A team of researchers from the University of Michigan has developed an AI model to predict treatment outcomes for tuberculosis (TB) patients. By analyzing diverse biomedical data including clinical tests, genomics, medical imaging, and drug prescriptions from TB patients worldwide, the team aims to personalize TB treatment. The multimodal AI model accurately predicted treatment prognosis and outperformed existing models by identifying drug regimens effective against specific sets of drug-resistant TB patients.

The researchers examined over 5, 000 patients, considering various factors such as demographics, prior treatment history, comorbidities, imaging features, and genomic features. The study also found that certain drug combinations and pharmaceutical interactions could impact treatment success. The AI model has the potential to guide personalized treatments and provide better outcomes for TB patients.