Abstract: In one example, an apparatus comprises a treatment metric model database that stores a plurality of treatment metric models, each of the plurality of treatment metric models being generated for a first cluster of cluster features vectors, each cluster feature vector representing a distribution for each of a plurality of data categories of patients data represented by patient feature vectors clustered in a second cluster. The apparatus is further configured to: receive patient characteristics features data of a patient; identify a first treatment metric model based on the patient characteristics features data of the patient; input the first patient characteristic features data and data representing a range of administrations of the treatment to the first treatment metric model to compute a range of the treatment metric; and select an administration of the treatment from the range of administrations of the treatment having optimal treatment metric.

Abstract: In one example, an apparatus comprises a treatment metric model database that stores a plurality of treatment metric models, each of the plurality of treatment metric models being generated for a first cluster of cluster features vectors, each cluster feature vector representing a distribution for each of a plurality of data categories of patients data represented by patient feature vectors clustered in a second cluster. The apparatus is further configured to: receive patient characteristics features data of a patient; identify a first treatment metric model based on the patient characteristics features data of the patient; input the first patient characteristic features data and data representing a range of administrations of the treatment to the first treatment metric model to compute a range of the treatment metric; and select an administration of the treatment from the range of administrations of the treatment having optimal treatment metric.

Abstract: In one example, an apparatus comprises a treatment metric model database that stores a plurality of treatment metric models, each of the plurality of treatment metric models being generated for a first cluster of cluster features vectors, each cluster feature vector representing a distribution for each of a plurality of data categories of patients data represented by patient feature vectors clustered in a second cluster. The apparatus is further configured to: receive patient characteristics features data of a patient; identify a first treatment metric model based on the patient characteristics features data of the patient; input the first patient characteristic features data and data representing a range of administrations of the treatment to the first treatment metric model to compute a range of the treatment metric; and select an administration of the treatment from the range of administrations of the treatment having optimal treatment metric.