ABSTRACT
If understanding sentiments is already a difficult task in human-human communication, this becomes extremely challenging when a human-computer interaction happens, as for instance in chatbot conversations. In this work, a machine learning neural network-based Speech Emotion Recognition system is presented to perform emotion detection in a chatbot virtual assistant whose task was to perform contact tracing during the COVID-19 pandemic. The system was tested on a novel dataset of audio samples, provided by the company Blu Pantheon, which developed virtual agents capable of autonomously performing contacts tracing for individuals positive to COVID-19. The dataset provided was unlabelled for the emotions associated to the conversations. Therefore, the work was structured using a sort of transfer learning strategy. First, the model is trained using the labelled and publicly available Italian-language dataset EMOVO Corpus. The accuracy achieved in testing phase reached 92%. To the best of their knowledge, thiswork represents the first example in the context of chatbot speech emotion recognition for contact tracing, shedding lights towards the importance of the use of such techniques in virtual assistants and chatbot conversational contexts for psychological human status assessment. The code of this work was publicly released at: https://github.com/fp1acm8/SER. © 2023 The Authors. Cognitive Computation and Systems published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and Technology and Shenzhen University.
ABSTRACT
Background: Patients with cancer are purported to be more vulnerable to coronavirus disease 2019 (COVID- 19). However, cancer encompasses a spectrum of heterogeneous tumor subtypes. The aim of this study was to investigate severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection risk and COVID-19 prevalence according to tumor subtype in the resident cancer patient population of the Province of Parma (Emilia Romagna Region, Northern Italy) during the first wave of COVID- 19 pandemic in Italy. Methods: We analyzed data from the Parma Province Cancer Registry, COVID-19 hospital medical records, and local surveillance system of all laboratory-confirmed cases tested positive for SARS-CoV-2 from the beginning of the outbreak (20th of February) to the 19th of July 2020. All the Parma resident population of cancer patients was classified as either “active” or “inactive” according to the evidence of any referral to health services, for any reason, during the observation period. Study analyses were adjusted for patient demographics, tumor subtype and period of cancer diagnosis. Results: 40,148 cancer patients (mean age 68 years;57.8% females;45.1% active) were analyzed. The cumulative risk of SARS-CoV-2 infection was 11.2% for cancer patients vs. 7% for non-cancer subjects (P < 0.0001). The overall COVID-19 attack rate was 2.2% (95% CI, 2.0-2.4) and 2.6% (95% CI, 2.4-2.9) for inactive and active cancer patients, respectively. The cumulative incidence of COVID-19 was higher in active vs. inactive cancer subjects (HR 1.18, P = 0.01). In the active cancer group, the cumulative incidence of COVID-19 was higher in lung cancer patients vs. other tumor subtypes (HR 4.3). In the same group, HR for breast cancer patients was 0.86. Interestingly, the subgroup analysis of COVID-19 cumulative incidence showed a significant interaction between active patient status and hematological malignancies. Conclusions: In our study, patients with cancer were more susceptible to SARS-CoV-2 infection. The cumulative incidence of COVID-19 was higher in active vs. inactive cancer subjects. However, cancer is a heterogeneous group of diseases and patients with different tumor types had differing susceptibility to COVID-19 phenotypes. COVID-19 fatality rates for subgroups will be reported at the meeting.
ABSTRACT
Background: Patients (pts) with cancer are purported to be more vulnerable to coronavirus disease 2019 (COVID-19). However, cancer encompasses a spectrum of heterogenous tumor subtypes. The aim of this study was to investigate severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection risk and COVID-19 prevalence according to tumor subtype in the resident cancer population of the Province of Parma (Emilia Romagna Region, Nothern Italy) during the first wave of COVID-19 pandemic in Italy. Methods: We analyzed data from the Parma Province Cancer Registry, COVID-19 hospital medical records, and local surveillance system of all laboratory-confirmed cases tested positive for SARS-CoV-2 from the beginning of the outbreak (February, 20) to July 19, 2020. All the Parma resident cancer population was classified as either “active” or “inactive” according to the evidence of any referral to health services, for any reason, during the observation period. Study analyses were adjusted for patient demographics, tumor subtype and period of cancer diagnosis. Results: 40,148 cancer pts (mean age 68;57.8% females;45.1% active) were analyzed. The cumulative risk of SARS-CoV-2 infection was 11.2% for cancer pts vs. 7% for non-cancer subjects (P < 0.0001). The overall COVID-19 attack rate was 2.2% (95% CI, 2.0-2.4) and 2.6% (95% CI, 2.4-2.9) for inactive and active cancer pts, respectively. The cumulative incidence of COVID-19 was higher in active vs. inactive cancer subjects (HR 1.18, P = 0.01). In the active cancer group, the cumulative incidence of COVID-19 was higher in lung cancer pts vs. other tumors (HR 4.3). In the same group, HR for breast cancer pts was 0.86. Interestingly, the subgroup analysis of COVID-19 cumulative incidence showed a significant interaction between active patient status and hematological malignancies. Conclusions: In our study, cancer pts were more susceptible to SARS-CoV-2 infection. The cumulative incidence of COVID-19 was higher in active vs. inactive cancer subjects. However, cancer is a heterogeneous group of diseases and pts with different tumor types had differing susceptibility to COVID-19 phenotypes. COVID-19 fatality rates for subgroups will be reported at the meeting. Legal entity responsible for the study: University Hospital of Parma. Funding: Has not received any funding. Disclosure: A. Musolino: Financial Interests, Institutional, Research Grant: Lilly;Financial Interests, Personal, Advisory Board: Roche;Financial Interests, Personal, Advisory Board: Macrogenics;Financial Interests, Personal, Advisory Board: Pfizer;Financial Interests, Personal, Advisory Board: Seagen. B. Pellegrino: Financial Interests, Personal, Research Grant: Roche. All other authors have declared no conflicts of interest.