CN117995430A - Multi-drug treatment response prediction method based on blood viscosity change - Google Patents
Multi-drug treatment response prediction method based on blood viscosity change Download PDFInfo
- Publication number
- CN117995430A CN117995430A CN202410127244.XA CN202410127244A CN117995430A CN 117995430 A CN117995430 A CN 117995430A CN 202410127244 A CN202410127244 A CN 202410127244A CN 117995430 A CN117995430 A CN 117995430A
- Authority
- CN
- China
- Prior art keywords
- drug
- patient
- treatment
- data
- medication
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000003814 drug Substances 0.000 title claims abstract description 379
- 229940079593 drug Drugs 0.000 title claims abstract description 352
- 238000011282 treatment Methods 0.000 title claims abstract description 180
- 239000008280 blood Substances 0.000 title claims abstract description 95
- 210000004369 blood Anatomy 0.000 title claims abstract description 95
- 238000000034 method Methods 0.000 title claims abstract description 62
- 230000008859 change Effects 0.000 title claims abstract description 32
- 230000004044 response Effects 0.000 title claims abstract description 26
- 201000010099 disease Diseases 0.000 claims abstract description 60
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims abstract description 60
- 238000012544 monitoring process Methods 0.000 claims abstract description 60
- 230000036541 health Effects 0.000 claims abstract description 41
- 230000003993 interaction Effects 0.000 claims abstract description 32
- 230000002439 hemostatic effect Effects 0.000 claims abstract description 25
- 206010067484 Adverse reaction Diseases 0.000 claims abstract description 20
- 230000006838 adverse reaction Effects 0.000 claims abstract description 20
- 230000008569 process Effects 0.000 claims abstract description 18
- 206010013710 Drug interaction Diseases 0.000 claims abstract description 16
- 239000000890 drug combination Substances 0.000 claims abstract description 15
- 238000000518 rheometry Methods 0.000 claims abstract description 10
- 238000002483 medication Methods 0.000 claims abstract description 6
- 239000002547 new drug Substances 0.000 claims abstract description 6
- 238000012827 research and development Methods 0.000 claims abstract description 4
- 230000000694 effects Effects 0.000 claims description 105
- 238000004422 calculation algorithm Methods 0.000 claims description 56
- 238000004458 analytical method Methods 0.000 claims description 45
- 238000005516 engineering process Methods 0.000 claims description 26
- 238000011269 treatment regimen Methods 0.000 claims description 26
- 238000007726 management method Methods 0.000 claims description 23
- 230000007246 mechanism Effects 0.000 claims description 21
- 238000004393 prognosis Methods 0.000 claims description 19
- 238000011156 evaluation Methods 0.000 claims description 18
- 230000001225 therapeutic effect Effects 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 13
- 238000010801 machine learning Methods 0.000 claims description 13
- 238000011160 research Methods 0.000 claims description 13
- 238000013527 convolutional neural network Methods 0.000 claims description 11
- 238000013135 deep learning Methods 0.000 claims description 11
- 238000002651 drug therapy Methods 0.000 claims description 10
- 235000012041 food component Nutrition 0.000 claims description 9
- 239000005417 food ingredient Substances 0.000 claims description 9
- 238000010276 construction Methods 0.000 claims description 8
- 238000007418 data mining Methods 0.000 claims description 8
- 238000003066 decision tree Methods 0.000 claims description 8
- 238000005065 mining Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 8
- 230000009471 action Effects 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 7
- 238000012706 support-vector machine Methods 0.000 claims description 7
- 230000017531 blood circulation Effects 0.000 claims description 6
- 238000007405 data analysis Methods 0.000 claims description 6
- 238000013523 data management Methods 0.000 claims description 6
- 238000001647 drug administration Methods 0.000 claims description 6
- 230000007774 longterm Effects 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 230000002411 adverse Effects 0.000 claims description 5
- 238000002790 cross-validation Methods 0.000 claims description 5
- 230000000004 hemodynamic effect Effects 0.000 claims description 5
- 230000004060 metabolic process Effects 0.000 claims description 5
- 229940126585 therapeutic drug Drugs 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000013473 artificial intelligence Methods 0.000 claims description 4
- 238000004159 blood analysis Methods 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 4
- 238000013136 deep learning model Methods 0.000 claims description 4
- 230000000857 drug effect Effects 0.000 claims description 4
- 230000008713 feedback mechanism Effects 0.000 claims description 4
- 235000016709 nutrition Nutrition 0.000 claims description 4
- 230000000737 periodic effect Effects 0.000 claims description 4
- 230000002688 persistence Effects 0.000 claims description 4
- 230000000144 pharmacologic effect Effects 0.000 claims description 4
- 238000012502 risk assessment Methods 0.000 claims description 4
- 238000007619 statistical method Methods 0.000 claims description 4
- 208000024891 symptom Diseases 0.000 claims description 4
- 238000012795 verification Methods 0.000 claims description 4
- 230000005856 abnormality Effects 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 claims description 3
- 238000013500 data storage Methods 0.000 claims description 3
- 238000011161 development Methods 0.000 claims description 3
- 235000005911 diet Nutrition 0.000 claims description 3
- 230000037213 diet Effects 0.000 claims description 3
- 238000009509 drug development Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 230000023597 hemostasis Effects 0.000 claims description 3
- 238000011835 investigation Methods 0.000 claims description 3
- 230000002503 metabolic effect Effects 0.000 claims description 3
- 239000002207 metabolite Substances 0.000 claims description 3
- 235000015097 nutrients Nutrition 0.000 claims description 3
- 230000035764 nutrition Effects 0.000 claims description 3
- 238000011369 optimal treatment Methods 0.000 claims description 3
- 238000003909 pattern recognition Methods 0.000 claims description 3
- 230000036470 plasma concentration Effects 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 230000008901 benefit Effects 0.000 claims description 2
- 238000012482 interaction analysis Methods 0.000 claims description 2
- 230000004797 therapeutic response Effects 0.000 claims description 2
- 238000013329 compounding Methods 0.000 claims 1
- 238000009223 counseling Methods 0.000 claims 1
- 238000013210 evaluation model Methods 0.000 claims 1
- 238000011337 individualized treatment Methods 0.000 claims 1
- BSYNRYMUTXBXSQ-UHFFFAOYSA-N Aspirin Chemical compound CC(=O)OC1=CC=CC=C1C(O)=O BSYNRYMUTXBXSQ-UHFFFAOYSA-N 0.000 description 7
- 229960001138 acetylsalicylic acid Drugs 0.000 description 7
- 238000007477 logistic regression Methods 0.000 description 5
- 206010020772 Hypertension Diseases 0.000 description 4
- 239000003146 anticoagulant agent Substances 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 230000015271 coagulation Effects 0.000 description 4
- 238000005345 coagulation Methods 0.000 description 4
- 230000035488 systolic blood pressure Effects 0.000 description 4
- 239000005541 ACE inhibitor Substances 0.000 description 3
- 206010061623 Adverse drug reaction Diseases 0.000 description 3
- 208000030453 Drug-Related Side Effects and Adverse reaction Diseases 0.000 description 3
- 206010000059 abdominal discomfort Diseases 0.000 description 3
- 229940044094 angiotensin-converting-enzyme inhibitor Drugs 0.000 description 3
- 229940127219 anticoagulant drug Drugs 0.000 description 3
- 230000036772 blood pressure Effects 0.000 description 3
- 230000000747 cardiac effect Effects 0.000 description 3
- 238000010219 correlation analysis Methods 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000003907 kidney function Effects 0.000 description 3
- 238000003058 natural language processing Methods 0.000 description 3
- 238000007637 random forest analysis Methods 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 2
- 102000004190 Enzymes Human genes 0.000 description 2
- 108090000790 Enzymes Proteins 0.000 description 2
- 229940121710 HMGCoA reductase inhibitor Drugs 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 239000000090 biomarker Substances 0.000 description 2
- 230000000740 bleeding effect Effects 0.000 description 2
- 230000037396 body weight Effects 0.000 description 2
- HVYWMOMLDIMFJA-DPAQBDIFSA-N cholesterol Chemical compound C1C=C2C[C@@H](O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2 HVYWMOMLDIMFJA-DPAQBDIFSA-N 0.000 description 2
- 230000006735 deficit Effects 0.000 description 2
- 238000012217 deletion Methods 0.000 description 2
- 230000037430 deletion Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 239000002934 diuretic Substances 0.000 description 2
- 229940030606 diuretics Drugs 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 239000004615 ingredient Substances 0.000 description 2
- ZCCUUQDIBDJBTK-UHFFFAOYSA-N psoralen Chemical compound C1=C2OC(=O)C=CC2=CC2=C1OC=C2 ZCCUUQDIBDJBTK-UHFFFAOYSA-N 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000002560 therapeutic procedure Methods 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 206010000087 Abdominal pain upper Diseases 0.000 description 1
- 206010053567 Coagulopathies Diseases 0.000 description 1
- 206010010774 Constipation Diseases 0.000 description 1
- 102000004328 Cytochrome P-450 CYP3A Human genes 0.000 description 1
- 108010081668 Cytochrome P-450 CYP3A Proteins 0.000 description 1
- 108010000437 Deamino Arginine Vasopressin Proteins 0.000 description 1
- 206010061818 Disease progression Diseases 0.000 description 1
- 108010049003 Fibrinogen Proteins 0.000 description 1
- 102000008946 Fibrinogen Human genes 0.000 description 1
- 208000012671 Gastrointestinal haemorrhages Diseases 0.000 description 1
- 206010062713 Haemorrhagic diathesis Diseases 0.000 description 1
- 208000032843 Hemorrhage Diseases 0.000 description 1
- 102000004286 Hydroxymethylglutaryl CoA Reductases Human genes 0.000 description 1
- 108090000895 Hydroxymethylglutaryl CoA Reductases Proteins 0.000 description 1
- 208000031226 Hyperlipidaemia Diseases 0.000 description 1
- 238000000342 Monte Carlo simulation Methods 0.000 description 1
- 206010028813 Nausea Diseases 0.000 description 1
- 208000025174 PANDAS Diseases 0.000 description 1
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 description 1
- 240000000220 Panda oleosa Species 0.000 description 1
- 235000016496 Panda oleosa Nutrition 0.000 description 1
- 238000010220 Pearson correlation analysis Methods 0.000 description 1
- 206010047700 Vomiting Diseases 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000000702 anti-platelet effect Effects 0.000 description 1
- 229940030600 antihypertensive agent Drugs 0.000 description 1
- 239000002220 antihypertensive agent Substances 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 235000012000 cholesterol Nutrition 0.000 description 1
- 230000035602 clotting Effects 0.000 description 1
- 238000011284 combination treatment Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 229960004281 desmopressin Drugs 0.000 description 1
- NFLWUMRGJYTJIN-NXBWRCJVSA-N desmopressin Chemical compound C([C@H]1C(=O)N[C@H](C(N[C@@H](CC(N)=O)C(=O)N[C@@H](CSSCCC(=O)N[C@@H](CC=2C=CC(O)=CC=2)C(=O)N1)C(=O)N1[C@@H](CCC1)C(=O)N[C@@H](CCCNC(N)=N)C(=O)NCC(N)=O)=O)CCC(=O)N)C1=CC=CC=C1 NFLWUMRGJYTJIN-NXBWRCJVSA-N 0.000 description 1
- 230000005750 disease progression Effects 0.000 description 1
- 230000036267 drug metabolism Effects 0.000 description 1
- 230000029142 excretion Effects 0.000 description 1
- 229940012952 fibrinogen Drugs 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 208000030304 gastrointestinal bleeding Diseases 0.000 description 1
