Body Scan images - Tuberculosis

Marie.IA platform with 1 million images (X-rays and CT scans) of patients with tuberculosis and other lung diseases. The platform contains about 300 algorithms that extract different features and variables due to the multimodal model for screening and/or support in early diagnosis. In this phase, we have a proof of concept of identifying whether previous training with X-ray and CT images is sufficient to identify tuberculosis characteristics and patterns in Body Scan images. Another important aspect of using the platform was the high specificity of the model. In this project, we obtained promising data, as the result of the model showed an specificity of 87,23% and sensitivity of 100% for identifying cases with suspected tuberculosis.