counterfactual Building and Evaluation via eXplainable Support Vector Data Description
No full textIncreasingly in recent times, the mere prediction of a machine learning algorithm is considered insufficient to gain complete control over the event being predicted. A machine learning algorithm should be considered reliable in the way it allows to extract more knowledge and information than just having a prediction at hand. In this perspective, the counterfactual theory plays a central role. By definition, a counterfactual is the smallest variation of the input such that it changes the predicted behaviour. The paper addresses counterfactuals through Support Vector Data Description (SVDD), empowered by explainability and metric for assessing the counterfactual quality. After showing the specific case in which an analytical solution may be found (under Euclidean distance and linear kernel), an optimisation problem is posed for any type of distances and kernels. The vehicle platooning application is the use case considered to demonstrate how the outlined methodology may offer support to safety-critical applications as well as how explanation may shed new light into the control of the system at hand
Counterfactual Building and Evaluation via eXplainable Support Vector Data Description
No full textIncreasingly in recent times, the mere prediction of a machine learning algorithm is considered insufficient to gain complete control over the event being predicted. A machine learning algorithm should be considered reliable in the way it allows to extract more knowledge and information than just having a prediction at hand. In this perspective, the counterfactual theory plays a central role. By definition, a counterfactual is the smallest variation of the input such that it changes the predicted behaviour. The paper addresses counterfactuals through Support Vector Data Description (SVDD), empowered by explainability and metric for assessing the counterfactual quality. After showing the specific case in which an analytical solution may be found (under Euclidean distance and linear kernel), an optimisation problem is posed for any type of distances and kernels. The vehicle platooning application is the use case considered to demonstrate how the outlined methodology may offer support to safety-critical applications as well as how explanation may shed new light into the control of the system at hand
Effects of orthostatic stress on forearm endothelial function in normal subjects and in patients with hypertension, diabetes, or both diseases
No full textBackground: Sympathetically mediated vasoconstriction, to compensate for reduced venous return and cardiac output, characterizes the circulatory adaptation to head-up tilting (HUT). It has not been clarified whether this is coupled with a modulating endothelial vasorelaxation and whether diseases causing endothelial dysfunction, such as diabetes and hypertension, may impair this counterregulatory mechanism. Methods: In patients with hypertension (group 1), diabetes (group 2), or both diseases (group 3) and in healthy control subjects (12 subjects per group) we investigated the brachial artery vasodilating response to the release of distal circulatory arrest (DCA) while they were supine and during 60 degrees HUT. Results: The supine increase in lumen was smaller (P <.01) in groups 1 (+4.5% +/- 1.5%), 2 (+4.8% +/- 1.4%), and 3 (+3.9% +/- 1.3%) than in the control group (+8.6% +/- 1.6%). Vasorelaxation by nitroglycerin was similar in each population. During HUT, the lumen response to DCA was enhanced (P <.01 v supine) in control subjects (+15.4% +/- 2.5%) and group 1 (+ 10.0 +/- 2.4%) and was reduced (P <.01 v supine) in groups 2 (+2.9% +/- 0.5%) and 3 (+2.1% +/- 0.4%), even though the hyperemic reaction to DCA was similar. The ratio of lumen changes to changes in flow (mm/mL/min X 1000) during reactive hyperemia to DCA increased (P <.01) with HUT, compared with that in the supine position, in control subjects (1.75 v 1.19) and group 1 (1.61 v 0.95), and decreased (P <.01) in groups 2 (0.62 v 0.87) and 3 (0.48 v 0.77). Conclusions: The HUT posture is characterized by an increased endothelium-dependent, flow-mediated vasodilation as a possible modulator of the neural vasoconstriction. This effect is persistent but blunted in hypertension and is abolished in diabetes, either alone or in association with high BP. Thus, vasoconstrictor factors could remain unmodulated during an event such as orthostasis, making the risk posed by these disorders more critical
Evaluation of a novel speech-in-noise test for hearing screening : classification performance and transducers' characteristics
No full textOne of the current gaps in teleaudiology is the lack of methods for adult hearing screening viable for use in individuals of unknown language and in varying environments. We have developed a novel automated speech-in-noise test that uses stimuli viable for use in non-native listeners. The test reliability has been demonstrated in laboratory settings and in uncontrolled environmental noise settings in previous studies. The aim of this study was: (i) to evaluate the ability of the test to identify hearing loss using multivariate logistic regression classifiers in a population of 148 unscreened adults and (ii) to evaluate the ear-level sound pressure levels generated by different earphones and headphones as a function of the test volume. The multivariate classifiers had sensitivity equal to 0.79 and specificity equal to 0.79 using both the full set of features extracted from the test as well as a subset of three features (speech recognition threshold, age, and number of correct