- 210000001035 gastrointestinal tract Anatomy 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000007614 genetic variation Effects 0.000 description 1
- 235000015201 grapefruit juice Nutrition 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 208000006750 hematuria Diseases 0.000 description 1
- 208000031169 hemorrhagic disease Diseases 0.000 description 1
- 235000009200 high fat diet Nutrition 0.000 description 1
- 230000001631 hypertensive effect Effects 0.000 description 1
- 230000002218 hypoglycaemic effect Effects 0.000 description 1
- 230000001077 hypotensive effect Effects 0.000 description 1
- 238000001727 in vivo Methods 0.000 description 1
- 239000003112 inhibitor Substances 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 229940080794 lovastatin 20 mg Drugs 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000010534 mechanism of action Effects 0.000 description 1
- 230000037323 metabolic rate Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 208000010125 myocardial infarction Diseases 0.000 description 1
- 230000008693 nausea Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000003334 potential effect Effects 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 229940126409 proton pump inhibitor Drugs 0.000 description 1
- 239000000612 proton pump inhibitor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000011285 therapeutic regimen Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000008673 vomiting Effects 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention provides a multi-drug treatment response prediction method based on blood viscosity change, which comprises the following steps: dynamically detecting blood rheology parameters through a full-automatic blood viscosity monitor, evaluating the blood viscosity change condition, and providing a basis for formulating a multi-drug treatment scheme; constructing a multi-drug interaction knowledge graph, collecting drug-drug, drug-food and drug-disease interaction relation data, and providing reference for compatibility selection of multi-drug combination; collecting compliance and life style information of patients in the multi-drug treatment process, establishing a model to evaluate the correlation of the factors with the drug efficacy and adverse reaction, and providing health management advice; an intelligent voice assistant for reminding and monitoring the medication is designed to carry out illness state tracking, consultation service and the like, so that a patient is helped to use various medications correctly; and detecting hemostatic parameters of the patient again, evaluating blood viscosity change conditions before and after multi-drug treatment, establishing a database, and collecting effective cases, thereby providing support for new drug research and development.
Description
Technical Field
The invention relates to the technical field of information, in particular to a multi-drug treatment response prediction method based on blood viscosity change.
Background
Currently, multi-drug therapies are widely used clinically, but face a number of challenges. Traditional clinical approaches mainly formulate treatment regimens based on symptoms of the disease and physiological indicators of the patient, ignoring complex correlations between individual differences and multiple drugs. This leads to uncertainty in the therapeutic effect and adverse effects of the patient during the course of treatment. Under the current technical framework, the introduction of blood viscosity monitors provides a new direction for multi-drug therapy. However, conventional blood parameter monitoring methods tend to be static and do not reflect dynamic blood rheology changes. Furthermore, the prior art is still contradictory with respect to a combination of individual differences and multi-drug interactions. The formulation of a therapeutic regimen requires more accurate guidance to ensure that the patient is able to achieve optimal efficacy and minimal adverse effects during the course of treatment. On the other hand, current treatment regimens lack personalized adjustments for different individuals. The existing curative effect prediction model is still not comprehensive in consideration of factors such as age, sex, basic diseases and the like of a patient, so that uncertainty of a curative effect is caused. In addition, the vital sign changes and patient lifestyle monitoring that exist in multi-drug therapy have not yet achieved real-time, comprehensive coverage. Under the complex application background of multiple drugs, the establishment of the existing multiple drug compatibility system and interaction knowledge graph still seems to be not intelligent enough. Lack of comprehensive consideration of individual parameters, severity of illness, etc., makes it difficult to provide accurate efficacy and risk prediction. In addition, assessment of patient compliance and lifestyle is lacking, limiting a comprehensive understanding of the various factors in the course of treatment. In summary, the prior art has contradictions and shortcomings in individuation, dynamization and intellectualization of multi-drug therapy. Therefore, there is a need to introduce an innovative approach that provides more accurate, personalized prediction and decision support for multi-drug therapies through comprehensive, dynamic, intelligent monitoring and analysis.
Disclosure of Invention
In view of the above, the present invention provides a multi-drug therapeutic response prediction method based on blood viscosity change, which mainly comprises the following steps:
Dynamically detecting blood rheology parameters through a full-automatic blood viscosity monitor, evaluating the blood viscosity change condition, and providing a basis for formulating a multi-drug treatment scheme;
constructing a large sample database containing patients with different ages, sexes and basic diseases, collecting clinical data such as treatment response, adverse reaction and the like of the patients in the multi-drug treatment process, analyzing the characteristics and the ending of the patients by using a deep learning technology, and establishing an accurate prognosis prediction model;
Designing an intelligent multi-drug compatibility system, predicting the curative effects and risks of different medication modes by using a machine learning algorithm according to individual parameters of patients, severity of illness and other factors, and providing decision support of a multi-drug treatment scheme;
The method comprises the steps of monitoring vital sign changes of a patient in a multi-drug treatment process in real time by adopting a wearable sensor, evaluating the safety of different drug administration schemes, and adjusting a drug administration plan as required by matching with the blood viscosity monitoring result;
constructing a multi-drug interaction knowledge graph, collecting drug-drug, drug-food and drug-disease interaction relation data, and providing reference for compatibility selection of multi-drug combination;
collecting compliance and life style information of patients in the multi-drug treatment process, establishing a model to evaluate the correlation of the factors with the drug efficacy and adverse reaction, and providing health management advice;
An intelligent voice assistant for reminding and monitoring the medication is designed to carry out illness state tracking, consultation service and the like, so that a patient is helped to use various medications correctly;
and detecting hemostatic parameters of the patient again, evaluating blood viscosity change conditions before and after multi-drug treatment, establishing a database, and collecting effective cases, thereby providing support for new drug research and development.
Further, the dynamic detection of the blood rheology parameter by the full-automatic blood viscosity monitor evaluates the blood viscosity change condition and provides a basis for formulating a multi-drug treatment scheme, comprising:
Detecting a patient blood sample with an automatic blood viscosity analyzer to collect data and evaluate blood flow; analyzing the blood rheology parameters, comparing to determine the difference between samples and setting a standard range of abnormal values; if the blood viscosity is detected to deviate from the standard range, predicting possible clinical diseases causing abnormality by using a support vector machine algorithm; accessing a medical drug database to acquire information of targeted therapeutic drugs according to the disease prediction result; if the predicted result shows a plurality of treatment options, performing utility and interaction analysis among various medicines by adopting a decision tree algorithm; analyzing the distinguished compatibility options of various medicines by using an algorithm, and calculating the influence of the compatibility options on the blood viscosity to obtain an optimal combination; the optimal medicine combination is compared and analyzed with the historical case data, and a final personalized treatment scheme is determined in an auxiliary way through an artificial intelligence system; inputting the screened treatment scheme into a drug management system, and automatically generating a personalized drug use plan and a corresponding early warning mechanism by the system; after the drug management protocol is implemented, the blood viscosity of the patient is monitored again using an automatic blood viscosity analyzer to verify the persistence and stability of the therapeutic effect.
Further, the construction of a large sample database containing patients with different ages, sexes and basic diseases, collecting clinical data such as treatment response and adverse reaction in the multi-drug treatment process, analyzing the characteristics and the ending of the patients by using a deep learning technology, and establishing an accurate prognosis prediction model, wherein the construction comprises the following steps:
The age, sex and existing basic diseases of the patient are recorded through the construction of a professional database and are used as important basis for primary screening of the patient; the collected clinical data is processed by using a data cleaning technology, and the key is to extract the information of the therapeutic drugs of the patient, the reactions to the drugs, adverse reactions possibly generated and the like; classifying the cleaned data, and creating a plurality of sub-databases to analyze the response caused by different drug treatments based on the received treatment regimen; through text mining technology, key descriptions describing treatment effects and detailed information about complications are deeply mined and arranged in clinical data; screening out characteristic factors closely related to the treatment effect and having large influence by utilizing a characteristic selection algorithm and combining with deep analysis of the characteristics of a patient; inputting the screened patient characteristics and treatment effect data into a deep learning model, performing first training of the model based on the characteristics and treatment effect data, and predicting prognosis of the patient; in order to ensure the stability and accuracy of the prognosis prediction model, a cross-validation method is used for evaluation, which can ensure that the model meets the statistical requirements of research; after the model evaluation accords with the expectation, further optimizing and correcting the model, wherein an incremental learning technology is adopted so as to continuously introduce new clinical data and improve the prediction precision of the model; the optimized prognosis prediction model is deployed into a clinical decision support system, so that the personalized treatment plan for the individual patient can be assisted by accurate prognosis information provided by the model.
