Application of artificial intelligence and machine learning in the medical field is something that has picked up a lot of steam in the last few years. From disease prediction to brain mapping, deploying models to decipher complex medical data has proven to be more practical and efficient, thus aiding the medical practitioners to deliver their tasks better.
For her work that uses AI to predict the outbreak of air-borne diseases based on climatic conditions, Mumbai-based professor, Dr Nupur Giri has been awarded Microsoft’s Artificial Intelligence (AI) for Earth Grant 2019.
Dr Giri, who is the head of the department of computer engineering at the Vivekanand Education Society’s Institute of Technology (VESIT) in Mumbai, collaborated with her students to devise an AI system that can detect the outbreak of tuberculosis (TB) due to climatic conditions.
The AI systems used multiple sources of data of varied frequency and the team spent a few months to smoothen, standardise, and normalise the data.
“There are around 600 students in the department of computer science and we do a lot of projects that are socially relevant. Detection of airborne diseases like tuberculosis is one among the AI project with Microsoft, that is currently underway,” Dr Giri says.
AI for detecting the propagation of TB
For the project, Dr Giri and her team began their work in 2018 and tried to understand the correlation between a different feature that leads to a disease outbreak in India. Since an estimated 2.74 million people in India were detected with TB in 2017 and due to the increased risk associated with the disease, the team first chose TB to work upon.
As a first step towards understanding the propagation of the disease better, data on external factors like climatic factors and pollution rate were collated from 725 districts over a 17-year time frame from various sources before running the AI algorithms.
Accurate and clean data being the key element that could determine the success of the project, the team relied on climate datasets from Skymet Weather, pollution and air quality datasets from Open Government Data Platform India, tuberculosis dataset from the Central Tuberculosis Division and population data from the national census, a report by Microsoft notes.
As a result of their continuous effort and with the help of important tools, the team was able to deploy the AI models on the data. Talking about the success of the project, Dr Giri said, “The initial results are good, but we are currently testing multiple machine learning and neural network models to improve the accuracy of their predictions for every district in India,” she adds.
Sophisticated tools for easy modelling
Microsoft AI for Earth Grant supports projects that change the way people and organisations monitor, model, and ultimately manage in the areas of agriculture, biodiversity, climate change and water. As a receptionist of the grant, the tech giant will provide Dr Giri and her team access to Microsoft Azure cloud computing resources to develop solutions.
With access to sophisticated tools for processes like data cleaning and crunching, Dr Giri notes that the tools proved to be greatly beneficial, “The data science virtual machine was a huge benefit, as it comes preloaded with all the tools required, and data crunching became easier. The Machine Learning Studio allowed us to minimise the time required to develop algorithms and write codes, as it has a drag-and-drop authoring environment,” she said.
The road ahead
Though there are enough medication and cure for TB in India, it is among the list of countries with a high concentration of Multi-drug Resistant TB (MDR-TB) cases. MDR- TB is a state when the virus becomes stronger due to irregular intake of medication by a patient.
Hence, tackling TB medicine adherence continues to be a challenge that needs to be addressed. Therefore providing a data visualisation dashboard to help those working on eradicating TB make more informed decisions and extending the application of the technology to other diseases are among Dr Giri’s future plans with regards to the AI application.
At a more granular level, the team also hopes to understand the environmental changes associated with the spread of the disease, “If we can employ AI algorithms to predict the spread of disease and understand how environmental changes, including climatic conditions and external factors like pollution impact the ecology and epidemiology of the disease, we can initiate precision public health at a granular level,” Dr Giri concludes.