Predicting severity of dengue using simple AI model

Dengue fever (DF) and dengue hemorrhagic fever (DHF) are mosquito-borne viral diseases that cause a significant public health burden, especially in Southeast Asia. In India, the prevalence of dengue has increased dramatically in recent decades, with a rapidly changing epidemiology. The first major DHF outbreak in India occurred in 1996, followed by a DF outbreak in 2003. Since then, there has been a marked increase in the number and frequency of dengue outbreaks, and the disease is now endemic in most states. All four serotypes of the dengue virus are now circulating in India, with the predominant serotype changing frequently. Despite this increasing trend, surveillance, reporting, and diagnosis of dengue in India remain largely passive. We need more technology driven solutions to help reduce fatality, early detection and reduce economic burden. We will review here an AI based approach to identify the severity of the disease.

Dengue is a viral infection transmitted by the bite of an infected Aedes mosquito. It is a serious public health problem in many tropical and subtropical countries, including India. According to World Health Organization, Dengue has emerged as the most widespread and rapidly increasing vector-borne disease in the world. Of the 3.5 billion people around the world living in dengue endemic countries and at risk of contracting dengue fever, more than 1.5 billion live in dengue endemic areas in 10 countries of the Southeast Asia Region.
In India, dengue was rare before the 1970s. However, the number of cases has increased steadily over the past few decades. According to the National Vector Borne Diseases Control Programme (NVBDCP), India reported over 2,33,251 dengue cases in 2022, with 303 deaths.

There are a number of factors that have contributed to the increase in dengue cases in India. These include:

  1. Climate change: Climate change is leading to warmer temperatures and more rainfall, which create ideal conditions for mosquito breeding.
  2. Urbanization: Rapid urbanization is leading to the growth of slums and other areas with poor sanitation and drainage, which provide breeding grounds for mosquitoes.
  3. International travel: Increased international travel has made it easier for dengue viruses to spread to new areas.

The economic loss due to dengue in India is significant. This includes the cost of direct medical care, indirect costs such as lost productivity, and the cost of premature deaths. According to the World Health Organization, the incidence of dengue globally has shot up 30-fold in the past 50 years. The cumulative dengue diseases burden has attained an unprecedented proportion in recent times with a sharp increase in the size of human population at risk. Dengue disease presents highly complex pathophysiological, economic, and ecologic problems. More than one billion people (15% of the world’s population) reside in India. India’s population is twice that of Southeast Asia, the region that currently reports the most dengue-related deaths. A recent study done at the University of Oxford using a map-based approach to model how many dengue cases were occurring in various parts of the world, estimated that India had the largest number of dengue cases, with about 33 million apparent and another 100 million asymptomatic infections occurring annually.

Dengue case definition
The previous classification of dengue divided cases into three categories: dengue fever (DF), dengue hemorrhagic fever (DHF) grades 1 and 2, and dengue shock syndrome (DSS) grades 3 and 4. DF is an acute febrile illness, DHF is characterized by increased vascular permeability, and DSS is a severe complication of DHF that can lead to hypovolemic shock. The new revised clinical classification of dengue divides cases into two categories: dengue with or without warning signs, and severe dengue infection.

Simple model to identify severity of dengue
Once a person is infected with dengue, it manifests many symptoms. Our aim is to capture these symptoms and a few parameters from recent laboratory tests and come up with a severity score. The severity score will help the patient or a paramedic to give directions to the patient to seek appropriate advice.

Our model uses a combination of factors like Absolute platelet value, MPV, body hydration level, comorbidity etc. to compute a severity score. The score is classified in four categories. Category I indicates it can be managed in OPD. Patient needs to consult a doctor. Category II indicates patients need periodic medical assistance like periodic visits to doctor or tele consultation. Category III indicates patient needs closed medical monitoring i.e., hospital admission. Category IV, indicates sever condition and will require critical care monitoring

We believe this simple but powerful tool will be able to help reduce fatality and economic burden for people in India as well as other dengue prone areas across the globe. Early detection of severity can help providers to provide appropriate treatment.

This simple tool can be deployed as a web bot or mobile bot or WhatsApp bot or any other digital channel depending on the region to benefit the maximum number of people.


Leave A Comment