Furthermore, the design of the intelligent multi-drug compatibility system predicts the curative effects and risks of different medication modes by using a machine learning algorithm according to factors such as individual parameters of patients, severity of illness and the like, and provides decision support of a multi-drug treatment scheme, which comprises the following steps:
collecting individual parameters such as medical history, vital signs and the like of a patient, and integrating the information to ensure the accuracy of each data; analyzing the historical medical data by using a convolutional neural network algorithm, including treatment effects for similar cases; based on these data, a primary drug compatibility model is constructed, an initial weight score is defined for each drug based on the individual parameters of the patient and the relevant medical literature and study, this score reflects the possible effectiveness and risk level of the drug, and by further adjusting these initial weights, the selection and compatibility of the drug can be optimized until a combination appropriate for the particular patient is found; the risk assessment is carried out according to the severity of the current illness state of the patient, so that whether the patient needs to take medicines urgently or not is judged, and possible side effects are prevented in advance; a convolutional neural network algorithm is applied to predict and analyze potential curative effects under different medication modes, and a plurality of treatment schemes are proposed; the long-term and short-term risks of these regimens are assessed, and the depth analysis is compared to identify the risks that the patient may be at; to ensure therapeutic safety, drug compatibility in these regimens is checked using a drug interaction database to avoid potential drug adverse reactions; continuously analyzing the change of physical signs in the treatment process according to the feedback of the real-time medical monitoring data, and adjusting the drug scheme and the dosage according to the change; the expert system is used as a decision support tool, the expert knowledge is combined with the scheme generated by the algorithm, and a custom-made multi-drug treatment scheme is provided; a feedback mechanism is designed to continuously track the treatment effect of a patient, and the convolutional neural network algorithm is optimized through the actual result, so that the accuracy and the safety of a treatment scheme are improved.
Further, the use of wearable sensors to monitor patient vital sign changes during multi-drug therapy in real time, evaluate safety of different dosing schemes, and adjust medication plans as needed in cooperation with the above blood viscosity monitoring results, including:
The method comprises the steps of monitoring vital signs of a patient by adopting a wearable sensor technology, and collecting real-time data, wherein the data is processed by an intelligent data analysis system to obtain detailed physiological parameters and viscosity monitoring results; the acquired data are transmitted to a data storage and management module in real time, and the module is responsible for recording health history information of a patient and taking the information as an auxiliary basis of a clinical decision support system; the personalized medical strategy module analyzes the monitoring data according to the specific condition of the patient and judges whether the current administration scheme needs to be adjusted so as to ensure the optimal treatment effect; when the monitoring data show that the patient has potential inadaptation symptoms or risks, the risk early warning system can timely give an alarm to doctors or patients; the safety evaluation module continuously evaluates the real-time vital sign data of the patient to determine whether adverse reactions caused by the medicines exist; if the evaluation result shows the potential risk, starting a medication plan adjustment module to reformulate a treatment scheme, and evaluating the interaction among different medicines through a multi-medicine treatment analysis module to optimize a treatment strategy; the treatment effect monitoring module continuously tracks the new treatment scheme to evaluate the curative effect of the scheme and ensure that the treatment target is achieved; if the curative effect monitoring module feeds back that the curative effect is poor, the medication plan adjusting module is started again, and the medication plan adjusting module is continuously optimized until the treatment scheme which is most suitable for the patient is found.
Further, the construction of the multi-drug interaction knowledge graph, the collection of drug-drug, drug-food, drug-disease interaction relationship data, for providing a reference for the compatibility selection of multi-drug combination, comprises:
Analyzing the medical literature and the drug instruction book by using a text mining technology, and extracting the attribute and action mechanism information of the drug from the medical literature and the drug instruction book; determining the possible effect of a specific food ingredient on the efficacy of the medicament by using information in a food ingredient database in combination with relevant knowledge of the nutrition; deep analysis is carried out on the relevant data of the diseases so as to identify the correlation between the action mechanism of the medicine and the disease characteristics; by identifying these correlations, it is possible to learn which drugs are at risk of treatment or side effects for a particular disease; the metabolic process of the drug in the body, particularly the interaction of the metabolite with other drugs or nutrients, can be better understood by the drug metabolic analysis tool; by analyzing the medical guidelines and the treatment manual, the advice of the combined use of a plurality of medicines can be provided for medical professionals, thereby helping to make the most reasonable combined treatment scheme; by tracking the drug concentration monitoring data, the influence of the combined drug on the drug plasma concentration can be observed in real time, and the fluctuation trend of the drug concentration under various conditions is analyzed; in analyzing the side effect data, disease characteristics will be considered to determine the risk of side effects that different patient populations may encounter when using a particular drug; based on the data collected above and previous research results, a model of efficacy influence assessment is constructed, which aims at predicting interactions when different drugs are combined and assessing how they influence the therapeutic effect; integrating the data and the evaluation results obtained in all the steps, and deploying a set of dynamic knowledge map updating mechanism by using a pattern recognition and machine learning algorithm; the mechanism can automatically incorporate new research results and real-time data, thereby continuously perfecting the knowledge graph and providing scientific decision support for medical workers about drug combination.
Further, the collecting compliance and lifestyle information of patients during multi-drug treatment process, establishing a model to evaluate the correlation of these factors with drug efficacy and adverse reaction, providing health management advice, including:
Carrying out detailed acquisition of the drug treatment information of the patient, recording the drug name, the dosage, the drug taking frequency and the specific taking time period, and establishing a basic database for subsequent analysis; collecting patient compliance data relating to whether the patient took the medication on time according to the order, monitoring and marking those missed or misadministered events, which provides direct evidence for assessing the actual performance of medication; developing a comprehensive investigation of the patient's lifestyle, including evaluating and recording diet, exercise, sleep, and stress-coping patterns, forming a lifestyle dataset related to health conditions; combining basic health information of patients, such as disease history and family history, constructing a comprehensive database, wherein the comprehensive database comprehensively records the overall health condition of individuals and provides necessary background information for further analysis; deep analysis of medication compliance and relationships with patient health using data mining techniques is aimed at identifying key factors that may affect the efficacy of medication; the statistical means is used for researching the relation between life style and medicine curative effect and adverse reaction, and potential relation between life habit and medicine safety is mined; based on the analysis, a prediction model is established, the model can evaluate the possibility of influence of different living modes on the drug reaction of a patient, and then the calibration and verification operation is carried out on the model, so that the accuracy of the model is ensured; providing personalized health management advice for the patient according to the result of the model prediction, helping the patient to make adjustment on the use of the medicine and encouraging the patient to improve life habits; dynamic monitoring of patient medication compliance and lifestyle changes is achieved, and these data are reviewed periodically to continuously optimize the predictive accuracy of the model and adjust health management measures accordingly.
Further, the intelligent voice assistant for reminding and monitoring the designed medication carries out illness state tracking, consultation service and the like, helps patients to correctly use various medicines, and comprises the following steps:
Collecting health and drug usage records of the patient through a disease tracking system for monitoring and analyzing the progress of the patient and the rules of drug usage; using the health data stored in the patient medical profile to compare the drug characteristics of the drug information database, the system automatically checks potential interactions of the current drug to prevent unsafe drug combinations; according to the analysis result of the drug interaction, calculating safe and effective drug dosage by utilizing a compound pharmacological algorithm, and adjusting a personalized medication reminding plan for a patient to ensure medication time and dosage accuracy; the intelligent voice interaction function enables patients to verbally inquire about the medication records and medical orders of the patients through a voice recognition technology, and quick self-management is achieved; when a patient has problems on medicines or diseases, the patient is automatically connected to related medical professionals through the consultation service interface to provide immediate professional consultation services; extracting proper information from a medication instruction knowledge base, educating a patient about a correct use method of a medicament, and enhancing the ability of the patient to follow medical advice; periodically generating disease course reports containing disease progress information of the patient by using a disease course management statistical function, and adjusting a treatment plan of the patient according to the reports so as to help doctors evaluate treatment effects; the feedback and health data of the patient are synchronized in real time, so that all relevant medical information is kept up to date, and medical staff can effectively monitor and intervene in time; and continuously monitoring the system performance, evaluating potential risks or existing problems, and periodically optimizing the system functions according to the collected data and user feedback to continuously improve the service accuracy and user experience.
Further, the step of detecting the hemostatic parameters of the patient again, evaluating the blood viscosity change condition before and after the multi-drug treatment, establishing a database to collect effective cases, and providing support for new drug development, comprises the following steps:
in order to accurately monitor the change of the hemostatic parameters of a patient, an automatic blood analysis system is deployed to track the hemostatic parameters of the patient, so that detailed data before and after treatment is collected; using the collected hemostatic parameters as a research basis, performing skin-level analysis on blood states before and after multi-drug treatment by using hemodynamic simulation software so as to grasp specific changes of blood viscosity; in order to improve the data management efficiency, an effective database architecture is constructed, hemostasis parameters and blood viscosity change data are integrated, and the integrity and consistency of the key information are ensured; the method has the advantages that through the inspiration of the existing data, the optimization is performed by applying a data mining algorithm, and real-time and effective decision support is provided for the development of new medicines in progress; defining and perfecting efficacy evaluation criteria, and analyzing and verifying the actual efficacy of the treatment regimen recorded in the database; further analyzing the correlation between the hemostatic parameters and the blood viscosity before and after treatment by using a defined curative effect evaluation standard and a statistical method to identify a significant difference; aiming at the data trend obtained by analysis, the existing multi-drug treatment scheme is subjected to necessary adjustment so as to improve the accuracy and pertinence of the treatment scheme; realizing a dynamic updating mechanism, ensuring that the latest research findings, treatment results and related parameters can be reflected in a database in real time; a periodic reporting regime is formulated to automatically generate a comprehensive treatment effect report containing treatment regimen adjustment recommendations and patient hemostatic parameter changes.
The technical scheme provided by the embodiment of the invention can have the following beneficial effects:
The invention discloses a multi-drug treatment response prediction method based on blood viscosity change. By introducing the full-automatic blood viscosity monitor, the problem of static monitoring of traditional blood parameters is solved, dynamic detection of blood rheology parameters is realized, and a more accurate basis is provided for formulating a multi-drug treatment scheme. And secondly, by constructing a large sample database and applying a deep learning technology, the problem that the existing curative effect prediction model is insufficient in comprehensive consideration of individual differences and multi-drug correlations is solved, and a more accurate prognosis prediction model is established. Most importantly, the intelligent multi-drug compatibility system is designed, the problem of insufficient intelligence of the traditional compatibility system is solved by combining a machine learning algorithm, a personalized treatment scheme is provided for a patient, and meanwhile, the safety and patient compliance problems in the treatment process are solved by monitoring vital signs and intelligent voice assistants in real time. In the whole, the invention effectively solves the problems of individuation, dynamics, intellectualization and the like of multi-drug treatment, and provides a comprehensive solution for improving the treatment effect and the life quality of patients.
