Digital representations of past, current, and future health using vectors

An apparatus comprising at least one processor and at least one memory including computer program code for one or more programs, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following:

• access a plurality of instances of medical information encoded using medical codes and corresponding to a population of patients comprising a plurality of patients;
• generate a plurality of medical sentences corresponding to the population of patients by, for each patient of the population of patients, generating one medical sentence based on one or more instances of the plurality of instances of medical information corresponding to the patient and a timestamp associated with each instance of the one or more instances such that the one medical sentence corresponding to the patient comprises one or more medical codes in a chronological order;
• generate a vector dictionary comprising a plurality of multi-dimensional vectors (a) based on a vector generation model trained using machine learning and the plurality of medical sentences, and (b) in a multi-dimensional space having a configurable dimensionality, each multi-dimensional vector corresponding to a medical code;
• generate a digital representation of a first patient of the population of patients using an anagram model and the multi-dimensional vectors of the vector dictionary;
• determine a health state of the first patient based on the digital representation of the first patient;
• and
• provide an output indicating the health state of the first patient.

A computer program product comprising at least one non-transitory computer-readable storage medium having computer-executable program code portions stored therein, the computer-executable program code portions comprising program code instructions, the computer program code instructions, when executed by a processor of a computing entity, are configured to cause the computing entity to at least:

• access a plurality of instances of medical information encoded using medical codes and corresponding to a population of patients comprising a plurality of patients;
• generate a plurality of medical sentences corresponding to the population of patients by, for each patient of the population of patients, generating one medical sentence based on one or more instances of the plurality of instances of medical information corresponding to the patient and a timestamp associated with each instance of the one or more instances such that the one medical sentence corresponding to the patient comprises one or more medical codes in a chronological order;
• generate a vector dictionary comprising a plurality of multi-dimensional vectors (a) based on a vector generation model trained using machine learning and the plurality of medical sentences, and (b) in a multi-dimensional space having a configurable dimensionality, each multi-dimensional vector corresponding to a medical code;
• generate a digital representation of a first patient of the population of patients using an anagram model and the multi-dimensional vectors of the vector dictionary;
• determine a health state of the first patient based on the digital representation of the first patient;
• and
• provide an output indicating the health state of the first patient.