How Kenyan innovators are using AI to come up with health solutions

For a minute, forget all the hype and the endless debate on  artificial intelligence (AI).

Just what is the future of AI in healthcare, not only in Kenya but also on the continent?

Speaking at the recently concluded 2023 Grand Challenges Annual Meeting that was held in Dakar, Senegal, Juliana Rotich, a leading global expert in AI and the head of M-Pesa at Safaricom,  explained that AI refers to the simulation of human intelligence processes by computer systems.

This basically means that it involves the development of algorithms, software and hardware that enable computers to perform tasks typically requiring human intelligence. 

AI can include various capabilities such as problem-solving, learning from data, understanding natural language, recognising patterns and making decisions. It is a broad field that encompasses machine learning, deep learning, natural language processing and computer vision, among other subfields, with the goal of enabling machines to perform tasks that normally require human intelligence.

 In one of the panel discussions on the future of AI in Africa, Ms Rotich highlighted that she was particularly inspired by an innovation by Theofrida Maginga, a Tanzanian lecturer at Sokoine University of Agriculture, who has added an Internet of Things (IoT) component to and developed a Swahili powered Chat Generative Pre-Trained Transformer (ChatGPT) chatbot that can help local communities and rural farmers in early detection of diseases on maize farms.

IoT refers to the collective network of connected devices and the technology that facilitates communication between devices and the cloud, as well as between the devices themselves.

“The data that comes from IoT in farms to do early detection of diseases and then putting that data into a large language model that can also help to inform farmers about a problem is really interesting because it’s a mix of technology and has the potential to help increase crop yields. I had a great conversation with Theofrida and we are thinking of how to infuse that data around crop yields and diseases to lower insurance premiums,” Ms Rotich said.

Theofrida is one of the 50 grantees mostly from Africa who the Bill & Melinda Gates Foundation gave $100,000 each earlier this year to help them innovate and address issues around food and healthcare on the continent.

In addition, during the Grand Challenges Annual Meeting,  Bill Gates announced a commitment of $30 million to develop an AI platform in Africa.

The AI platform, he said, will equip African researchers and scientists with technological and operational support, helping them turn innovative ideas into scalable health solutions.

“In the coming years, rapid advances in AI will play an increasingly large role in global health and development, with the potential to fundamentally alter the way people communicate, work, learn and improve their overall health and well-being.

“The world needs to make sure that everyone—and not just people who are well-off—benefits from AI. This is why governments and philanthropists will need to play a major role in ensuring that AI reduces inequity and doesn’t contribute to it,” Mr Gates said.

This is even as more Kenyan and African innovators embrace AI to solve health challenges.

The innovators who Healthy Nation spoke to are of the view that “the world must remember that AI is only a tool and it can never replace human beings”. They say the future of AI in healthcare is promising as it has the potential to revolutionise medical diagnosis, drug discovery, personalised treatment plans and patient care. It can improve efficiency, reduce errors and enhance patient outcomes. 

Despite the benefits, they point out that AI also comes with challenges related to data privacy, ethical considerations and regulatory frameworks.

As AI technologies continue to advance, we can expect increased integration into healthcare systems, making healthcare more precise and accessible.
Here are healthcare innovations from Kenya and the continent that experts believe will redefine the future of AI in healthcare.

SOPHIEBOT

For young people who have sexual reproductive health questions and are afraid of talking about sex with adults, SophieBot is here for you.

According to Irvine Mukasa, the developer, SophieBot is an AI chatbot that is fed with verified information on sexuality and sexual reproductive health and relays the information to its users through conversations that are driven by text or voice chats.

Her features include anonymous forums and digital chatbots built in on the app, Telegram, X (@misssophiebot) and web-site (sophiebot.ml).

“We like people to think of us as Siri (a virtual assistant developed by Apple that uses voice recognition and natural language processing to answer users’ questions ) for sexual reproductive health. We have users asking us questions for free because we found out that our value lies in automating customer support for businesses.