Drawings
FIG. 1 is a flow chart of a method for predicting a response to multi-drug therapy based on blood viscosity changes according to the present invention.
FIG. 2 is a schematic diagram of a method for predicting a response to multi-drug therapy based on blood viscosity changes according to the present invention.
FIG. 3 is a schematic diagram of a method for predicting a response to multi-drug therapy based on blood viscosity changes according to the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings for a clear and intuitive understanding to those skilled in the art.
In order to make the technical solutions in the present specification better understood by those skilled in the art, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only some embodiments of the present specification, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.
The multi-drug treatment response prediction method based on blood viscosity change in this embodiment specifically may include:
step S101, dynamically detecting blood rheology parameters through a full-automatic blood viscosity monitor, evaluating blood viscosity change conditions, and providing a basis for formulating a multi-drug treatment scheme.
An automated blood viscosity analyzer is used to detect a patient's blood sample to collect data and evaluate blood flow. The haemorheology parameters are analysed and a comparison is made to determine the differences between the samples and to set a standard range of outliers. If blood viscosity deviation from the standard range is detected, a support vector machine algorithm is used to predict possible clinical diseases causing abnormalities. According to the disease prediction result, the medical drug database is accessed to acquire the information of the targeted therapeutic drug. If the predicted outcome shows multiple treatment options, a decision tree algorithm is used to analyze the effects and interactions between the various drugs. The different drug compatibility options are analyzed by an algorithm, and the influence of the drug compatibility options on the blood viscosity is calculated to obtain the optimal combination. The optimal drug combination is compared and analyzed with the historical case data, and the final personalized treatment scheme is determined in an auxiliary way through an artificial intelligence system. The screened treatment schemes are input into a drug management system, and the system automatically generates personalized drug use plans and corresponding early warning mechanisms. After the drug management protocol is implemented, the blood viscosity of the patient is monitored again using an automatic blood viscosity analyzer to verify the persistence and stability of the therapeutic effect.
For example, in medical practice, it is critical that the rheological parameters of the patient's blood be accurately measured by automated equipment. For example, using an automatic blood viscosity analyzer, a blood sample can be measured with a viscosity of 3.5cP (milliPascals) and a normal range of 2.5-3.5cP. When the blood viscosity measurement of a patient is 4.2cP, i.e., outside the normal range, a Support Vector Machine (SVM) is used to compare the data points to pathological data in the training set, and predict possible disease factors, such as hypertension or hyperlipidemia. On this basis, the artificial intelligence system may link to a medical drug database, gathering drug information matching the predicted results. For example, if hypertension is predicted, drug information such as ACE inhibitors, diuretics, etc. may be extracted. A decision tree algorithm is then employed to analyze the utility and interaction of the different drugs. Taking diuretics and ACE inhibitors as examples, a decision tree is constructed, the nodes of which represent drug selection, while the nodes represent parameters affecting decision making, such as drug effect, side effects and the like. The algorithm will calculate the final score for each path based on the utility function and the patient's specific conditions (age, weight, complications, etc.), making the best choice. Next, the selected drug combinations are analyzed in depth. Quantitative pharmacological models, such as Hill's equation, are used to estimate the effect of drug compatibility on blood viscosity. For example, for ACE inhibitors, their effect on blood viscosity may be described by Hill equation V=Vmax. N/(Kdn+. n), where V is the effect of the drug, vmax is the maximum potential effect of the drug, is the drug concentration, kd is the half maximum effect drug concentration, and n is the slope of the drug response curve. And adjusting parameters in Hill equation according to the drug response data of the subjects to predict the curative effect of the optimal drug combination. Thereafter, machine learning techniques such as k-nearest neighbor (k-NN) algorithms can be used to refine the treatment plan by comparison with the historical case data. The system will analyze the treatment response in the historical cases similar to the predicted one, optimizing the personalized treatment regimen. The final optimized treatment regimen will be incorporated into a medication management system that will automatically generate a medication usage plan for the patient and set up an early warning mechanism, such as setting a time point for periodic blood pressure or blood viscosity detection. Such monitoring may rely on patient biomarkers, such as blood viscosity above 4.0cP for three consecutive days, the system will automatically send a reminder to visit the physician. Following administration of the drug management regimen, the persistence and stability of the therapeutic effect should be tracked. The real-time monitoring data is collected by an automatic blood viscosity analyzer and the treatment effect is determined by algorithm analysis. The blood viscosity data of the patient reflected by the results are compared with the data before and after treatment, for example, the blood viscosity value of the patient after treatment is stabilized at about 3.0cP, and the effectiveness and the safety of the treatment scheme are proved.
Step S102, constructing a large sample database containing patients with different ages, sexes and basic diseases, collecting clinical data such as treatment response and adverse reaction in the multi-drug treatment process, analyzing the characteristics and the ending of the patients by using a deep learning technology, and establishing an accurate prognosis prediction model.
Through the construction of a professional database, the age, sex and existing basic diseases of the patient are recorded and used as important basis for primary screening of the patient. The collected clinical data is processed by using a data cleaning technology, and the key is to extract information such as therapeutic drugs of patients, reactions to the drugs, adverse reactions possibly generated and the like. Classifying the cleaned data, based on the received treatment regimen, multiple sub-databases may be created to analyze the response elicited by different drug treatments. By text mining, key statements describing the effect of treatment, and detailed information about complications, will be mined and organized deep in clinical data. And (3) screening out characteristic factors closely related to the treatment effect and having large influence by utilizing a characteristic selection algorithm and combining with deep analysis of the characteristics of the patient. These screened patient characteristics and treatment effect data are input into a deep learning model, which is used as a basis for first training of the model and predicting prognosis of the patient. In order to ensure the stability and accuracy of the prognosis prediction model, the assessment is performed using a cross-validation method, which can ensure that the model meets the statistical requirements of the study. After the model evaluation meets the expectations, further optimization and adjustment are carried out, wherein an incremental learning technology is adopted, so that the prediction accuracy of the model is improved while new clinical data are continuously introduced. The optimized prognosis prediction model is deployed into a clinical decision support system, so that the personalized treatment plan for the individual patient can be assisted by accurate prognosis information provided by the model.
Illustratively, in constructing a patient database, structured queries are performed in the SQL language, and tables are created to store information about the patient's age, sex, and medical history, for example using "CREATETABLEPatients (PATIENTIDINT, AGEINT, GENDERCHAR (1), DISEASEHIST ORYVARCHAR (255)); ". Subsequently, the collected clinical data is subjected to data cleansing, and the deletion value and the abnormal value are removed by using pandas library of Python, for example, deletion data is deleted by using a' df. The effective information after data cleaning, such as the response of a patient to a certain drug a, is extracted and classified into a sub-database, such as ' DATAFRAME = ' DrugA ' ]. Next, a text mining tool in NLP technology, such as NLTK library or spaCy library, is applied to analyze the free text in the clinical data, extract key words describing the therapeutic effect, such as "significant effect", and encode these information into a data format that can be used for subsequent calculations. In the feature selection link, a feature importance assessment method in a machine learning algorithm such as logistic regression or decision tree is adopted, for example, the 'SelectFromModel' class of sklearn is used for matching with the 'LogisticRegression', so that the importance of the features is scored and screened. The filtered feature data is then loaded into a deep learning framework such as TensorFlow or PyTorch to construct a neural network model. For example, a multi-layer perceptron model with three hidden layers is designed, a reasonable number of neurons, such as 64-128-64, is set, and nonlinear transformation is performed by using a ReLU activation function. First training is performed with the prepared data set and prediction accuracy is measured using Mean Square Error (MSE) as a loss function, with model weights being continually optimized by a back propagation algorithm. And then, introducing a cross-validation technology validation model, namely randomly dividing the data set into K subsets possibly by using K-fold cross-validation, sequentially using one subset as a validation set and the rest as a training set to evaluate the generalization capability of the model. As the evaluation results accumulate, the overall performance metrics of the model, such as accuracy, recall, and F1 score, can be calculated, ensuring that they reach acceptable levels. To achieve continuous optimization of the model, new data may be introduced gradually using online learning or batch learning methods, model parameters may be adjusted, for example by adjusting super-parameters such as learning rate when new patient data is added, adapting the model to the latest data features, for example "optimizer =tf. Train. Adam optimizer (learning_rate=0.001). Minimize (loss)", where Adam optimizer is used to adjust the learning rate. Finally, the trained and optimized deep learning model is embedded into a clinical decision support system, real-time patient information is received through an interface of the system, and prognosis prediction is carried out by using the model. The system may assist the physician in developing a more personalized and accurate treatment plan based on the probabilistic output of the model predictions, such as predicting a patient's 5 year survival rate of 75% under drug a treatment.
Step S103, designing an intelligent multi-drug compatibility system, predicting the curative effects and risks of different medication modes by using a machine learning algorithm according to individual parameters of patients, severity of illness and other factors, and providing decision support of a multi-drug treatment scheme.
Individual parameters such as medical history, vital signs and the like of the patient are collected, and the information is integrated, so that the accuracy of each data is ensured. The historical medical data is analyzed using a convolutional neural network algorithm, including treatment effects for similar cases. Based on these data, a primary drug compatibility model is constructed, and based on individual parameters of the patient and related medical literature and studies, an initial weight score is defined for each drug, which score reflects the potential effectiveness and risk level of the drug, and by further adjusting these initial weights, the selection and compatibility of the drug can be optimized until a combination appropriate for the particular patient is found. The risk assessment is carried out according to the severity of the current illness state of the patient, so that whether the patient needs to take medicines urgently or not is judged, and possible side effects are prevented in advance. And a convolutional neural network algorithm is applied to predict and analyze potential curative effects under different medication modes, and various treatment schemes are proposed. The long-term and short-term risks of these regimens are assessed, and the depth analysis is compared to identify the risks that the patient may be at. To ensure therapeutic safety, drug compatibility in these regimens is checked using a drug interaction database to avoid potential drug adverse effects. According to the feedback of the real-time medical monitoring data, the change of physical signs in the treatment process is continuously analyzed, and the drug scheme and the dosage are adjusted according to the change. The expert system is used as a decision support tool to combine the expert knowledge with the algorithm generated scheme to provide a custom-made multi-drug treatment scheme. A feedback mechanism is designed to continuously track the treatment effect of a patient, and the convolutional neural network algorithm is optimized through the actual result, so that the accuracy and the safety of a treatment scheme are improved.
