AU2021102318A4 - System for Improving Prediction Accuracy of Healthcare Ontology - Google Patents
System for Improving Prediction Accuracy of Healthcare Ontology Download PDFInfo
- Publication number
- AU2021102318A4 AU2021102318A4 AU2021102318A AU2021102318A AU2021102318A4 AU 2021102318 A4 AU2021102318 A4 AU 2021102318A4 AU 2021102318 A AU2021102318 A AU 2021102318A AU 2021102318 A AU2021102318 A AU 2021102318A AU 2021102318 A4 AU2021102318 A4 AU 2021102318A4
- Authority
- AU
- Australia
- Prior art keywords
- data
- prediction accuracy
- healthcare
- ontology
- models
- 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.)
- Ceased
Links
- 238000010801 machine learning Methods 0.000 abstract description 3
- 238000013138 pruning Methods 0.000 abstract description 2
- 238000007405 data analysis Methods 0.000 abstract 1
- 238000000034 method Methods 0.000 description 8
- 230000010354 integration Effects 0.000 description 7
- 230000002068 genetic effect Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000011524 similarity measure Methods 0.000 description 3
- 238000013179 statistical model Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000001154 acute effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000013502 data validation Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000003285 pharmacodynamic effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Landscapes
- Complex Calculations (AREA)
Abstract
TITLE OF INVENTION: System for Improving Prediction Accuracy of
Healthcare Ontology
FIELS OF INVENTION: COMPUTER SCIENCE
ABSTRACT
There is considerable concern about the lack of data in this digital age because the web is packed
with a great deal of data in different formats and sizes. Healthcare data is used in online and other
databases on a regular basis. The Semantic Web intelligently parses its ontological details.
Machine Learning and Big Data Analysis have included tree-based algorithms like ID3, C4.5
etc. If the dataset quality is relatively low, mis-pruning of trees can decrease the prediction
accuracy considerably. It has been suggested that EMIA is a solution based on the principles of
HMM and Stochastic Automata Model (SAM). The proposed application incorporates the rules
of diagnosed models based on the thresholds into an integrated Healthcare ontology resulting
in a prediction accuracy of 86 percent.
Description
TITLE OF INVENTION: System for Improving Prediction Accuracy of Healthcare Ontology FIELD OF INVENTION: COMPUTER SCIENCE
[001] Internet holds vast amount of knowledge on a gigantic scale. Semantic web deals with information or meaning about data which is present in the web. Medicine is one of the major domains among other industries that produce data in large quantities every minute in day to day life. However, the created data cannot function effectively even within the same domain if it is not connected to other data which is present in the same web.
[002] The data available in semantic web is already vast and complex. It has many patterns, missing values, various labels and many other things in it. Therefore, it is not suitable for automated computer information exploration. The data featured in the proposed graphical technique contributes towards better decision-making and increases the prediction accuracy.
[003] ("US7899764B2") : A medical ontology may be used for computer assisted clinical decision support. Multi-level and/or semantically grouped medical ontology is incorporated into a machine learning algorithm. The resulting machine-learnt algorithm outputs information to assist in clinical decisions. For example, a patient record is input to the algorithm. Based on the incorporated medical ontology, similarities are aggregated in different groups. An aggregate similarity of at least one group is a function of an aggregate similarity of another group. One or more similar patients and/or outcomes are identified based on similarity. Probability based outputs may be provided.
[004] ("US20070178501A"): The system described herein enables clinicians and researchers to use aggregated genetic and phenotypic data from clinical trials and medical records to make the safest, most effective treatment decisions for each patient. This involves (i) the creation of a standardized ontology for genetic, phenotypic, clinical, pharmacokinetic, pharmacodynamic and other data sets, (ii) the creation of a translation engine to integrate heterogeneous data sets into a database using the standardized ontology, and (iii) the development of statistical methods to perform data validation and outcome prediction with the integrated data. The system is designed to interface with patient electronic medical records (EMRs) in hospitals and laboratories to extract a particular patient's relevant data. The system may be used in the context of generating phenotypic predictions and enhanced medical laboratory reports for treating clinicians. The system may also be used in the context of leveraging the huge amount of data created in medical and pharmaceutical clinical trials. The ontology and validation rules are designed to be flexible so as to accommodate a disparate set of clients. The system is designed to be flexible so that it can change to accommodate scientific progress and remain optimally configured.