responses). The analysis of the ear-level sound pressure levels showed substantial variability across transducer types and models, with earphones levels being up to 22 dB lower than those of headphones. Overall, these results suggest that the proposed approach might be viable for hearing screening in varying environments if an option to self-adjust the test volume is included and if headphones are used. Future research is needed to assess the viability of the test for screening at a distance, for example by addressing the influence of user interface, device, and settings, on a large sample of subjects with varying hearing loss.sponsorship: This work was supported in part by the Capita Foundation through Project WHISPER, Widespread Hearing Impairment Screening and PrEvention of Risk (2020 Auditory Research Grant) and in part by the European Research Council under the European Union's Horizon 2020 Research and Innovation Program or ERC Consolidator under Grant SONORA 773268. (Marco Zanet and Edoardo M. Polo authors contributed equally to this work.) (Capita Foundation through Project WHISPER, Widespread Hearing Impairment Screening and PrEvention of Risk (2020 Auditory Research Grant), European Research Council under the European Union's Horizon 2020 Research and Innovation Program or ERC Consolidator under Grant SONORA|773268)status: Publishe
Analysis of the Effect of Emotion Elicitation on the Cardiovascular System
No full textEmotions play an important role in our everyday life, influencing our decision-making process, and also affecting our physiology. Several studies in literature have proposed successful classification models for emotion recognition combining multimodal physiological measures without dwelling on the physiological significance of the measures. Our study aims at finding cardiovascular indices related to the autonomic nervous system that can explain how autonomic control of the heart responds with respect to specific emotions: happiness, fear, relaxation and boredom. Pulse arrival time and pulse pressure measurements have been shown to be significantly separating the 4 emotions, especially along the arousal dimension as expected from previous findings. Importantly, these blood pressure related indices also yielded relevant insights into characterizing the valence dimension when looking at high and low arousal subsets. In addition, these measures were found to be correlated with classical autonomic indices and explanatory in the cardiovascular and autonomic changes elicited by different emotions. Autonomic indices were then used to train a basic support vector machine model obtaining four-class test accuracy in discriminating happiness, relaxation, boredom and fear equal to 44%, 67%, 55%, 44% respectively
Emotion recognition from multimodal physiological measurements based on an interpretable feature selection method
No full textMany studies in literature successfully use classification algorithms to classify emotions by means of physiological signals. However, there are still important limitations in interpretability of the results, i.e. lack of feature specific characterizations for each emotional state. To this extent, our study proposes a feature selection method that allows to determine the most informative subset of features extracted from physiological signals by maintaining their original dimensional space. Results show that features from the galvanic skin response are confirmed to be relevant in separating the arousal dimension, especially fear from happiness and relaxation. Furthermore, the average and the median value of the galvanic skin response signal together with the ratio between SD1 and SD2 from the Poincarè analysis of the electrocardiogram signal, were found to be the most important features for the discrimination along the valence dimension. A Linear Discriminant Analysis model using the first ten features sorted by importance, as defined by their ability to discriminate emotions with a bivariate approach, led to a three-class test accuracy in discriminating happiness, relaxation and fear equal to 72%, 67% and 89% respectively.Clinical relevance This study demonstrates the ability of physiological signals to assess the emotional state of different subjects, by providing a fast and efficient method to select most important indexes from the autonomic nervous system. The approach has high clinical relevance as it could be extended to assess other emotional states (e.g. stress and pain) characterizing pathological states such as post traumatic stress disorder and depression
A novel method to derive personalized minimum viable recommendations for type 2 diabetes prevention based on counterfactual explanations
No full textDespite the growing availability of artificial intelligence models for predicting type 2 diabetes, there is still a lack of personalized approaches to quantify minimum viable changes in biomarkers that may help reduce the individual risk of developing disease. The aim of this article is to develop a new method, based on counterfactual explanations, to generate personalized recommendations to reduce the one-year risk of type 2 diabetes. Ten routinely collected biomarkers extracted from Electronic Medical Records of 2791 patients at low risk and 2791 patients at high risk of type 2 diabetes were analyzed. Two regions characterizing the two classes of patients were estimated using a Support Vector Data Description classifier. Counterfactual explanations (i.e., minimal changes in input features able to change the risk class) were generated for patients at high risk and evaluated using performance metrics (availability, validity, actionability, similarity, and