Illustratively, in constructing a personalized treatment regimen, a detailed medical history is first collected for the patient, e.g., the historical blood pressure record for patient J may be represented by a list of values: [120,118,122,119,124] mmHg. At the same time, vital signs of J were monitored and heart rate was recorded as [78,82,81,77,80] bpm. Deep learning algorithms such as Convolutional Neural Networks (CNNs) are used to train the collected data and historical medical data of similar patients to extract features for predicting the effectiveness of the treatment. In quantifying risk assessment of J disease, a cardiac risk scoring formula may be used: (r=w\ timesH), where W is body weight, normalized to a standard body weight of 1, H is systolic blood pressure, normalized to normal systolic blood pressure of 1. And predicting the curative effect of different treatment schemes by using a deep learning algorithm, and scoring the curative effect under different drug combinations, wherein the higher the score is, the better the curative effect is predicted. For example, a combination treatment of drugs A and B may result in an effect score of 0.9 and a combination of drugs A and C of 0.85, which score may be generated by a model trained from historical data. Monte Carlo simulation is further used to assess long-term and short-term risk of treatment regimens. Considering that long-term administration of drug A may increase the probability of kidney function impairment, 10000 patients may be simulated for drug A treatment, and simulation results show that 2% of patients have kidney function impairment. In terms of drug compatibility, a drug interaction database may be used to query the index of interaction between different drugs, such as the CYP450 enzyme impact index, and if the index is greater than a certain threshold, such as 1.5, then significant interactions are considered likely to exist. For the treatment regimen of J, if the CYP450 enzyme impact index of drugs a and B is 1.2 and a and C is 1.6, then the combination of drugs a and B is selected to reduce the risk of interaction. As treatment progresses, the real-time medical monitoring data reflects an increase in the heart rate of J to 88bpm, and if the heart rate continues to be higher than 85bpm, the dosage of drug a is reduced to 80% of the original dosage according to a preset decision tree model. The compatibility of medicines can be further adjusted by combining the opinion of clinical specialists and model prediction through the decision support of an expert system. If an expert system recommends that in high risk cardiac patients the use of drug D can reduce the probability of occurrence of a cardiac event, even if it is not directly related to the treatment effect score, this information should be taken into account in the treatment regimen. In the treatment process, the designed feedback mechanism can finely adjust the deep learning algorithm according to the treatment effect. If a deviation between the drug compatibility effect score and the true response is observed during treatment J, the algorithm can be retrained based on the response data of J to optimize the predictive model.
Step S104, the wearable sensor is adopted to monitor the vital sign change of the patient in the multi-drug treatment process in real time, the safety of different drug administration schemes is evaluated, and the drug administration plan is adjusted as required by matching with the blood viscosity monitoring result.
The wearable sensor technology is adopted to monitor vital signs of patients and collect real-time data, and the data is processed by the intelligent data analysis system to obtain detailed physiological parameters and viscosity monitoring results. The acquired data are transmitted to a data storage and management module in real time, and the module is responsible for recording health history information of a patient and taking the information as an auxiliary basis of a clinical decision support system. The personalized medical strategy module analyzes the monitoring data according to the specific condition of the patient and judges whether the current administration scheme needs to be adjusted so as to ensure the optimal treatment effect. The risk early warning system can alert a doctor or patient in time when the monitored data indicate that a potentially inadaptive symptom or risk is present to the patient. The safety assessment module continuously assesses patient real-time vital sign data to determine whether adverse reactions due to the drug are present. If the evaluation result shows the potential risk, a medication plan adjustment module is started to reformulate a treatment plan, and interaction among different medicines is evaluated through a multi-medicine treatment analysis module so as to optimize a treatment strategy. The treatment effect monitoring module continuously tracks the new treatment plan to evaluate the efficacy of the plan and ensure that the treatment objective is achieved. If the curative effect monitoring module feeds back that the curative effect is poor, the medication plan adjusting module is started again, and the medication plan adjusting module is continuously optimized until the treatment scheme which is most suitable for the patient is found.
Illustratively, in monitoring vital signs, the patient heart rate data recorded by the wearable sensor is a series of time series values, and the average heart rate per minute is calculated using a sliding window algorithm. The heart rate value per minute is obtained by dividing 60 by summing the number of heartbeats over 60 seconds, and the calculation result is 72 times per minute. The intelligent data analysis system will then compare this result to the heart rate baseline of the patient, mark the data as abnormal and trigger a risk warning if the deviation exceeds a set threshold, such as 5 times/min. If the patient needs to adjust the dosage regimen, the personalized medicine strategy module can apply a medicine dosage adjustment algorithm, such as a overcoming rate equation (Cockcroft-Gaultequation), to calculate the renal function clearance according to the information of the age, sex, weight and the like of the patient and the pharmacokinetics characteristics of the medicine, so as to further calculate the personalized medicine dosage. The module may recommend that the dose of a certain drug be reduced from 5mg to 4 mg once a day. The safety assessment module continuously monitors vital sign data, such as matching heart rate to known data in a drug side effect database, applies a machine learning algorithm such as a Support Vector Machine (SVM) to determine if there is evidence of drug adverse effects, and prompts adjustment of the medication plan if the predicted probability exceeds an acceptable threshold, such as 20%. The medication plan adjustment module, upon receiving the adjustment signal, evaluates the current prescription using a decision tree algorithm and combines the medication interaction data provided by the multi-medication therapy analysis module, for example, using a medication interaction checker (DrugInteractionChecker), to ensure that the adjusted regimen avoids potential medication interactions. This may involve reducing interactivity by time-staggered administration of two potentially interacting drugs, e.g., adjusting the administration time of drug a from morning to evening. To continuously evaluate the efficacy of the new treatment regimen, the treatment effect monitoring module periodically examines the key physiological parameters and uses logistic regression analysis to predict the treatment response to determine whether the intended treatment objective has been achieved. For example, for a hypertensive patient, based on an average systolic blood pressure reading over three consecutive days, if the drop is less than 10 mmHg, the module will signal poor efficacy. Finally, if the efficacy monitoring module prompts that further optimization of treatment is required, the medication plan adjustment module will update the efficacy probability distribution of the patient using the bayesian network and consider new treatment options. For example, if the current regimen is to use the drug X, it may be better to infer the drug Y from the new data obtained, by calculating the posterior probabilities of the two regimens, if the probability of the effect of the drug Y is significantly higher, it is recommended to adjust the treatment regimen to the drug Y.
Step S105, constructing a multi-drug interaction knowledge graph, collecting drug-drug, drug-food and drug-disease interaction relation data, and providing reference for compatibility selection of multi-drug combination.
The medical literature and the drug specifications are analyzed by using text mining technology, and the attribute and action mechanism information of the drug are extracted from the analysis. The information in the food ingredients database, in combination with relevant knowledge of the nutrition, is used to determine the effect that a particular food ingredient may have on the efficacy of a drug. And carrying out deep analysis on the disease-related data so as to identify the correlation between the action mechanism of the medicine and the disease characteristics. By identifying these correlations, it is possible to learn which drugs are at risk of treatment or side effects for a particular disease. The metabolic processes of drugs in vivo, particularly the interactions of metabolites with other drugs or nutrients, can be better understood by drug metabolic analysis tools. By analyzing medical guidelines and handbooks, the medical professionals can be provided with suggestions for the combined use of various medicines, and the most reasonable combined treatment scheme can be prepared. By tracking the drug concentration monitoring data, the influence of the combined drug on the drug plasma concentration can be observed in real time, and the fluctuation trend of the drug concentration under various conditions can be analyzed. In analyzing the side effect data, disease characteristics will be considered to determine the risk of side effects that different patient populations may encounter when using a particular drug. Based on the data collected above and previous research results, a model of efficacy impact assessment was constructed, which was aimed at predicting interactions when different drugs were combined and assessing how they affected the therapeutic effect. Integrating the data and the evaluation results obtained in all the steps, and deploying a set of dynamic knowledge map updating mechanism by using a pattern recognition and machine learning algorithm. The mechanism can automatically incorporate new research results and real-time data, thereby continuously perfecting the knowledge graph and providing scientific decision support for medical workers about drug combination.
For example, in analyzing medical literature and drug specifications using text mining techniques, natural Language Processing (NLP) algorithms can be utilized to identify key information such as drug name, dose, and mechanism of action, as by solid state identification (NER). For example, information about "aspirin", "antiplatelet", and "reduce myocardial infarction" can be extracted from an article of aspirin using a Conditional Random Field (CRF) based model. Next, in combination with food ingredient information in a database, such as the food ingredient database of the United States Department of Agriculture (USDA), the inhibition of the drug CYP3A4 enzyme by a specific ingredient, such as furocoumarin in grapefruit juice, can be determined, which can affect the metabolic rate of the drug. Through the synergistic analysis of the nutritional ingredients and the drug effects, the probability of the interaction of the food ingredients with the specific drugs can be predicted by using a logistic regression model, and the influence degree can be further estimated. In correlation analysis of disease characteristics and drug action mechanisms, genomic correlation analysis (GWAS) can be used to identify the therapeutic efficacy of drugs in specific disease states based on gene expression profile data and disease biomarkers. For example, for statin, an inhibitor of the key enzyme HMG-CoA reductase in cholesterol synthesis, it was found by GWAS analysis that patients with certain genetic variations may respond better to statin. In understanding the metabolic processes of drugs in depth, pharmacokinetic (PK) models, such as nonlinear mixed effect models (NONMEM), can be used to model and predict the absorption, distribution, metabolism, and excretion (ADME) processes of drugs in different individuals. For example, using this model, it can be predicted that drug A's half-life increases from 8 hours to 10 hours when drug A and drug B are present simultaneously in the same patient. In analyzing drug side effect data, the severity of side effect of a patient on a drug is predicted and classified using a Back Propagation Neural Network (BPNN) model using information in the adverse event reporting system (FAERS) database. For example, it may be found that in elderly diabetics who are treated with certain blood pressure medications, a hypoglycemic response occurs with a 25% probability. The construction of the drug effect influence assessment model can combine various models and data, including the food-drug, disease-drug, drug metabolism model and drug side effect model, and the treatment effect and side effect risk when different drugs are combined can be predicted by using an integrated learning method such as a random forest model. When the medicine A and the medicine B are combined, the model calculates that the risk of side effects of the combination is improved by 15 percent, and the treatment effect is improved by 10 percent. Finally, the construction and updating of the knowledge graph can be realized by using a graph database Neo4j, and the connection relation between medicines is optimized by combining the analysis results and the Cypher query language built in the graph database. For example, the drug interactions found by the new study are automatically added when updating the profile, and the therapeutic effect improvement index for the combination of drug A and drug B is calculated, and based thereon, decision support is provided to the medical professional. The knowledge graph synthesizes various models, database information, real-time clinical data and the like to form a self-updating and dynamic decision-making auxiliary system so as to improve the safety and effectiveness of the combined medication of the medicines. Through the series of comprehensive data analysis and automation models, a more precise and self-learning medical information technology framework can be constructed.
