Determination of health sciences recommendations based on multidimensional nonlinear manifold clustering
A method and apparatus can include a system controller and a system processor. The system controller can retrieve a health science related dataset from at least one database, the retrieved dataset including information associated with at least one of a patient medical information, healthcare provider clinical information, health related publications and treatment information, and pharmaceutical information, and transmit to a user equipment a recommendation of at least one of health diagnosis and treatment for a patient. The system processor can utilize multidimensional nonlinear manifold clustering on at least one element from the retrieved dataset, assign an entity formulated from the at least one element of the retrieved dataset into a decision hyper-volume based on the multidimensional nonlinear manifold clustering, and determine the recommendation of at least one of health diagnosis and treatment for the patient based on the assignment of the entity into the decision hyper-volume.
1 . An apparatus, comprising:
a system controller to
retrieve a health science related dataset from at least one database, the retrieved dataset including a plurality of attributes, each of the plurality of attributes associated with at least one of a patient medical information, healthcare provider clinical information, health related publications and treatment information, and pharmaceutical information, and
transmit to a user equipment a recommendation of at least one of health diagnosis and treatment for a particular patient, wherein the user equipment is separate from the system controller; and
a system processor to
formulate a plurality of entities, the plurality of entities formulated based on the retrieved dataset and based on a particular patient entity formulated for the particular patient from at least one attribute of the plurality of attributes,
separate, utilizing multidimensional nonlinear manifold clustering, the plurality of entities into a plurality of hyper-volumes along a nonlinear manifold, each hyper-volume including at least one entity of the plurality of entities, an entity of the at least one entity being defined by a set of the plurality of attributes defining a point in the respective hyper-volume along the nonlinear manifold, a decision hyper-volume of the plurality of hyper-volumes including a plurality of decision hyper-volume entities, the plurality of decision hyper-volume entities including the particular patient entity, and
determine the recommendation of at least one of health diagnosis and treatment for the particular patient based on assignment, using the multidimensional nonlinear manifold clustering, of the particular patient entity into the decision hyper-volume.
2 . The apparatus of claim 1 , wherein the system processor further formulates the particular patient entity from at least one attribute of the plurality of attributes of the retrieved dataset.
3 . The apparatus of claim 1 , wherein the system processor further performs adjudication to minimize a loss function in order to optimize the assignment of the particular patient entity by adjusting a decision boundary of the decision hyper-volume.
4 . The apparatus of claim 1 , wherein the utilization of the multidimensional nonlinear manifold clustering further includes utilizing the multidimensional nonlinear manifold clustering and at least one other type of multidimensional clustering including at least one of linear clustering, linear manifold clustering, and nonlinear clustering.
5 . The apparatus of claim 1 , wherein the assignment of the particular patient entity into the decision hyper-volume is further based on at least one other type of multidimensional clustering including at least one of linear clustering, linear manifold clustering, and nonlinear clustering.
6 . The apparatus of claim 1 , wherein the system processor further performs adjudication to minimize a loss function in order to optimize the assignment of the particular patient entity by adjusting a decision boundary of the decision hyper-volume for the multidimensional nonlinear manifold clustering and at least one other type of multidimensional clustering including at least one of linear clustering, linear manifold clustering, and nonlinear clustering, and selects at least one of the multidimensional nonlinear manifold clustering and the at least one other type of multidimensional clustering based on a minimum loss function and optimum entity assignment.
7 . The apparatus of claim 1 , wherein the system processor further performs adjudication to control a number of iterations of selection of the retrieved dataset and compares multiple iterations of the multidimensional nonlinear manifold clustering and at least one other type of multidimensional clustering.
8 . The apparatus of claim 1 , wherein the system processor further utilizes a neural network to define the decision hyper-volume.
9 . The apparatus of claim 1 , wherein the at least one of health diagnosis includes at least one of a disease and a health condition and the treatment is at least one of surgical, radiological, pharmaceutical, deoxyribonucleic acid modification, ribonucleic acid modification, immunotherapy, environment change, and life style change.
10 . The apparatus of claim 1 , wherein the treatment includes at least one of discovery of a new drug, new application for an existing drug, repurpose an existing drug, drug interactions, drug side-effects, drug dosage, drug efficacy, drug sensitivity, and discovery of a complementary drug.
11 . The apparatus of claim 1 , wherein the at least one of health diagnosis and treatment is based on at least one of deoxyribonucleic acid, ribonucleic acid, phenotype, genotype, biomarker, exosome, environment, and life style.
12 . The apparatus of claim 1 , wherein the system processor further places the particular patient entity into an entity universe based on Monte Carlo insertion with or without measure of importance.
13 . The apparatus of claim 1 , wherein the system processor further divides the particular patient entity into two or more sub-entities.
14 . The apparatus of claim 1 , wherein the system processor further formulates the particular patient entity using text processing including performing at least one of Information Theoretic, semantic, and syntactic, processing, or extracting numerically encoded text features from at least one of bulk text, structured text, unstructured text, biological sequences, chemical sequences, deoxyribonucleic acid, and ribonucleic acid.
15 . The apparatus of claim 1 , wherein the system processor further applies the assignment of the particular patient entity regionally to an adjacent hyper-volume sharing a decision boundary.
16 . The apparatus of claim 1 , wherein the system processor further formulates a confidence region to rank an order of a plurality of the recommendation of the at least one of health diagnosis and treatment.
17 . The apparatus of claim 1 , wherein the system processor further assigns a decision of the at least one of health diagnosis and treatment to the decision hyper-volume based on at least one of supervised training and unsupervised training.
18 . The apparatus of claim 1 , wherein the system processor further performs adjudication to formulate at least one of the retrieved dataset and the particular patient entity, to optimize the recommendation of at least one of diagnosis and treatment.