[005] ("W02006130162A2"): An information system using a healthcare ontology to provide a standardized representation for healthcare data is disclosed. One embodiment of the information system comprises a digital logic platform for storing and using the healthcare ontology. The healthcare ontology describes concepts and relationships between the contents derived from the corpus of domain specific knowledge and linking with standardized terminological systems.
[006] ("US8024128B2"): The information management system disclosed enables caregivers to make better decisions by using aggregated data. The system enables the integration, validation and analysis of genetic, phenotypic and clinical data from multiple subjects. A standardized data model stores a range of patient data in standardized data classes comprising patient profile, genetic, symptomatic, treatment and diagnostic information. Data is converted into standardized data classes using a data parser specifically tailored to the source system. Relationships exist between standardized data classes, based on expert rules and statistical models, and are used to validate new data and predict phenotypic outcomes. The prediction may comprise a clinical outcome in response to a proposed intervention. The statistical models and methods for training those models may be input according to a standardized template. Methods are described for selecting, creating and training the statistical models to operate on genetic, phenotypic, clinical and undetermined data sets.
[007] ("US20120095300Al"): Methods, systems, and computer storage media are provided for predicting a probability of acute deterioration for a specific patient. Various discrete measurements are taken regarding the patient's current health. Those measurements are used to determine a PPOD score, which is displayed for clinicians.
[008] The following presents a simplified summary of the invention in order to provide a basic understanding of some aspects of the invention. This summary is not an extensive overview of the present invention. It is not intended to identify the key/critical elements of the invention or to delineate the scope of the invention. Its sole purpose is to present some concept of the invention in a simplified form as a prelude to a more detailed description of the invention presented later.
Work progresses on extending the proposal for enhancing prediction accuracy across ontologies upon interaction connecting the semantic data at schema level for efficient utilization of the available datasets.
[009] The following description is of exemplary embodiments only and is not intended to limit the scope, applicability or configuration of the invention in any way. Rather, the following description provides a convenient illustration for implementing exemplary embodiments of the invention. Various changes to the described embodiments may be made in the function and arrangement of the elements described without departing from the scope of the invention.
Algorithm: Effective Model Integration Algorithm (EMIA) to improve prediction Accuracy
Input: Training dataset (T), Number of sets (M), Number of Models in each set (N) Output: Enhanced Model based on C4.5 classification rules
1. Let DS be the Dataset created from T of varied size, where IDSI= M 2. Let A be a vector representing prediction 3. Let BSM be a list with 3 data fields representing Binary Similarity Measure 4. Let Tv represent threshold of all models in particular setof M 5. Vm eM, IN/ = C / C is a decimal constant > 5 for optimal solution 6. for i= 1 to |MI do 7. for] = 1 to INI do 8. Using DSi, execute C4.5 algorithm 9. Document rules of model () 10. Document Prediction Accuracy of model(j) in A 11. end for 12. end for 13. VNE M, Compute Threshold (Tv) based on A grouped on M, 14. for i = 1 to |MI do 15. forj= 1 to INI do 16. if A(N) 1 TN then 17. Mark model n eligible for integration / where meN */ 18. end if 19. end for 20. end for 21. for i = 1 to IMI do 22. V (a,b) e N, /Ni gets updated with integrated models every iterations 23. ifmodels (a and b) are marked as eligible for integration then 24. Compute Binary Similarity Measure of a and b and store in BSM 25. Compute Threshold (TN) based on BSM scores from recent construct 26. V(a,b) e N present in BSM /* BSM from recent construct *7 27. if BSM(a,b) > TNthen / Thfrom recent construct 28. Integrate the models a &b and rename as Integrated model 29. end for 30. Repeat the preceding for loop until no more intra models are eligible for integration 31. Consider M itself as N as result of previous step 32. Repeat the preceding for loop until no more inter models are eligible for integration 33. Finally from the resultant integrated models, choose the model with highest prediction accuracy by verifying on C4.5 results as final Enhanced Model 34. Return the Enhanced Model
[010] The proposed idea starts with m sets with each set having n models equivalent to states of HMM in a SAM form. Instead of probability functions in HMM, the suggested approach considers the threshold values for integration of models if found similar proved upon computation of binary similarity scores between the models for models generalization. To begin with accuracies of all models in all the sets will be documented at level-i. The intra models get selected to further levels for integrating based on the following steps: For all a belongs to n in each model m • At Level-1: calculate accuracy, apply threshold to select the model a further. • From level-2 to level-p, using the Asymmetric Binary Similarity measure method, the concept of clustering is applied for determining the pair of rules to be integrated and the process continues till the last level to get final set of rules so as to attain better decision accuracy.