Step S106, collecting compliance and life style information of patients in the multi-drug treatment process, establishing a model to evaluate the correlation of the factors with the drug efficacy and adverse reaction, and providing health management advice.
Carrying out detailed acquisition of the drug treatment information of the patient, recording the drug name, the dosage, the drug taking frequency and the specific taking time period, and establishing a basic database for subsequent analysis; collecting patient compliance data relating to whether the patient took the medication on time according to the order, monitoring and marking those missed or misadministered events, which provides direct evidence for assessing the actual performance of medication; developing a comprehensive investigation of the patient's lifestyle, including evaluating and recording diet, exercise, sleep, and stress-coping patterns, forming a lifestyle dataset related to health conditions; combining basic health information of patients, such as disease history and family history, constructing a comprehensive database, wherein the comprehensive database comprehensively records the overall health condition of individuals and provides necessary background information for further analysis; deep analysis of medication compliance and relationships with patient health using data mining techniques is aimed at identifying key factors that may affect the efficacy of medication; the statistical means is used for researching the relation between life style and medicine curative effect and adverse reaction, and potential relation between life habit and medicine safety is mined; based on the analysis, a prediction model is established, the model can evaluate the possibility of influence of different living modes on the drug reaction of a patient, and then the calibration and verification operation is carried out on the model, so that the accuracy of the model is ensured; providing personalized health management advice for the patient according to the result of the model prediction, helping the patient to make adjustment on the use of the medicine and encouraging the patient to improve life habits; dynamic monitoring of patient medication compliance and lifestyle changes is achieved, and these data are reviewed periodically to continuously optimize the predictive accuracy of the model and adjust health management measures accordingly.
For example, in building a base database, an electronic health record system may be used to automatically capture and record patient medication information, including the use of certain antihypertensive agents, such as lovastatin 20mg, once a day in the evening. Data on compliance may then be collected using intelligent kits that record the time the medication was removed and automatically mark missed medication events as compared to the scheduled administration time, such as when the patient had a day that was not taken 8 pm in five consecutive days. Further, by monitoring the lifestyle of the patient through his self-report and wearable device, it can be known that the average number of steps per week of the patient is 6000 steps, adhering to sleep and high-fat diet for 7 hours per night. In combination with disease history, such as 10 years of history of hypertension and family history of heart disease, a comprehensive database containing comprehensive health information of individuals is obtained. The relationship between compliance with medication therapy and patient health indicators can be analyzed by data mining tools and algorithms, such as logistic regression analysis. For example, the model may show that every 1% decrease in medication compliance, the systolic blood pressure flattens up to 0.5mmHg. By using statistical correlation analysis, the relationship between life style and drug efficacy can be studied, for example, the increase of exercise amount and drug hypotensive effect can be found to have positive correlation, the correlation coefficient is 0.3. Next, a machine learning model, such as random forest, is established according to the collected data and analysis result to predict the influence of different life styles on drug response. Through cross verification, the performance of the model on different data sets is checked to adjust parameters, and the accuracy and generalization capability of the model are improved. For example, the model accuracy reaches 80%, and the F1 fraction is 0.75. According to this model, personalized health management advice is provided, for example, for patients with insufficient exercise amount, gradually increasing to 8000 steps per day, and changes in their drug efficacy are monitored. Finally, the compliance of the long-term drug treatment and the life style change of the patient are monitored in real time by utilizing a big data technology, and the prediction model is continuously optimized by a deep learning method. The model is recalibrated every quarter to accommodate new patient data and treatment response patterns. For example, adjusting the model weights increases the effect of drug compliance by 5% to reflect its significance in the therapeutic effect.
Step S107, an intelligent voice assistant for reminding and monitoring medication is designed to carry out illness state tracking, consultation service and the like so as to help patients to use various medications correctly.
Health and drug usage records of the patient are collected by a condition tracking system for monitoring and analyzing the patient's course of disease progression and the rules of drug usage. Using the health data stored in the patient medical profile, the system automatically checks the potential interactions of the current medication against the medication characteristics of the medication information database to prevent unsafe medication combinations. According to the analysis result of the drug interaction, the safe and effective drug dosage is calculated by utilizing a compound pharmacological algorithm, and a personalized medication reminding plan is adjusted for the patient, so that the medication time and the accuracy of the dosage are ensured. The intelligent voice interaction function enables patients to verbally inquire about the medication records and medical orders of the patients through a voice recognition technology, and quick self-management is achieved. When a patient has problems with medicines or diseases, the patient is automatically connected to relevant medical professionals through the consultation service interface to provide immediate professional consultation services. Appropriate information is extracted from the medication instruction knowledge base to educate the patient about the correct method of use of the medication, enhancing their ability to follow medical advice. By using the disease course management statistical function, disease course reports containing disease progress information of patients are generated regularly, and treatment plans of the patients are adjusted according to the reports so as to help doctors evaluate treatment effects. And the feedback and health data of the patient are synchronized in real time, so that all relevant medical information is ensured to be kept up to date, and medical staff can effectively monitor and intervene in time. And continuously monitoring the system performance, evaluating potential risks or existing problems, and periodically optimizing the system functions according to the collected data and user feedback to continuously improve the service accuracy and user experience.
Illustratively, patient A has a weight of 60 kg, a height of 7 meters, and an age of 35 years. Patient a is currently taking a drug called "aspirin" at a dose of 100 mg per day. Doctor B has told patient a that this drug may cause several side effects, including gastrointestinal discomfort and bleeding. The following health data were collected, weighing 60 kg, height 1.7 m, age 35 years, drug name aspirin, drug dosage 100 mg daily. The potential interactions of the current medication are checked using the medication characteristics in the medication information database. The drug information database shows that aspirin may cause the following side effects, gastrointestinal discomfort including nausea, vomiting, gastralgia, constipation, etc., and bleeding including bleeding tendency, hematuria, etc. The potential interaction risk of the current medication is calculated to help the patient prevent possible side effects. Based on the calculations, patient a was advised to increase gastrointestinal tract protection when taking aspirin, such as daily use of proton pump inhibitors to reduce gastrointestinal discomfort. And calculating the safe and effective drug dosage according to the analysis result of drug interaction. If patient A needs to take 80 mg of aspirin daily, the therapeutic effect is achieved. Checking whether the current medication dose is safe and effective, and automatically adjusting the medication dose of the patient A according to the calculation result. Finally, according to the disease progress of the patient, generating a disease course report containing the disease progress information of the patient, and adjusting the treatment plan of the patient according to the report. If patient a does not have any serious side effects during the administration, doctor B may consider increasing the amount of patient a administered to achieve a better therapeutic effect.
Step S108, detecting hemostatic parameters of the patient again, evaluating blood viscosity change conditions before and after multi-drug treatment, establishing a database to collect effective cases, and providing support for new drug research and development.
In order to accurately monitor the change of the hemostatic parameters of the patient, an automatic blood analysis system is deployed to track the hemostatic parameters of the patient, so as to ensure that detailed data before and after treatment is collected. And (3) performing skin-level analysis on the blood state before and after multi-drug treatment by using the collected hemostatic parameters as a research basis through hemodynamic simulation software so as to grasp specific changes of blood viscosity. In order to improve the data management efficiency, an effective database architecture is constructed, hemostasis parameters and blood viscosity change data are integrated, and the integrity and consistency of the key information are guaranteed. The method is characterized in that the existing data is used for inspiring, a data mining algorithm is applied for optimization, and real-time and effective decision support is provided for the development of new medicines. Efficacy assessment criteria are defined and refined and the actual efficacy of the treatment regimen recorded in the database is analyzed and verified accordingly. The correlation between hemostatic parameters and blood viscosity before and after treatment is further analyzed using defined efficacy assessment criteria in combination with statistical methods to identify significant differences. Aiming at the data trend obtained by analysis, the existing multi-drug treatment scheme is subjected to necessary adjustment so as to improve the accuracy and pertinence of the treatment scheme. A dynamic update mechanism is implemented to ensure that the latest findings, treatment results, and related parameters can be reflected in the database in real time. A periodic reporting regime is formulated to automatically generate a comprehensive treatment effect report containing treatment regimen adjustment recommendations and patient hemostatic parameter changes.
