Enhancing physicians' radiology diagnostics of COVID-19's effects on lung health by leveraging artificial intelligence.

Introduction: This study aimed to develop an individualized artificial intelligence model to help radiologists assess the severity of COVID-19's effects on patients' lung health. Methods: Data was collected from medical records of 1103 patients diagnosed with COVID-19 using RT- qPCR between March and June 2020, in Hospital Madrid-Group (HM-Group, Spain). By using Convolutional Neural Networks, we determine the effects of COVID-19 in terms of lung area, opacities, and pulmonary air density. We then combine these variables with age and sex in a regression model to assess the severity of these conditions with respect to fatality risk (death or ICU). Results: Our model can predict high effect with an AUC of 0.736. Finally, we compare the performance of the model with respect to six physicians' diagnosis, and test for improvements on physicians' performance when using the prediction algorithm. Discussion: We find that the algorithm outperforms physicians (39.5% less error), and thus, physicians can significantly benefit from the information provided by the algorithm by reducing error by almost 30%.

Role of Innate and Adaptive Cytokines in the Survival of COVID-19 Patients.

SARS-CoV-2 is a new coronavirus characterized by a high infection and transmission capacity. A significant number of patients develop inadequate immune responses that produce massive releases of cytokines that compromise their survival. Soluble factors are clinically and pathologically relevant in COVID-19 survival but remain only partially characterized. The objective of this work was to simultaneously study 62 circulating soluble factors, including innate and adaptive cytokines and their soluble receptors, chemokines and growth and wound-healing/repair factors, in severe COVID-19 patients who survived compared to those with fatal outcomes. Serum samples were obtained from 286 COVID-19 patients and 40 healthy controls. The 62 circulating soluble factors were quantified using a Luminex Milliplex assay. Results. The patients who survived had decreased levels of the following 30 soluble factors of the 62 studied compared to those with fatal outcomes, therefore, these decreases were observed for cytokines and receptors predominantly produced by the innate immune system-IL-1α, IL-1α, IL-18, IL-15, IL-12p40, IL-6, IL-27, IL-1Ra, IL-1RI, IL-1RII, TNFα, TGFα, IL-10, sRAGE, sTNF-RI and sTNF-RII-for the chemokines IL-8, IP-10, MCP-1, MCP-3, MIG and fractalkine; for the growth factors M-CSF and the soluble receptor sIL2Ra; for the cytokines involved in the adaptive immune system IFNγ, IL-17 and sIL-4R; and for the wound-repair factor FGF2. On the other hand, the patients who survived had elevated levels of the soluble factors TNFß, sCD40L, MDC, RANTES, G-CSF, GM-CSF, EGF, PDGFAA and PDGFABBB compared to those who died. Conclusions. Increases in the circulating levels of the sCD40L cytokine; MDC and RANTES chemokines; the G-CSF and GM-CSF growth factors, EGF, PDGFAA and PDGFABBB; and tissue-repair factors are strongly associated with survival. By contrast, large increases in IL-15, IL-6, IL-18, IL-27 and IL-10; the sIL-1RI, sIL1RII and sTNF-RII receptors; the MCP3, IL-8, MIG and IP-10 chemokines; the M-CSF and sIL-2Ra growth factors; and the wound-healing factor FGF2 favor fatal outcomes of the disease.

Impact of the Innate Inflammatory Response on ICU Admission and Death in Hospitalized Patients with COVID-19.

OBJECTIVE: To describe the capacity of a broad spectrum of cytokines and growth factors to predict ICU admission and/or death in patients with severe COVID-19. DESIGN: An observational, analytical, retrospective cohort study with longitudinal follow-up. SETTING: Hospital Universitario Príncipe de Asturias (HUPA). PARTICIPANTS: 287 patients diagnosed with COVID-19 admitted to our hospital from 24 March to 8 May 2020, followed until 31 August 2020. MAIN OUTCOME MEASURES: Profiles of immune response (IR) mediators were determined using the Luminex Multiplex technique in hospitalized patients within six days of admission by examining serum levels of 62 soluble molecules classified into the three groups: adaptive IR-related cytokines (n = 19), innate inflammatory IR-related cytokines (n = 27), and growth factors (n = 16). RESULTS: A statistically robust link with ICU admission and/or death was detected for increased serum levels of interleukin (IL)-6, IL-15, soluble (s) RAGE, IP10, MCP3, sIL1RII, IL-8, GCSF and MCSF and IL-10. The greatest prognostic value was observed for the marker combination IL-10, IL-6 and GCSF. CONCLUSIONS: When severe COVID-19 progresses to ICU admission and/or death there is a marked increase in serum levels of several cytokines and chemokines, mainly related to the patient's inflammatory IR. Serum levels of IL-10, IL-6 and GCSF were most prognostic of the outcome measure.