Claims:1. A method for detect of anomalies in health record, the method comprising:
In the invention, method to find different methods available for predicting and
classifying chronic illnesses in healthcare records and what if there are irregularities
such as anomalies in the healthcare data itself. The invention findings can't be
undermined at any expense. Any mistake can amount to loss of life; thus,
identification of irregularities such as anomalies in healthcare records is so essential.
A Med-Claim data set MD containing records of the patient with respect to inpatient,
outpatient, and carrier claims. The pre-process and model the given MD into chronic
and non-chronic condition categories using ICD and HCC codes. The data MD should
be processed to select the records of the member patient suffering from respective
chronic diseases using ICD. The five chronic diseases that should be analyzed are
Diabetes, Heart, Liver, Kidney, and cancer. All the ICD codes in MD should be
mapped to HCC codes. MD data should be transformed into TDM data. Detect the
anomalies in TDM data and validate the same using appropriate approaches.
2. An method according to claim 1,Capability of system implement HCC model on
health records .
3. An system according to claim 2,Capability of system to analyse, record, and store
health records to cloud server .
4. An system according to claim 1,Capability of system to detect anomalies in TDM
data and validate the same using appropriate approaches. , Description:In the invention, Healthcare has been a vital entity as it defines the quality of life. Health
care is an effort made to achieve wellbeing of physical, mental, or emotional needs by
trained and licensed professionals. One of the obstacles in today’s modern health care
services is dealing with inaccurate diagnosis of health conditions and hurriedly recorded
patient information. As per Medical records, the records give insight into the past, clinical
results and prescription of the patient. During legal-battles, health records often play a
pivotal role, whenever a mishap happens. There have been different methods and models
available for predicting, classifying, and categorizing chronic illnesses in health care but
what if there are irregularities in the medical record. Even A well designed model
wouldn’t work as expected if input data possess anomalies. Any erroneous decision on
regulating health care might even lead to fatality.
Anomalies are unpleasant component of a well recorded medical record. Detection of
abnormalities includes the issue of finding trends in data that do not correspond to normal
behavior. Abnormalities are typical issues which can arise in unplanned, nonstandardized
repositories. Elimination of anomalies in medical data is essential for
medical treatment providers for efficiently determine and treat persons who suffer from
chronic illness. The widespread availability of electronic form of patient health records,
it helps to over plan patient treatment procedures followed in healthcare setup. On the
very same instance, however it develops necessary to take care of any anomalies based
on these trends to effectively control and treat patients. Experts have established
principles from different fields such as analytics, ML, data mining, evolutionary
computation, textual theory to establish an effective approach for resolving anomalies
that occur in various fields of study such as banking, health, security, insurance, etc.
In this invention, a variety of anomaly detection and evaluation methods is used
to clean medical records so that the lives of healthcare stakeholders are less strenuous.
The CMS Medi-claim dataset is employed for representing illnesses in terms of
classification of Diseases (ICD) and Hierarchical Condition Category (HCC) to show the
2
nature of patient health. Three different algorithms are developed in this work, such as
the Graph model (GM), Principal Component Analysis (PCA) method and the Game
Theory (GT) method. The very first two methods are for identification of phenomena,
thus the next method is for evaluation of the effects of the initial two methods, the data
input of the mentioned techniques is Document Term Matrix (TDM) [3]. The outcome
of GM and PCA method is evaluated using GT approach. What appears to be important
is that an empty set is the output of a juncture of HCC codes after those in the results of
anomaly detection, and GT. The conclusion of the incorporation of GM and PCA model
is a successful procedure to find anomalies in the medical history. Compared with already
known methods, the outcomes from this research have found to be hopeful. As this study
has an impact on the lives of patients, it is strongly suggested that the findings should be
extensively wetted by domain expert.