ABSTRACT
This paper proposes a real-time hand gesture recognition interface called VUCIS (Virtual User Computer Interaction System) to operate a mouse, keyboard and stylus. OpenCV and MediaPipe Hands are used to achieve this. OpenCV is utilized for its powerful image processing capabilities and computer vision. MediaPipe Hands library is used to generate 21 landmarks on a hand. By using a combination of various landmarks, a distinct gesture is assigned to a task performed by the input devices. The proposed interface allows users to interact with computers easily, reduces the need for keyboard and mouse as external hardware in certain circumstances and eliminates any form of contact with a surface, minimizing the spread of COVID-19.
ABSTRACT
BackgroundSevere Corona virus disease (COVID-19) is associated with high mortality. Although single centre intensive care units (ICU) have reported clinical characteristics and outcomes, no large scale multicentric study from India has been published. The present retrospective, multi-centre study was aimed to describe the predictors and outcomes of COVID-19 patients requiring ICU admission from COVID-19 Registry of Indian council of Medical Research (ICMR), India.MethodsProspectively collected data from multiple participating institutions was entered in the electronic National Clinical Registry of COVID 19. We enrolled patients aged>18 years with COVID-19 pneumonia requiring ICU admission between March 2020 and August 2021. Exclusion criteria were negative RT PCR, death within 24 hours of ICU admission, or patients with incomplete data in the registry Their demographic characteristics, laboratory variables, ICU severity indices, treatment strategies and outcomes were analysed.ResultsA total of 5865 patients, with mean age 56±15 years, with 3840/5865 (65.4%) men, were enrolled in the ICMR registry.. Overall mortality was 2535/5865 (43.5%). Non-survivors were older than survivors (58.2±15.4 years vs 53.6 ±14.7 years; P=0.001). Non-survivors had multiple comorbidities (n=1951, 52.9%) with hypertension (47.2%) and diabetes (45.6%) being the most common, higher creatinine (1.6 ± P=0.001, high D-dimer (1.56 vs 1.37, P=0.001), higher CT severity index (16.8±5.2 vs 13.5 ±5.47 ) compared to survivors. Non survivors had longer hospital and ICU stay (P=0.001). On multivariate regression analysis, high NLR (HR 1.017, 95% CI 1.005- 1.029, P=0.001), high CRP (HR 1.008, 95% CI 1.006- 1.010, P=0.001), high D dimer ((HR 1.089, 95% CI 1.065- 1.113, P=0.001) were associated with mechanical ventilation while younger age, (HR 0.974, CI 0.965-0.983, p=0.001), high D dimer (HR-1.014, CI 1.001-1.027, P=0.035) and use of prophylactic LMWH (HR 0.647, CI 0.527-0.794, p=0.001) were independently associated with mortality. ConclusionIn this large retrospective study of 5865 critically ill COVID 19 patients admitted to ICU, overall mortality was 2535/5865 (43.5%). Age, high D dimer, CT Severity score and use of prophylactic LMWH were independently associated with mortality.