BAU researchers make first-ever use of AI to combat Brucellosis

For the first time in Bangladesh, researchers at Bangladesh Agricultural University (BAU) claim to have successfully utilised machine learning, a form of artificial intelligence, to identify risk factors, diagnose, and propose effective preventive measures for Brucellosis, which is a deadly zoonotic disease affecting both animals and humans.

 

The research was conducted by Professor Dr. Md. Siddiqur Rahman of the Department of Medicine under the Faculty of Veterinary Science, along with his PhD researcher Colonel (Retd) SM Azizul Karim Hussaini.

 

They assert that this AI-driven approach has the potential to revolutionise Brucellosis control and prevention in the country.

 

"Brucellosis is a globally significant infectious disease that affects domestic livestock, wildlife, and even marine mammals. It causes considerable economic losses to farmers due to reduced milk production, abortions, and declining reproductive performance," said Dr. Rahman.

 

"Moreover, as a zoonotic disease, it can be easily transmitted from animals to humans, posing serious public health risks."

 

The study was recently published in the Asian Journal of Agriculture and Biology, a Scopus-indexed journal with an impact factor of 1.6. Dr. Rahman served as the principal investigator of the study, while Dr. Heinrich Neubauer of Germany's Friedrich-Loeffler-Institut was the co-supervisor. The project also received technical support from King Faisal University in Saudi Arabia.

 

Addressing the media, Dr. Rahman explained, “There is currently no effective treatment for this disease, and farmers have long suffered due to its widespread impact. By applying five machine learning algorithms - including MLP, Deep Learning 4j, AdaBoost MI, and J48 Tree - we successfully identified key risk factors associated with Brucellosis. Among them, these four models proved to be the most effective.”

 

He added, “While machine learning is already being used worldwide to detect heart disease, kidney disorders, diabetes, and respiratory conditions in humans, its application in animal health is still emerging in Bangladesh. Given the country’s economic limitations and religious sentiments, mass culling of infected animals is not a feasible option. Instead, we recommend targeted treatment for valuable animals, especially those with a high proportion of exotic genetics.”

 

The study found that among the 12 species of Brucella bacteria, B. abortus, B. suis, B. melitensis, and B. canis are the most pathogenic and harmful.

 

Dr. Rahman further highlighted that although live vaccines like B. abortus (S-19) and RB51 are currently in use, they pose a risk of side effects. In contrast, heat-killed vaccines have proven to be safer and more effective in the study, paving the way for the development of a new generation of vaccines for Brucellosis prevention in Bangladesh.

 

PhD researcher Colonel (Retd) SM Azizul Karim Husaini stated, “Under the direct supervision of Professor Dr. Md. Siddiqur Rahman, we have demonstrated that machine learning techniques can be effectively applied to understand the transmission dynamics and preventive strategies for Brucellosis.”

 

He added, “We followed the correct protocols and achieved the desired outcomes. This success stands as a proud milestone for both the nation and the veterinary science community.”