19 . The apparatus of claim 1 , wherein at least one of the system processor and system controller are at least partially implemented in at least one of a local computing, distributed computing, mobile computing, cloud-based computing, Graphics Processing Unit (GPU), array processing, Field Programmable Gate Arrays (FPGA), tensor processing, Application Specific Integrated Circuits (ASIC), quantum computing, and a software program.
20 . A method, comprising:
retrieving, by a system controller, a health science related dataset from at least one database, the retrieved dataset including a plurality of attributes, each of the plurality of attributes associated with at least one of a patient medical information, healthcare provider clinical information, health related publications and treatment information, and pharmaceutical information;
formulating, by a system processor, a plurality of entities, the plurality of entities formulated based on the retrieved dataset and based on a particular patient entity formulated for a particular patient from at least one attribute of the plurality of attributes;
separating, by the system processor, utilizing multidimensional nonlinear manifold clustering, the plurality of entities into a plurality of hyper-volumes along a nonlinear manifold, each hyper-volume including at least one entity of the plurality of entities, an entity of the at least one entity being defined by a set of the plurality of attributes defining a point in the respective hyper-volume along the nonlinear manifold, a decision hyper-volume of the plurality of hyper-volumes including a plurality of decision hyper-volume entities, the plurality of decision hyper-volume entities including the particular patient entity;
determining, by the system processor, a recommendation of at least one of health diagnosis and treatment for a particular patient based on assignment, using the multidimensional nonlinear manifold clustering, of the particular patient entity into the decision hyper-volume; and
transmitting, by the system controller, to a user equipment the recommendation of at least one of health diagnosis and treatment for the particular patient, wherein the user equipment is separate from the system controller.
21 . The method of claim 20 , further comprising formulating, by the system processor, the particular patient entity from at least one attribute of the plurality of attributes of the retrieved dataset.
22 . The method of claim 20 , further comprising performing adjudication, by the system processor, to minimize a loss function in order to optimize the assignment of the particular patient entity by adjusting a decision boundary of the decision hyper-volume.
23 . The method of claim 20 , wherein the utilization of the multidimensional nonlinear manifold clustering further includes utilizing the multidimensional nonlinear manifold clustering and at least one other type of multidimensional clustering including at least one of linear clustering, linear manifold clustering, and nonlinear clustering.
24 . The method of claim 20 , wherein the assignment of the particular patient entity into the decision hyper-volume is further based on at least one other type of multidimensional clustering including at least one of linear clustering, linear manifold clustering, and nonlinear clustering.
25 . The method of claim 20 , further comprising:
performing adjudication, by the system processor, to minimize a loss function in order to optimize the assignment of the particular patient entity by adjusting a decision boundary of the decision hyper-volume for the multidimensional nonlinear manifold clustering and at least one other type of multidimensional clustering including at least one of linear clustering, linear manifold clustering, and nonlinear clustering, and
selecting, by the system processor, at least one of the multidimensional nonlinear manifold clustering and the at least one other type of multidimensional clustering based on a minimum loss function and optimum entity assignment.
26 . The method of claim 20 , further comprising performing adjudication, by the system processor, to control a number of iterations of selection of the retrieved health science related dataset and comparing, by the system processor, multiple iterations of the multidimensional nonlinear manifold clustering and at least one other type of multidimensional clustering.
27 . The method of claim 20 , further comprising utilizing, by the system processor, a neural network to define the decision hyper-volume.
28 . The method of claim 20 , wherein the at least one of health diagnosis includes at least one of a disease and a health condition and the treatment is at least one of surgical, radiological, pharmaceutical, deoxyribonucleic acid modification, ribonucleic acid modification, immunotherapy, environment change, and life style change.
29 . The method of claim 20 , wherein the treatment includes at least one of discovery of a new drug, new application for an existing drug, repurpose an existing drug, drug interactions, drug side-effects, drug dosage, drug efficacy, drug sensitivity, and discovery of a complementary drug.
30 . The method of claim 20 , wherein the at least one of health diagnosis and treatment is based on at least one of deoxyribonucleic acid, ribonucleic acid, phenotype, genotype, biomarker, exosome, environment, and life style.
31 . The method of claim 20 , further comprising placing, by the system processor, the particular patient entity into an entity universe based on Monte Carlo insertion with or without measure of importance.
32 . The method of claim 20 , further comprising dividing, by the system processor, the particular patient entity into two or more sub-entities.
33 . The method of claim 20 , further comprising formulating, by the system processor, the particular patient entity using text processing including performing at least one of Information Theoretic, semantic, and syntactic, processing, or extracting numerically encoded text features from at least one of bulk text, structured text, unstructured text, biological sequences, chemical sequences, deoxyribonucleic acid, and ribonucleic acid.
34 . The method of claim 20 , further comprising applying, by the system processor, the assignment of the particular patient entity regionally to an adjacent hyper-volume sharing a decision boundary.
35 . The method of claim 20 , further comprising formulating, by the system processor, a confidence region to rank an order of a plurality of the recommendation of at least one of health diagnosis and treatment.
36 . The method of claim 20 , further comprising assigning, by the system processor, a decision of the at least one of health diagnosis and treatment to the decision hyper-volume based on at least one of supervised training and unsupervised training.
37 . The method of claim 20 , further comprising performing adjudication, by the system processor, to formulate at least one of the retrieved dataset and the particular patient entity, to optimize the recommendation of at least one of diagnosis and treatment.
38 . The method of claim 20 , wherein at least one of the system processor and system controller are at least partially implemented in at least one of a local computing, distributed computing, mobile computing, cloud-based computing, Graphics Processing Unit (GPU), array processing, Field Programmable Gate Arrays (FPGA), tensor processing, Application Specific Integrated Circuits (ASIC), quantum computing, and a software program.