Traditional evaluations of human disease model mice for drug development have primarily focused on hematological and biochemical analyses, as well as pathological assessments. Recently, non-invasive live imaging techniques such as MRI and CT have been introduced. However, there is a pressing need to identify early disease pathogenesis promptly when utilizing model animals for drug research. Thus, the development of a new animal experimental method is essential, which involves continuous, comprehensive monitoring of animal behavior through digital analysis.  We propose the use of high-resolution IR cameras and RFID readers in custom-designed cages to acquire continuous and comprehensive behavioral data for disease model mice. By embedding chips in individual mice, we can analyze both individual and group behavior among multiple animals. Furthermore, this approach allows for the quantitative capture of symptom exacerbation and improvement over time. Additionally, the behavioral data obtained in this study will be subjected to "unsupervised learning" using artificial intelligence (AI). We expect that AI-based analysis will enable the detection of subtle behavioral changes in disease model mice that may go unnoticed to the human eye.