![Asla – Watan Sahi [Official MV] Latest Punjabi Song – K Million Music](https://i.ytimg.com/vi/sCuLojys0n4/maxresdefault.jpg "Asla – Watan Sahi [Official MV] Latest Punjabi Song – K Million Music")
Shahid Akhter, editor, ETHealthworld, spoke to Nakul Jain, Director, Solutions, Wadhwani Institute of Artificial Intelligence, to figure out the power of AI and its utility for tasks such as triage, generating differential diagnoses, and even offering personalized treatment recommendations.
AI in Public Health care
From a public health perspective, harnessing the power of advanced computing and the possibility to super-compute the vast amount of data collected through digital systems, has opened up numerous opportunities to work on a lot of things. As of today, artificial intelligence is playing a pivotal role in various aspects of public health. It begins with applications in diagnostics and screening, which can encompass image-based or sound-based methods. Additionally, there is a significant focus on clinical decision support, utilizing AI for tasks such as triage, generating differential diagnoses, and even offering personalized treatment recommendations. Beyond that significant efforts have been made in terms of surveillance and disease modeling, heavily relying on artificial intelligence.
Recently, the emergence of large language models like ChatGPT has paved the way for the development of health assistants. These health assistants are being considered for various levels of healthcare workers and even for everyday citizens who can use the power of artificial intelligence in their day-to-day life.
AI in public Health care: Challenges
There are numerous challenges which emerge when public health and artificial intelligence come together. This is primarily due to the widespread deployment of digital systems that collect vast amounts of data within the ecosystem. One significant challenge revolves around data quality and data integrity. While data on various health aspects and programmatic data concerning different patients and individuals is being collected, the manner in which it’s gathered in many settings might not be optimal for training artificial intelligence. Therefore, there is a need to develop and deploy specific solutions to ensure that the collected data can effectively train artificial intelligence models.Another challenge that public health models and AI models in general often face is the issue of bias in the data. Given the complex and diverse nature of the data we collect, it is essential to ensure that there is no bias present. Otherwise, the artificial intelligence system may produce results that are biased toward a particular outcome and that might not be the actual output that would be optimal for use in a public health setting.
The other issue that is very pertinent specially in this setting is ethical and legal consideration, approvals, and accreditations. This is because we handle a substantial amount of sensitive and proprietary data, often collected from diverse ecosystems. So, it is very important for us to understand what kind of product we are building and how to make sure that we are building it in the most ethical manner, adhering to all regulatory requirements, and having taken utmost care around data security and privacy. Addressing these factors poses a significant and genuine challenge that organizations usually face.
Another significant challenge we encounter relates to the ability to integrate our solutions effectively. Again this is because while we develop these solutions a lot of user centricity is at the core of it, but given the kind of population and geography we live in, they bring their unique challenges and pain points. Trying to create a solution that fits for all is inherently challenging. Moreover, deploying such a solution on a large scale, in a production environment, presents its own set of difficulties. The multitude of systems, processes, programs, and ecosystems in the field makes it exceedingly complex to create and implement a uniform solution. So, all of these are challenges that need to be solved before considering the deployment of impactful solutions on a large scale.
AI in public Health care: Rural Outreach
Another aspect of the challenges is ensuring that the solutions we develop effectively reach rural populations. While the challenges we’ve mentioned apply broadly to the rural setting as well, there are certain initial hurdles which include securing acceptance, promoting adoption, and providing on-field training. We need to solve these issues to make sure that the solutions reach the last mile and genuinely start benefiting the people who are actually in need of these solutions.
Wadhwani AI
Wadhwani AI is an independent non for profit institute which has been working since 2018 in the area of artificial intelligence for social good. We aim at serving the underserved population across the developing nations. We have a clear mandate to collaborate with the public and governments, comprehensively understand the challenges and pain points people face in various domains including healthcare, agriculture, education, urban affairs, and climate. Our goal is to identify areas where artificial intelligence can make a substantial impact in addressing these issues and develop sustainable, scalable solutions that benefit all the people in the community.
Wadhwani AI collaborations
Our journey as an institute even if it is not very long, it is very impactful. We have worked with various government departments and integrated our team within the ecosystem. We are today collaborating with the ministry of health and family welfare, ministry of agriculture and farmer welfare, and with ministry of education.Additionally, we are partnering with multiple state governments in the fields of healthcare, education, and agriculture. Our involvement extends to organizations like the National Skill Development Corporation, among many other ministries.
We are supported by some of the leading donors in the ecosystem. As of today, we are working on 30+ solutions across various domains. Notably, one of our solutions, focused on clinical decision support, has already been seamlessly integrated into the national telemedicine application known as e-sanjeevani. Through this platform, we have successfully conducted over 30 million triage assessments. There are five other solutions which are already deployed and the remaining solutions are also in advanced stages of development and implementation.
As an institute our mandate is to positively impact as many lives as possible. We want to be an applied AI institute. Our philosophy has taken a slight change in the last couple of years, where we have placed application of all our solutions at the core. Our ideology is to build solutions that are field-ready, deployable, and capable of making a genuine impact. This is the path we intend to continue on in the future. We will continue to work and collaborate with different governments, developing nations, communities, and individuals to make sure that we enrich their lives using the cutting edge solutions that we build.