We were the finalist for the ‘i-am initiative’,  a UNFPA programme run by Nailab aimed at getting innovators solving sexual reproductive health issues.Winning the competition guaranteed us $10,000 in seed funding plus space and acceleration from Nailab ,” Irvine told Healthy Nation. Nailab is a business incubator based in Nairobi that provides business advice, technical training as well as professional mentoring and gives access to markets apart from linking entrepreneurs to investors.

ASKNIAH.COM

Thanks to AI, a Kenyan developer has made it possible for you to simply send a WhatsApp message to Askniah.com’s assistant  and you instantly receive health information and support free of charge.

“I have built a healthcare assistant powered by open AI’s powerful language model.

All you have to do is to visit Askniah.com and type in your symptoms —  like if you have a headache just type ; ‘I have a headache’ — and you will receive professional and authentic feedback from Niah on what to do, it’s that simple,” Kevin Oduor and Annick Gulenywa, the developers,  explained to Healthy Nation.

“Skip the hassle of booking doctor appointments for minor concerns. with our whatsapp-based symptom checker, you can easily access help 24/7, right at your fingertips,”Askniah.com highlights on the website while assuring that the goal is to empower you with knowledge about health conditions, symptoms and preventive measures. 

 DIAGNOSOFT

 According to Diagnososft CEO Kabuu Kihara, the  healthtech start-up uses AI Machine Learning models to diagnose diseases from medical imaging scans with high accuracy.

“The AI-driven diagnostic solution originally set its sights on early tuberculosis  detection using chest X-rays. This groundbreaking technology is a testament to the power of advanced machine learning algorithms, which swiftly and accurately analyse medical images. In regions where resources are limited and healthcare access is a constant struggle, our solution bridges these gaps, providing critical support where it is needed most,” Kabuu told Healthy Nation.

Diagnosoft is able to detect abnormalities in less than five seconds compared to a radiologists’ 30 minutes;  with high accuracy levels, Kabuu explains , adding that  the innovation reduces misdiagnosis rates ; meaning less lives lost and less medical claims paid out by hospitals. “Diagnosoft uses deep learning models to predict and diagnose scanned images for brain tumours, brain and chest trauma and tuberculosis.”

Kabuu explains they are making a difference in four main ways:

Detecting diseases earlier: Our AI models are designed to provide radiologists with tools that help detect diseases earlier. By providing accurate and timely diagnoses, we aim to improve the chances of successful treatment and save lives.

Improved patient outcomes: Our innovation enables faster and more precise diagnoses. 

Better life balance for radiologists: By providing  radiologists with AI tools, we aim to relieve the burden of manual analysis, allowing them to focus on more critical and nuanced aspects of patient care. This, in turn, improves their work-life balance and job satisfaction.

Reduced healthcare costs: Through early detection and precise diagnosis, AI helps alleviate the need for extensive surgeries, prolonged hospital stay, and costly treatments. By identifying health issues at an earlier stage, AI not only enhances patient outcomes but also eases the strain on healthcare facilities, allowing them to allocate resources more efficiently and ultimately leading to a healthier and more cost-effective healthcare ecosystem.  The developer highlights that they are looking at AI as a tool rather than a replacement of human healthcare workers.

“By harnessing the power of technology, we aim to bridge healthcare gaps, improve medical outcomes, and create a healthier Africa for all. We were finalists at the United Nation AI for good innovation competitions,” the CEO says.

DIABETES MANAGEMENT MEDICAL CENTER

For Duncan Motanya, a personal tragedy that saw him lose a parent to diabetes propelled him to leverage AI and set up a diabetes management clinic with an online pharmacy in an effort to ensure affordable access to diabetes treatment. The clinic is located in Ngara, Nairobi.

“ Diabetes Management Medical Center ( DMMC ) is a small giant in the healthcare space whose services include an Online Pharmacy (www.dmrckenya.co.ke) and a TeleHealth platform (www.teledoc.co.ke), which allows people with non-communicable diseases to consult doctors, nutritionists and laboratory service providers among others from the comfort of their homes.

