The invention relates to the technical field of
anorexia risk prediction, and discloses an
anorexia risk
prediction system based on multi-
modal data, which comprises a multi-dimensional monitoring module and an intelligent evaluation module. According to the
system, health management data and daily
life management data of all individuals are obtained through a multi-dimensional monitoring module and are classified to form a
data set, and an intelligent evaluation module evaluates the actual
energy demand required by the individuals for maintaining the current
body weight, generates basic
energy consumption, evaluates the individual
physical activity level coefficient, generates an energy intake gap, and obtains the health management data and the daily
life management data. Metabolic characteristics of different ages, genders and groups are fully considered, the multi-dimensional assessment precision is high, the intelligent assessment module sets a fixed monitoring period, dynamically tracks the change trend of health key indexes, accurately quantifies and assesses the health injury degree of an individual caused by
anorexia behaviors, generates injury indexes, quantifies risk levels, and improves the assessment accuracy. Intendencies of individuals to different anorexia subtypes are distinguished, corresponding early warning signals and feeding suggestions are output, and grading early warning screening efficiency is high.