[011] The enhanced rules obtained using EMIA can be utilized efficiently by following the sequence of methodology starting with enhanced rule generation, followed by building Ontology - by applying the rules defined in .pie file using tools like GraphDB or using Cellfie tool available in protege and querying the constructed knowledge graph using SPARQL queries for efficient knowledge retrieval. Observations conclude that the ontology built using the stated methodology is witnessed with the increased derived instances of up to 10% compared to the derived instances of well-known decision algorithms resulting in good prediction rate comparatively.
[012] Semantic web instead of working on existing data files, works on concepts represented in the documents. Despite the advancements in artificial intelligence and machine learning, current algorithms are not to predict with accuracy without rule pruning. So to address these disadvantages, an Effective MIA (EMIA) has been proposed based on the principle of Hidden Markov Model and the Stochastic Automata Model (SAM). Thresholds are determined by accuracy ratings and Binary Similarity Tests. Different models from different sets are combined, until the baseline hypothesis is supported.
The experimental results show that LUCC is better than current techniques. The proposed model worked very well with an overall accuracy of 86%. Hence, our proposed solution performs better in any domain no matter which domain you are operating in.
Claims (1)
1. System for Improving Prediction Accuracy of Healthcare Ontology, generates improved decision rules by integrating the rules of chosen models based on threshold in a Healthcare ontology
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| IN202141011671 | 2021-03-19 | ||
| IN202141011671 | 2021-03-19 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| AU2021102318A4 true AU2021102318A4 (en) | 2021-06-17 |
Family
ID=76327713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| AU2021102318A Ceased AU2021102318A4 (en) | 2021-03-19 | 2021-05-02 | System for Improving Prediction Accuracy of Healthcare Ontology |
Country Status (1)
| Country | Link |
|---|---|
| AU (1) | AU2021102318A4 (en) |
-
2021
- 2021-05-02 AU AU2021102318A patent/AU2021102318A4/en not_active Ceased
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yuan et al. | Constructing biomedical domain-specific knowledge graph with minimum supervision | |
| US10949456B2 (en) | Method and system for mapping text phrases to a taxonomy | |
| Asiaee et al. | A framework for ontology-based question answering with application to parasite immunology | |
| Megdiche et al. | An extensible linear approach for holistic ontology matching | |
| Guan et al. | Natural language processing and recurrent network models for identifying genomic mutation-associated cancer treatment change from patient progress notes | |
| Cao et al. | Multi-information source hin for medical concept embedding | |
| Mayya et al. | Multi-channel, convolutional attention based neural model for automated diagnostic coding of unstructured patient discharge summaries | |
| Schäfer et al. | Multilingual ICD-10 Code Assignment with Transformer Architectures using MIMIC-III Discharge Summaries. | |
| CN110299194B (en) | Similar case recommendation method based on comprehensive feature representation and improved wide-depth model | |
| Boukhari et al. | Approximate matching-based unsupervised document indexing approach: application to biomedical domain | |
| Zeng et al. | Medical data mining in sentiment analysis based on optimized swarm search feature selection | |
| Memarzadeh et al. | A study into patient similarity through representation learning from medical records | |
| Almasoud et al. | Automated multidocument biomedical text summarization using deep learning model | |
| Zaghir et al. | Real-world patient trajectory prediction from clinical notes using artificial neural networks and UMLS-based extraction of concepts | |
| AU2021102318A4 (en) | System for Improving Prediction Accuracy of Healthcare Ontology | |
| Chraibi et al. | A Deep Learning Framework for Automated ICD-10 Coding. | |
| Ferreira et al. | Active learning for medical code assignment | |
| Devkota et al. | Knowledge of the ancestors: Intelligent ontology-aware annotation of biological literature using semantic similarity | |
| Hemati et al. | CRFVoter: gene and protein related object recognition using a conglomerate of CRF-based tools | |
| Rama Sree et al. | A Comparative Study on a Disease Prediction System Using Machine Learning Algorithms | |
| Dao et al. | Patient Similarity using Electronic Health Records and Self-supervised Learning | |
| Venkatesh et al. | A review on biomedical mining | |
| Kalita et al. | Fundamentals of Data Science: Theory and Practice | |
| Chitra et al. | Systematic Literature Review on Deep Learning Techniques in Electronic Health Records | |
| Weiss et al. | Predictive rule discovery from electronic health records |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FGI | Letters patent sealed or granted (innovation patent) | ||
| MK22 | Patent ceased section 143a(d), or expired - non payment of renewal fee or expiry |