Digital Screening Tool to detect Covid-19 infected People

Nalavade, J. E.; Ieee,

Covid-19 pandemic is increasing day by day in world. Government agencies informing people to stay safe don't go outside, keep social distancing. Means we want to break chain of corona virus affected people. Till date there is no any automation tool is designed to detect Covid-19 patients. Such type of automation tool is required to detect patient at earlier stage(during first week of infection). If we detect Covid-19 patients in earlier stage and take required actions, we can save all patients. In this proposal we want to design automated digital screening tool to identify the people in the first week of the pandemic. Based on machine learning techniques, system will be trained as per the day to day symptoms. ASHA workers from Rural Health Centre can collect the all the peoples information related to Covid-19. By using Classification algorithm system will classify the people into different categories like healthy people (no any symptoms), first day of covid-19, second day covid-19, upto fourteenth day of covid-19. Data set required for training the model will be created by studying covid-19 patients per day history. Like first day he had aches and pains, nasal congestion, runny nose, sore throat or diarrhea etc., related any symptoms. Similarly, we will prepare dataset of 5000 patients. Using this dataset model will be trained. After training the model whenever such type of pandemic occurs we will detect the infected people in early stage. We will reach to each of the citizen of India through social workers, ASHA worker, Rural Health care staff. Collect information of each person by using attributes which is decided for dataset. We will provide collected dataset (each citizen information) to our designed system. Designed system will classify according to classes mentioned above. If person don't have any symptoms related to covid-19 means he is healthy. If he had any symptoms according to day we have classified then such type of people immediately informed and according to requirement they will be kept in isolation, quarantine or hospitalized immediately. If we able to identify during the first week of pandemic then we can break chain at early stage and we can avoid spreading of coronavirus to large number of people. If we use such type of system then no need of lockdown for long period is required. Time required only for collecting the information, identifying the infected people and take appropriate action as mentioned above. Maximum 1 week of lockdown will be sufficient and we can stop chain. No any major loss in economy. In future this type of model can be used for any type of pandemic. We will detect infected people at early stage and we can stop chain for any type of pandemic. Here we are designing this type of tool using machine learning, data mining and IoT concepts. Till date nobody has been designed such type of tool to detect covid-19 within less time. Till date nobody has been designed such type of tool to detect covid-19 within less time.