For the first time in Bangladesh, researchers at Bangladesh Agricultural University (BAU) have successfully used machine learning, a form of artificial intelligence (AI), to identify risk factors, diagnose, and propose effective preventive measures for brucellosis, a deadly zoonotic disease affecting both animals and humans.
The groundbreaking research was conducted by Professor Dr Md Siddiqur Rahman from the Department of Medicine, Faculty of Veterinary Science, alongside his PhD researcher and former Colonel SM Azizul Karim Hussaini. Their study, recently published in the Asian Journal of Agriculture and Biology, utilised five machine learning algorithms—including MLP, Deep Learning 4j, AdaBoost MI, and J48 Tree—to identify key risk factors associated with brucellosis.
Dr Siddiqur highlighted the disease’s severe economic impact due to reduced milk production, abortions, and poor reproductive performance in livestock, as well as its significant public health risk as a zoonotic infection. “There is currently no effective treatment for brucellosis, and mass culling is not feasible due to economic limitations and religious sentiments. Our AI-driven approach recommends targeted treatment for valuable animals, particularly those with high exotic genetics,” he explained.
The study also found that while live vaccines such as B. abortus (S-19) and RB51 carry side effects, heat-killed vaccines have proven safer and more effective, paving the way for a new generation of vaccines in Bangladesh.
PhD researcher Hussaini described the study as a milestone for Bangladesh’s veterinary science community, demonstrating AI’s potential to revolutionise disease control in animal health, similar to its established role in diagnosing human diseases like heart and kidney disorders globally.
The research received technical support from Germany’s Friedrich-Loeffler-Institut and Saudi Arabia’s King Faisal University, reflecting international collaboration in advancing Bangladesh’s veterinary diagnostics and public health protection.
