In order to effectively use the parallel processing capabilities of the cloud platform Hadoop framework, an improved data mining algorithm based on Map Reduce parallel mode is proposed by analyzing and improving a traditional association rules algorithmApriori algorithm which belong to big data mining technology, which is suitable for the analysis and application of medical big data. First, the Boolean arrangement is used to optimize the storage mode of transaction data in the database, and it will effectively reduce the number of database scanned. Then, the association rule optimization is used to reduce the redundant subsets in the Apriori algorithm. In order to verify the effectiveness of the improved algorithm, medical history data is used to verify the experiment. Finally, the simulation results show that the proposed algorithm is more efficient and has better reliability and validity as compared with the traditional Apriori algorithm.