Artificial Intelligence based Healthcare Knowledge Structure for Automatic Medical Report Generation

Automatic medical report production is a key text summarization application. Traditional text summarizing systems cannot completely interpret and use medical texts’ complicated vocabulary due to the disparities between medical consultation data and general domain data. The crucial knowledge in medical consultations has not been successfully harnessed. Traditional text summary algorithms cannot automatically identify, and filter key information based on medical reports’ structure and generate structured text. An AI-based knowledge-assisted structured medical report creation solution addresses these difficulties. This method uses entity-guided previous domain knowledge and structure-guided task decoupling to fully utilize medical consultation data and medical report structure. The suggested strategy improves ROUGE scores by 2% to 3% compared to similar methods, resulting in more accurate medical reports, according to IMCS21 dataset experiments.