Non-invasive Anemia Detection Tool

NIADAT is a non-invasive, point-of-care, real-time solution that uses Artificial Intelligence to detect and monitor anemia using a smartphone app.

  • 0 Raised
  • 560 Views
  • 0 Judges

Tags

  • No tag

Categories

  • This hackathon has categories available. Please select one if necessary.

Description

Globally, 2+ billion people suffer from anemia.  Globally anemia prevalence has been particularly persistent among reproductive age women and young children despite intervention. WHO estimates 571M women and 269M young children are anemic. 

There are no specific symptoms for anemia. The risks associated with anemia increase as hemoglobin levels decrease. If undetected for long enough, anemia often becomes fatal. 

  1. Anemia in women & children:
  • CDC WIC program supports pregnant & postpartum women and more than 50% of all infants born in the USA with nutritional supplements to help prevent iron deficiency anemia. And yet, WIC reports 13% increase in anemia in the last decade impacting minority communities disproportionately . 
  • 800 million children around the world are anemic. In Africa , 50% of all children will experience anemia one time or another. These kids and our future are at risk of impaired cognitive development and cognitive loss, obesity and increased years lived with disabilities.

Solution.

We propose a radical solution to this multi-dimensional long-standing anemia prevalence problem by developing a novel artificial intelligence system - Non-invasive Anemia Detection with AI  

The tool is a smartphone app that analyzes eyelid images to estimate and display hemoglobin level on the screen in real-time. Hemoglobin level is used to decide whether a user has anemia or not. The estimates can be saved for regular and long term monitoring for an individual or at the population level.

It is an SaMD (Software as Medical Device) . It is built upon a framework of integrated solution for anemia screening, monitoring  and intervention.

The solution is developed with deep learning as the core technology , which is securely deployed on AWS cloud with full redundancy and data privacy in place.  App uses the cloud AI & Data platform through an API and hence it is integration ready for other telemedicine and remote monitoring app and websites.


2. Anemia & maternal health:

 Today, the world sees 1 maternal death every 2 mins. 

  • Maternal anemia in pregnancy is a reversible risk factor. And yet, maternal anemia is estimated to contribute to 20% of maternal deaths (Black, et al., 2008).
  • Growing evidence that anemia is linked to greater risk of postpartum hemorrhage(PPH) (Kavle, et al., 2008). 
    • PPH is responsible for 25 percent of maternal deaths globally 
  • Anemia in pregnancy can lead to premature birth, poor brain development in infants.
  • Antepartum anemia is most prevalent among historically marginalized pregnant patients, contributes to disparities in severe maternal morbidity(SMM) and postpartum depression.
  • While anemia prevalence more than doubled from 2011 to 2020 for all racial groups in the USA, Black and American Indian–Alaska Native mothers are affected most. A significant and persistent disparity(12 percent point difference) in anemia exists between Black(21.5% prevalence) and White(9.6%) pregnant patients over the last decade.

Attachments