Woxsen University Researchers alongside US Scientist, develop AI-Based Prediction Model to Detect Obesity Risk

Hyderabad, 17 January: Obesity is becoming a significant global health issue, driven by lifestyle challenges such as diets rich in processed foods and low physical activity, which are further exacerbated by technological advancements and rapid urbanization. To tackle this problem, a team of researchers from Woxsen University, Hyderabad, including Dr. Shahid Mohammad Ganie, Dr. Hemachandran Kannan, and student Bobba Bharath Reddy alongside US Scientist Prof. Manjeet Rege, have explored how combining multiple machine learning methods can predict the risk of obesity based on lifestyle data. Their latest research introduces an AI-based prediction model that leverages boosting techniques to assess the risk of obesity more effectively.
Their research paper titled “An Investigation of Ensemble Learning Techniques for Obesity Risk Prediction Using Lifestyle Data,” was published in the Decision Analytics Journal by Elsevier.

Notably, Prof. Manjeet Rege, a globally acclaimed expert in Artificial Intelligence and Data Science and Chair Professor of the Department of Software Engineering and Data Science at the University of St. Thomas, USA, recently visited Woxsen University in Hyderabad. During his visit, he evaluated innovative research at the Manjeet Rege Analytics Lab, which is focused on the transformative use of Explainable AI and Machine Learning in healthcare.

According to the research, identifying the underlying causes of obesity risk in its early stages has become a challenge for medical practitioners. In the healthcare sector, online medical repositories and hospitals are generating vast amounts of data, providing valuable resources for researchers to explore and leverage AI techniques to address real-life health issues.

The researchers explained that they selected three algorithms from each ensemble method, each possessing distinct characteristics and strengths, to demonstrate the effectiveness of the proposed model from multiple perspectives. Additionally, preprocessing techniques were employed to enhance the quality of the data.
“Our goal at the AI Research Centre at Woxsen, has always been to use technology to create meaningful solutions to real-world problems. This research represents a step toward a healthier future, where we leverage AI not just as a tool but as a partner in addressing complex health challenges like obesity. By focusing on lifestyle factors and ensemble learning methods, we aim to provide healthcare professionals with actionable insights while empowering individuals to make informed decisions about their well-being.” Said Dr. Hemachandran Kannan.

The team found that their approach contributes to a more comprehensive understanding of obesity risk factors, aiding healthcare providers in delivering targeted interventions based on specific obesity levels.
The researchers observed that while BMI (body mass index) is commonly used as the primary indicator of obesity risk, it has limitations in capturing the full complexity of obesity, which is influenced by behavioral, environmental, and genetic factors. They noted that BMI does not account for critical health indicators such as muscle mass, fat distribution, or other variables, potentially reducing the precision of obesity classification.

Additionally, their study highlighted that while some machine learning models for obesity classification incorporate lifestyle factors, sex, and 3D body scans without relying solely on BMI, the datasets used in these models often stem from smaller sample sizes in specific regions or countries. This limited scope restricts the models’ ability to generalize findings to broader or global populations.

The study also noted that the perfect climate for obesity to flourish has been created by traditional diets heavy in processed foods and low physical activity as technology develops and urbanization picks up speed. The resultant effects include an increase in disorders linked to obesity, such as diabetes, heart problems, and other ailments.
Rege emphasized that detecting illnesses early and identifying risk factors can serve as a strong motivator for individuals. “Armed with this knowledge, people are more empowered to make healthier choices regarding their diet, lifestyle, and exercise,” he stated. “Early identification of obesity-related issues is particularly advantageous, as it enables timely interventions and lifestyle adjustments.”

The researchers utilized a publicly available dataset drawn from diverse populations in countries like Colombia, Peru, and Mexico, incorporating factors such as eating habits, age, sex, physical condition, water and alcohol consumption, and the frequency of vegetable intake. This dataset included a broad range of features, including dietary patterns, physical activity levels, mental health, and sleep habits.

The team is hopeful that their findings will aid in developing more effective strategies for preventing and addressing obesity. They also proposed that future studies could leverage deep learning methods to improve the accuracy of obesity risk detection and prediction.

This development highlights Woxsen University’s commitment to tackling real-world challenges through innovation, fostering international collaborations, and advancing technology for the greater good.