Illustratively, in order to accurately monitor hemostatic parameters of a patient, an automated blood analysis system, such as a coagulation function monitor, e.g., DDAVP, used to treat the patient with an anticoagulant, is used to continuously track and record the patient's Clotting Time (CT) and Fibrinogen Time (FT), so that detailed data can be obtained by comparing the CT and FT values before and after treatment. For example, patient a had a pre-treatment CT of 120 seconds, FT of 600 seconds, and post-treatment CT decreased to 60 seconds, FT decreased to 400 seconds, demonstrating the effectiveness of the treatment. After these coagulation parameters are collected, a simulation analysis of the patient's blood status before and after treatment is performed by hemodynamic simulation software, such as Hemodyn. By inputting patient blood parameters and rheological properties, such as viscosity and density, software can demonstrate changes in blood flow conditions in computer simulations. By analyzing the blood flow rate patterns simulated by the software before and after treatment, specific changes in viscosity can be detected, such as a drop from 4.5cP to 2.8cP. To efficiently manage these data, a database architecture is constructed that integrates all coagulation parameters and blood viscosity data. The SQL statement is used for establishing association, such as associating various parameters with a 'JOIN' command, so that the integrity and consistency of data are ensured. For example, a "hemodynamics" table is created that contains fields for patient ID, treatment time point, CT, FT, and viscosity, etc., with data association maintained by patient ID. Based on these data, a data mining algorithm, such as a random forest classifier, is used for optimization. Based on this, a real-time decision support system was developed that was able to provide feedback to the drug development team based on real-time data. For example, the system may indicate that decreasing the dosage of an anticoagulant drug achieves the same therapeutic effect when the viscosity falls below a certain threshold. Defining new curative effect evaluation standard, for example, setting viscosity to be reduced by more than 5% to obtain obvious curative effect, and analyzing actual effect of therapeutic scheme recorded in database. By this method, the accuracy and applicability of the database collection data in the treatment strategy can be verified. Statistical methods, such as Pearson correlation analysis, are used to further study the relationship between hemostatic parameters and blood viscosity. The correlation coefficient of the hemostatic parameters of the patients and the blood viscosity is analyzed, and the correlation is found to be more than 0.85, which shows that the hemostatic parameters and the blood viscosity have a remarkable positive correlation. Existing treatment regimens are adjusted according to trends in the data analysis. For example, if an improvement in coagulation parameters is observed with a less pronounced decrease in viscosity, it is possible to reduce the anticoagulant drug dosage while increasing the blood circulation promoting drug. Finally, a dynamic update mechanism is implemented, for example by periodically extracting data from the laboratory instrument by writing an automated script, and updating new findings, treatment results, and parameter changes in real-time into the database.
The foregoing is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical scheme of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.
Claims (9)
1. A method for predicting a multi-drug therapeutic response based on blood viscosity changes, comprising the steps of:
Dynamically detecting blood rheology parameters through a full-automatic blood viscosity monitor, evaluating the blood viscosity change condition, and providing a basis for formulating a multi-drug treatment scheme;
Constructing a large sample database containing patients with different ages, sexes and basic diseases, collecting clinical data of treatment reactions and adverse reactions of the patients in the multi-drug treatment process, analyzing the characteristics and the ending of the patients by using a deep learning technology, and establishing an accurate prognosis prediction model;
Designing an intelligent multi-drug compatibility system, predicting the curative effects and risks of different medication modes by using a machine learning algorithm according to individual parameters of patients and the severity of illness, and providing a decision support which looks like a multi-drug treatment scheme;
the method comprises the steps of monitoring vital sign changes of a patient in a multi-drug treatment process in real time by adopting a wearable sensor, evaluating the safety of different drug administration schemes, and adjusting a drug administration plan as required by matching with the blood viscosity monitoring result;
Constructing a multi-drug interaction knowledge graph, collecting drug-drug, drug-food and drug-disease interaction relation data, and providing reference for compatibility selection of multi-drug combination;
Collecting compliance and life style information of patients in the multi-drug treatment process, establishing a model to evaluate the correlation of the factors with the drug efficacy and adverse reaction, and providing health management advice;
An intelligent voice assistant for reminding and monitoring medication is designed to carry out illness state tracking and consultation services so as to help patients to use various medications correctly;
And detecting hemostatic parameters of the patient again, evaluating blood viscosity change conditions before and after multi-drug treatment, establishing a database, and collecting effective cases, thereby providing support for new drug research and development.
2. The method of claim 1, wherein the dynamically detecting the blood rheology parameter by the fully automatic blood viscosity monitor, evaluating the blood viscosity change, provides a basis for formulating a multi-drug treatment regimen, comprising:
Detecting a patient blood sample with an automatic blood viscosity analyzer to collect data and evaluate blood flow; analyzing the blood rheology parameters, comparing to determine the difference between samples and setting a standard range of abnormal values; if the blood viscosity is detected to deviate from the standard range, predicting possible clinical diseases causing abnormality by using a support vector machine algorithm; accessing a medical drug database to acquire information of targeted therapeutic drugs according to the disease prediction result; if the predicted result shows a plurality of treatment options, performing utility and interaction analysis among various medicines by adopting a decision tree algorithm; analyzing the distinguished compatibility options of various medicines by using an algorithm, and calculating the influence of the compatibility options on the blood viscosity to obtain an optimal combination; the optimal medicine combination is compared and analyzed with the historical case data, and a final personalized treatment scheme is determined in an auxiliary way through an artificial intelligence system; inputting the screened treatment scheme into a drug management system, and automatically generating a personalized drug use plan and a corresponding early warning mechanism by the system; after the drug management protocol is implemented, the blood viscosity of the patient is monitored again using an automatic blood viscosity analyzer to verify the persistence and stability of the therapeutic effect.
3. The method of claim 1, wherein said constructing a large sample database containing patients of different ages, sexes, and underlying diseases, collecting clinical data of treatment response and adverse reaction during multi-drug treatment, analyzing the characteristics and outcome of the patients using deep learning technique, and establishing accurate prognosis prediction model, comprising:
The age and sex of the patient and the existing basic diseases are recorded through the construction of a professional database and used as important basis for primary screening of the patient; the collected clinical data is processed by using a data cleaning technology, and the key is to extract the treatment drugs of patients and the information of the reactions to the drugs and adverse reactions possibly generated; classifying the cleaned data, and creating a plurality of sub-databases to analyze the response caused by different drug treatments based on the received treatment regimen; through text mining technology, key descriptions describing treatment effects and detailed information about complications are deeply mined and arranged in clinical data; screening out characteristic factors closely related to the treatment effect and having large influence by utilizing a characteristic selection algorithm and combining with deep analysis of the characteristics of the patient; inputting the screened patient characteristics and treatment effect data into a deep learning model, performing first training of the model based on the characteristics and treatment effect data, and predicting prognosis of the patient; the stability and the accuracy of the prognosis prediction model are evaluated by using a cross-validation method, so that the model can be ensured to meet the statistical requirements of research; after the model evaluation accords with the expectation, adopting an incremental learning technology to carry out further optimization and adjustment on the model so as to continuously introduce new clinical data and improve the prediction precision of the model; the optimized prognosis prediction model is deployed into a clinical decision support system to assist in formulating an individualized treatment regimen for an individual patient in accordance with the accurate prognosis information provided by the model.
4. The method of claim 1, wherein the designing an intelligent multi-drug compounding system for predicting efficacy and risk of different modes of administration using a machine learning algorithm based on factors such as individual patient parameters and severity of illness, provides decision support for multi-drug treatment regimen, comprising:
Collecting individual parameters of medical history and vital signs of a patient, and integrating the information to ensure the accuracy of each data; analyzing the historical medical data by using a convolutional neural network algorithm, including treatment effects for similar cases; based on these data, a primary drug compatibility model is constructed, an initial weight score is defined for each drug based on the individual parameters of the patient and the relevant medical literature and study, this score reflects the possible effectiveness and risk level of the drug, and by further adjusting these initial weights, the selection and compatibility of the drug can be optimized until a combination appropriate for the particular patient is found; the risk assessment is carried out according to the severity of the current illness state of the patient, so that whether the patient needs to take medicines urgently or not is judged, and possible side effects are prevented in advance; a convolutional neural network algorithm is applied to predict and analyze potential curative effects under different medication modes, and a plurality of treatment schemes are proposed; the long-term and short-term risks of these regimens are assessed, and the depth analysis is compared to identify the risks that the patient may be at; checking the drug compatibility in these regimens using a drug interaction database ensures therapeutic safety; continuously analyzing the change of physical signs in the treatment process according to the feedback of the real-time medical monitoring data, and adjusting the drug scheme and the dosage according to the change; the expert system is used as a decision support tool, the expert knowledge is combined with the scheme generated by the algorithm, and a custom-made multi-drug treatment scheme is provided; a feedback mechanism is designed to continuously track the treatment effect of a patient, and the convolutional neural network algorithm is optimized through the actual result, so that the accuracy and the safety of a treatment scheme are improved.
5. The method of claim 1, wherein the monitoring vital sign changes of the patient during the multi-drug treatment in real time using the wearable sensor, evaluating safety of different dosing regimens, and adjusting the medication schedule as needed in combination with the blood viscosity monitoring results, comprises:
The method comprises the steps of monitoring vital signs of a patient by adopting a wearable sensor technology, and collecting real-time data, wherein the data is processed by an intelligent data analysis system to obtain detailed physiological parameters and viscosity monitoring results; the acquired data are transmitted to a data storage and management module in real time, and the module is responsible for recording health history information of a patient and taking the information as an auxiliary basis of a clinical decision support system; the personalized medical strategy module analyzes the monitoring data according to the specific condition of the patient and judges whether the current administration scheme needs to be adjusted so as to ensure the optimal treatment effect; when the monitoring data show that the patient has potential inadaptation symptoms or risks, the risk early warning system can timely give an alarm to doctors or patients; the safety evaluation module continuously evaluates the real-time vital sign data of the patient to determine whether adverse reactions caused by the medicines exist; if the evaluation result shows the potential risk, starting a medication plan adjustment module to reformulate a treatment scheme, and evaluating the interaction among different medicines through a multi-medicine treatment analysis module to optimize a treatment strategy; the treatment effect monitoring module continuously tracks the new treatment scheme to evaluate the curative effect of the scheme and ensure that the treatment target is achieved; if the curative effect monitoring module feeds back that the curative effect is poor, the medication plan adjusting module is started again, and the medication plan adjusting module is continuously optimized until the treatment scheme which is most suitable for the patient is found.