«At the moment we only use AI to narrow down to the ailment a platform user may be suffering from based on the symptoms that they key in when signing up,” he told Healthy Nation while disclosing that all this began on Facebook.

“Before 2020, if you asked me I did not think a Facebook page I created in 2014 to educate people about diabetes would make a concrete business case, leading to my decision to go all out in 2020 by setting up a pharmacy, which is today a medical center focusing on offering quality and affordable primary healthcare services to people with diabetes and hypertension.

« I have invested about Sh15 million in setting up the medical center, a laboratory, among others. But we only became profitable this year” Duncan, who is also the chairperson of the Fintech Association of Kenya, disclosed.

“YOUR CHOICE” HIV APP

Caroline Govathson, a health economics and epidemiology research officer at HE2RO in South Africa,  noted that“Your choice” app prototype, an AI-powered chatbot that aims to nip in the bud the stigmatisation of HIV patients, is what the continent needs.

 
“We have leveraged on LLM (a deep learning algorithm that can perform a variety of natural language processing tasks) to reduce stigma and improve precision in HIV risk assessment so as to provide a potential, tailored HIV prevention plan,» she explained while observing that 189,000 people (64 per cent women) were newly infected with HIV in South Africa in 2022.

“34 per cent of women newly infected are aged below 24 years. High risk populations are being left behind,” she told Healthy Nation.

She noted that thechatbot is simple and easy to use. “This is an intermediary between clients and healthcare providers , with 100 per cent privacy guaranteed. Imagine talking to your phone and no one else about your HIV treatment plan, medication and clinic visits without being judged or stigmatised.

«The LLM persona is non-judgmental, knowledgeable healthcare provider with a goal of assessing HIV risk; pre-exposure prophylaxis (PrEP) interest and PrEP eligibility,” she explained while disclosing that they have used ChatGPT, Llama and Claude AI technologies to build the chatbot. 

“For the design, we looked at prompt engineering, rules for completing assessment and summarising relevant risk information for providers.»

How global medical field is benefitting from AI

Globally, here are specific ways that AI is currently being used to improve healthcare , and how the tool is likely to be used in the future.

 Disease diagnosis: AI can analyse large medical images and patient datasets to identify patterns and correlations that human doctors might miss. This can lead to more accurate and earlier diagnoses, saving lives and improving patient outcomes.

Drug discovery: AI can be used to screen millions of potential drug compounds to identify those that are most likely to be effective against a particular disease. This can significantly accelerate the drug discovery process and bring new treatments to patients faster. 

 Patient monitoring: AI can be used to monitor patients' vital signs and other health data in real-time to identify potential problems early on. This can help to prevent complications and improve patient safety.

Personalised medicine: AI can be used to tailor medical treatments to individual patients based on their unique genetic profile and other factors. This can lead to more effective treatments and fewer side effects. In the future, AI is likely to play an even greater role in healthcare. For example, AI-powered robots could perform surgery with greater precision and accuracy than human surgeons. AI could also be used to develop new diagnostic tools and treatments for diseases that are currently incurable.

Below are examples of some of the leading AI companies dedicated to revolutionising healthcare.

Babylon Health: Babylon Health is an AI-powered healthcare app that can diagnose and treat over 100 common conditions. The app uses a chatbot to ask patients questions about their symptoms and then uses AI to generate a diagnosis and treatment plan..

DeepMind Health: DeepMind Health is a Google AI company that is developing AI-powered tools to improve healthcare. One of DeepMind's most promising projects is an AI system that can detect diabetic retinopathy from eye images. Diabetic retinopathy is a leading cause of blindness, and early detection is essential for preventing vision loss.

IBM Watson Health: IBM Watson Health is a division of IBM, a global technology innovator leading advances in AI, automation and hybrid cloud solutions that is developing AI-powered tools to improve healthcare. One of Watson Health's most successful products is Watson for Oncology, which is used by oncologists to help personalise cancer treatment plans.

[email protected]