discriminative power) and a qualitative survey administered to seven expert clinicians. Results showed that, on average, the requested minimum viable changes implied a significant reduction of fasting blood sugar, systolic blood pressure, and triglycerides and a significant increase of high-density lipoprotein in patients at risk of diabetes. A significant reduction in body mass index was also recommended in most of the patients at risk, except in females without hypertension. In general, greater changes were recommended in hypertensive patients compared to non-hypertensive ones. The experts were overall satisfied with the proposed approach although in some cases the proposed recommendations were deemed insufficient to reduce the risk in a clinically meaningful way. Future research will focus on a larger set of biomarkers and different comorbidities, also incorporating clinical guidelines whenever possible. Development of additional mathematical and clinical validation approaches will also be of paramount importance
Effects of cardioversion of atrial fibrillation on endothelial function in hypertension or diabetes
No full textCardioversion (CV) to sinus rhythm corrects endothelial dysfunction secondary to atrial fibrillation (AF). As AF often complicates hypertension and diabetes (disorders associated with impaired endothelial function) the study probed whether these comorbidities to AF produced an additive effect and to what extent CV might be advantageous. MATERIALS AND METHODS: Brachial artery flow-mediated dilatation (FMD) was evaluated before and after CV in 17 lone AF patients (group 1), 16 patients with AF + hypertension (group 2) and 17 patients with AF + diabetes type II (group 3), while in supine and head-up tilting (HUT) positions, as this is when endothelial vasodilation is emphasized as a counterbalance to neurogenic vasoconstriction. RESULTS: After 2 weeks, CV in group 1 increased (P 9.50%) and restored its HUT potentiation (from 9.31-->17.22%). In group 2, FMD also improved significantly with CV (supine from 4.92-->7.11% and HUT from 5.29-->11.83%; P 4.92% and HUT from 4.98-->4.73%). After 3 months, FMD improvement persisted in groups 1 and 2 with enduring sinus rhythm, but not in those with AF relapse. In group 3, FMD remained unchanged regardless of cardiac rhythm. CONCLUSIONS: Cardioversion persistently increases supine shear stress endothelial responsiveness and restores the orthostatic modulation in AF alone or in association with hypertension, but not with diabetes. Differences in background endothelial impairment may explain the presence (hypertension) or the absence (diabetes) of an additive AF effect in comorbidities, as well as CV result
Endothelial dysfunction and exercise performance in lone atrial fibrillation or associated with hypertension or diabetes : different results with cardioversion
No full textEndothelial dysfunction and underperfusion of exercising muscle contribute to exercise intolerance, hyperventilation, and breathlessness in atrial fibrillation (AF). Cardioversion (CV) improves endothelial function and exercise performance. We examined whether CV is equally beneficial in diabetes and hypertension, diseases that cause endothelial dysfunction and are often associated with AF. Cardiopulmonary exercise and pulmonary and endothelial (brachial artery flow-mediated dilation) function were tested before and after CV in patients with AF alone (n= 18, group 1) or AF with hypertension (n= 19, group 2) or diabetes (n= 19, group 3). Compared with group 1, peak exercise workload, O2 consumption (Vo2), O2 pulse, aerobic efficiency (Delta Vo2/Delta WR), and ratio of brachial diameter changes to flow changes (Delta D/Delta F) were reduced in group 2 and, to a greater extent, in group 3; exercise ventilation efficiency (Ve/Vco2 slope) and dead space-to-tidal volume ratio (Vd/Vt) were similar among groups. CV had less effect on peak workload (+7% vs. +18%), peak Vo2 (+12% vs. +17%), O2 pulse (+33% vs. +50%), Delta Vo2/Delta WR (+7% vs. +12%), Ve/Vco2 slope (-6% vs. -12%), Delta D/Delta F (+7% vs. +10%), and breathlessness (Borg scale) in group 2 than in group 1 and was ineffective in group 3. The antioxidant vitamin C, tested in eight additional patients in each cohort, improved flow-mediated dilation in groups 1 and 2 before, but not after, CV and was ineffective in group 3, suggesting that the oxidative injury is least in lone AF, greater in hypertension with AF, and greater still in diabetes with AF. Comorbidities that impair endothelial activity worsen endothelial dysfunction and exercise intolerance in AF. The advantages of CV appear to be inversely related to the extent of the underlying oxidative injury
Characterization of Type 2 Diabetes Using Counterfactuals and Explainable AI
No full textType 2 diabetes mellitus is a metabolic disorder of glucose management, whose prevalence is increasing inexorably worldwide. Adherence to therapies, along with a healthy lifestyle can help prevent the onset of disease. This preliminary study proposes the use of explainable artificial intelligence techniques with the aim of (i) characterizing diabetic patients through a set of easily interpretable rules and (ii) providing individualized recommendations for the prevention of the onset of the disease through the generation of counterfactual explanations, based on minimal variations of biomarkers routinely collected in primary care. The results of this preliminary study parallel findings from the literature as differences in biomarkers between patients with and without diabetes are observed for fasting blood sugar, body mass index, and high-density lipoprotein levels
- …