6. The method of claim 1, wherein constructing a multi-drug interaction knowledge graph, collecting drug-drug, drug-food, drug-disease interaction relationship data, and providing a reference for compatibility selection of a multi-drug combination, comprises:
Analyzing the medical literature and the drug instruction book by using a text mining technology, and extracting the attribute and action mechanism information of the drug from the medical literature and the drug instruction book; determining the possible effect of a specific food ingredient on the efficacy of the medicament by using information in a food ingredient database in combination with relevant knowledge of the nutrition; deep analysis is carried out on the relevant data of the diseases so as to identify the correlation between the action mechanism of the medicine and the disease characteristics; by identifying these correlations, it is possible to learn which drugs are at risk of treatment or side effects for a particular disease; the metabolic process of the drug in the body, particularly the interaction of the metabolite with other drugs or nutrients, can be better understood by the drug metabolic analysis tool; by analyzing medical guidelines and treatment manuals, the medical professional is provided with suggestions for the combined use of various medicines, and the most reasonable combined treatment scheme is helped to be established; by tracking the drug concentration monitoring data, the influence of the combined drug on the drug plasma concentration can be observed in real time, and the fluctuation trend of the drug concentration under various conditions is analyzed; in analyzing the side effect data, disease characteristics will be considered to determine the risk of side effects that different patient populations may encounter when using a particular drug; constructing a drug effect influence evaluation model based on the collected data and the research results of the former; integrating the data and the evaluation results obtained in all the steps, and deploying a set of dynamic knowledge map updating mechanism by using a pattern recognition and machine learning algorithm; the mechanism can automatically incorporate new research results and real-time data, thereby continuously perfecting the knowledge graph and providing scientific decision support for medical workers about drug combination.
7. The method of claim 1, wherein collecting compliance and lifestyle information during multi-medication of the patient, modeling the assessment of the correlation of these factors with medication efficacy and adverse effects, providing health management advice, comprises:
Carrying out detailed acquisition of the drug treatment information of the patient, recording the drug name, the dosage, the drug taking frequency and the specific taking time period, and establishing a basic database for subsequent analysis; collecting patient compliance data relating to whether the patient took the medication on time according to the order, and monitoring and marking those missed or misadministered events; developing a comprehensive investigation of the patient's lifestyle, including evaluating and recording diet, exercise, sleep, and stress-coping patterns, forming a lifestyle dataset related to health conditions; combining basic health information of a patient to construct a comprehensive database, wherein the comprehensive database comprehensively records the overall health condition of an individual; deep analysis of medication compliance and relationships with patient health using data mining techniques is aimed at identifying key factors that may affect the efficacy of medication; the statistical means is used for researching the relation between life style and medicine curative effect and adverse reaction, and potential relation between life habit and medicine safety is mined; based on the analysis, a prediction model is established, the model can evaluate the possibility of influence of different living modes on the drug reaction of a patient, and then the calibration and verification operation is carried out on the model, so that the accuracy of the model is ensured; providing personalized health management advice for the patient according to the result of the model prediction, helping the patient to make adjustment on the use of the medicine and encouraging the patient to improve life habits; dynamic monitoring of patient medication compliance and lifestyle changes is achieved, and these data are reviewed periodically to continuously optimize the predictive accuracy of the model and adjust health management measures accordingly.
8. The method of claim 1, wherein the intelligent voice assistant for designing medication reminders and monitoring to perform condition tracking and counseling services to assist patients in properly administering multiple medications, comprises:
Collecting health and drug usage records of the patient through a disease tracking system for monitoring and analyzing the progress of the patient and the rules of drug usage; using the health data stored in the patient medical profile to compare the drug characteristics of the drug information database, the system automatically checks potential interactions of the current drug to prevent unsafe drug combinations; according to the analysis result of the drug interaction, calculating safe and effective drug dosage by utilizing a compound pharmacological algorithm, and adjusting a personalized medication reminding plan for a patient to ensure medication time and dosage accuracy; the intelligent voice interaction function enables patients to verbally inquire about the medication records and medical orders of the patients through a voice recognition technology, and quick self-management is achieved; when a patient has problems on medicines or diseases, the patient is automatically connected to related medical professionals through the consultation service interface to provide immediate professional consultation services; extracting proper information from a medication instruction knowledge base, educating a patient about a correct use method of a medicament, and enhancing the ability of the patient to follow medical advice; periodically generating disease course reports containing disease progress information of the patient by using a disease course management statistical function, and adjusting a treatment plan of the patient according to the reports so as to help doctors evaluate treatment effects; the feedback and health data of the patient are synchronized in real time, so that all relevant medical information is kept up to date, and medical staff can effectively monitor and intervene in time; and continuously monitoring the system performance, evaluating potential risks or existing problems, and periodically optimizing the system functions according to the collected data and user feedback to continuously improve the service accuracy and user experience.
9. The method of claim 1, wherein re-detecting patient hemostatic parameters, evaluating blood viscosity changes before and after multi-drug therapy, creating a database to collect effective cases, providing support for new drug development, comprises:
Firstly, an automatic blood analysis system is deployed to track hemostatic parameters of a patient so as to accurately monitor the change of the hemostatic parameters of the patient and ensure that detailed data before and after treatment is collected; using the collected hemostatic parameters as a research basis, performing skin-level analysis on blood states before and after multi-drug treatment by using hemodynamic simulation software so as to grasp specific changes of blood viscosity; in order to improve the data management efficiency, an effective database architecture is constructed, hemostasis parameters and blood viscosity change data are integrated, and the integrity and consistency of the key information are ensured; the method has the advantages that through the inspiration of the existing data, the optimization is performed by applying a data mining algorithm, and real-time and effective decision support is provided for the development of new medicines in progress; defining and perfecting efficacy evaluation criteria, and analyzing and verifying the actual efficacy of the treatment regimen recorded in the database; further analyzing the correlation between the hemostatic parameters and the blood viscosity before and after treatment by using a defined curative effect evaluation standard and a statistical method to identify a significant difference; aiming at the data trend obtained by analysis, the existing multi-drug treatment scheme is subjected to necessary adjustment so as to improve the accuracy and pertinence of the treatment scheme; realizing a dynamic updating mechanism, ensuring that the latest research discovery and treatment results and related parameters can be reflected in a database in real time; a periodic reporting regime is formulated to automatically generate a comprehensive treatment effect report containing treatment regimen adjustment recommendations and patient hemostatic parameter changes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410127244.XA CN117995430A (en) | 2024-01-30 | 2024-01-30 | Multi-drug treatment response prediction method based on blood viscosity change |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410127244.XA CN117995430A (en) | 2024-01-30 | 2024-01-30 | Multi-drug treatment response prediction method based on blood viscosity change |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117995430A true CN117995430A (en) | 2024-05-07 |
Family
ID=90897190
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410127244.XA Pending CN117995430A (en) | 2024-01-30 | 2024-01-30 | Multi-drug treatment response prediction method based on blood viscosity change |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117995430A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105160199A (en) * | 2015-09-30 | 2015-12-16 | 刘毅 | Continuous blood sugar monitoring based method for processing and displaying diabetes management information with intervention information |
CN112041934A (en) * | 2018-03-23 | 2020-12-04 | 柏奥赛私人有限公司 | System and method for personalized medication management |
CN116705264A (en) * | 2023-06-20 | 2023-09-05 | 广州医科大学附属第三医院(广州重症孕产妇救治中心、广州柔济医院) | Medical record timing medicine taking reminding system |
CN116721730A (en) * | 2023-06-15 | 2023-09-08 | 医途(杭州)科技有限公司 | Whole-course patient management system based on digital therapy |
-
2024
- 2024-01-30 CN CN202410127244.XA patent/CN117995430A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105160199A (en) * | 2015-09-30 | 2015-12-16 | 刘毅 | Continuous blood sugar monitoring based method for processing and displaying diabetes management information with intervention information |
CN112041934A (en) * | 2018-03-23 | 2020-12-04 | 柏奥赛私人有限公司 | System and method for personalized medication management |
CN116721730A (en) * | 2023-06-15 | 2023-09-08 | 医途(杭州)科技有限公司 | Whole-course patient management system based on digital therapy |
CN116705264A (en) * | 2023-06-20 | 2023-09-05 | 广州医科大学附属第三医院(广州重症孕产妇救治中心、广州柔济医院) | Medical record timing medicine taking reminding system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20230178203A1 (en) | Systems and methods for patient-specific dosing | |
US20180039726A1 (en) | Computer based system for predicting treatment outcomes | |
US20200221990A1 (en) | Systems and methods for assessing and evaluating renal health diagnosis, staging, and therapy recommendation | |
JP5209714B2 (en) | System and method for creating treatments specific to each patient based on patient physiology modeling | |
CA3118297A1 (en) | Systems, methods, and apparatuses for managing data for artificial intelligence software and mobile applications in digital health therapeutics | |
US20060173663A1 (en) | Methods, system, and computer program products for developing and using predictive models for predicting a plurality of medical outcomes, for evaluating intervention strategies, and for simultaneously validating biomarker causality | |
US20100324874A9 (en) | Simulating patient-specific outcomes | |
JP2008546117A (en) | System and method for dynamic quantification of disease prognosis | |
US20090150134A1 (en) | Simulating Patient-Specific Outcomes | |
CN105005709A (en) | Process mining method based on single-disease treatment process | |
US20160358282A1 (en) | Computerized system and method for reducing hospital readmissions | |
CN113506622A (en) | Intelligent diagnosis and analysis system and use method thereof | |
US11705230B1 (en) | Assessing health risks using genetic, epigenetic, and phenotypic data sources | |
Lin et al. | The impact of a medication record sharing program among diabetes patients under a single-payer system: The role of inquiry rate | |
CN117995430A (en) | Multi-drug treatment response prediction method based on blood viscosity change | |
AU2022339235A1 (en) | Digital twin systems, devices, and methods for treatment of the musculoskeletal system | |
US11901083B1 (en) | Using genetic and phenotypic data sets for drug discovery clinical trials | |
US20030181792A1 (en) | Healthcare outcome measurement, analysis and improvement system | |
Jalali et al. | Predicting Clinical Deterioration in Hospitals | |
Baca | HarborPath: Examining Reach and Pathways to Data Improvement | |
CN117672516A (en) | Cardiovascular disease health management system based on user behaviors | |
Sinnige et al. | Clinical Medication Reviews in the general practice population: who and why? | |
CN116434906A (en) | Clinical comprehensive evaluation method for medicine based on dual-channel data acquisition | |
CN118016312A (en) | Intelligent curative effect evaluation cloud platform | |
Dió | Development of an Intelligent Tele-Diabetology System |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |