Fitbit Wearable Devices and CGM Integration into Diabetes Using Multi Criteria Optimization Approach for Decision Support System Based on IoMT Clinical Data

Fitbit Wearable Devices and CGM Integration into Diabetes Using Multi Criteria Optimization Approach for Decision Support System Based on IoMT Clinical Data

Introduction

The standard practice of CGM (Continuous Glucose Monitoring) is now beneficial for the diabetic patients. The CGM allows intensive control over the blood glucose level and the data is collected as well as monitored with the help of CGM. It allows recurrent accessibility to the daily glucose patterns and improving the control over the glucose levels of the body. Asymptomatic hypoglycemia is also detected with the help of CGM. The improved self-management of the patient’s health is made possible with the help real-time advancements in existing CGM technology. The betterments in the system such as glucose threshold pumps for the prevention of hypoglycemia and auto responding insulin pump that acts as artificial pancreas (AP) allow maintaining the glucose level of the body in diabetic patients. The term artificial pancreas that is used in the research is still under the consideration of prominent research of medical field to combine artificial intelligence and internet of medical things with real time biological needs of the human body. The sense and responding nature of IoMT allows better management and control of glucose levels in diabetic patients [1]. The limitation to the sense and respond mechanism is that the delay between sensing and administrating the action for giving current dose of insulin still takes time and scientists are working their best to minimize the time to respond. The pump controlling algorithm and the response time is to be synchronized for the compensation of the issues that the insulin pumps face due to lag in their response. Certain other parameters are still to be considered in detail such as state of the patient (resting or exercising) for monitoring the situations such as nocturnal hypoglycemia and predictive algorithm for suspended low glucose values. Diabetic patients can be monitored continuously with the help of CGM and the display interface.

Problem Statement

This research will focus on Tele-Medicinal Integration of CGM for the easy access. The advanced evaluation techniques and the focused research will allow the patients to monitor their health status remotely with the help of remote CGM’s. These CGM’s will reduce the pain of sensors to be installed on the body of patients. Applications that are remotely attached to the CGM will be beneficial for both doctor and patient to monitor the status of his/her glucose levels. The Remote Patient Monitoring (RPM) can be done easily with the help of proposed innovation in the existing CGM model.

Research Question

CGM RPM Remote Patient Monitoring to increase the heath system of type 2 diabetes patients and evaluation of existing applications and systems for proposing better solution in the field of Telemedicine.

Literature Review

Fitbit wearable watch can help preventing the patient with the condition of hypoglycemia during the exercise. A range-controlled AP system can then be triggered if the patient at any time faces the condition of Hypoglycemia during the exercise. The glucose monitoring and managing it while doing a physical activity with the help of sensors is the core concept of this research. Number of fitness trackers and smart watches use interactive software packages to monitor and manage the diabetic condition of a patient. These smart watches and trackers have multiple sensors that are linked with CGM to monitor and control the condition of patients and maintaining their glucose level. A healthy lifestyle is encouraged with these trackers and a lot of persons are using these trackers for the monitoring and management of their activities and their physical achievements. People with or without diabetes both use these watches for the promotion of a healthy lifestyle and monitoring the physical standards of their body. These trackers also monitor blood pressure, heart rate, oxygen level, number of steps and various other parameters for staying healthy. The increased amount of glucose level allows triggering rapid action for maintaining the dose of insulin and allowing the patient to stay healthy. 

The signals from the CGM can be seen on the screen of the wearable watches which allow the person to continuously monitor and decide his/her actions for a certain condition. The glucose level and the upper and lower range that is set by the user on the watch allows the watch to beep of the level goes up or drops down from the routine values. This allows the patient to continuously have an account of his glucose level and take precautionary actions if the values fluctuate from the regular values. Certain exercises that are practiced for attaining a healthy lifestyle are incorporated with the algorithms of the CGM as to see their dependency with the blood glucose level. The artificial pancreas control algorithm combined with other algos allow detection of abnormal patterns that are linked with glucose excursions of the patient’s need. In this research it focused that either the prediction of future glucose levels is enhanced with the help of increased number of physiological sensors or not. The data gathered from fitness trackers and their combined values with that of CGM and Insulin Pumps are to be optimized by eradicating the problems in their advancement. The difficulties that are linked with use of this system are the heat from the band that can cause burns on the wrist. An increased number of sensors mean overheating and increase in size that can be a challenge for the smart bands. The Application Programming Interface (API) allows considering raw data for real time applications but there is a need to keep the smartphone within range while recording the values from the smart band. The gaps in the evaluation data can be reduced with the help of multiple sensors and raw data evaluation.

 

 

 

 

 

The compact size of smart bands allows it to wear it with comfort and provides the user an experience of continuous monitoring his/her health status [4]. The challenge with smaller size is that maintaining the battery life is hard and it must be powerful enough to keep the watch powered up for maximum time so that there must be minimum gaps in the collected data [2]. Physiological sensors are a source of raw data that is continuously monitored with the help of smart band. Raw data signals, applications and algorithms allow development of wearable devices as well as CGM integration for filling the gap between research and innovation for telemedicinal purposes.  

Multicriteria Optimization Approach

For the optimal estimation of performance two objective functions were considered named as Productive Efficiency (PE) and Quality Effectiveness (QE). The Goal (G) is set assuming that it is the computational outcome of PE and QE. G = A+B1*PE+B2*QE+B3*(PE*QE) is the value of G. In the formula for attaining the goal (G), term A refers to the constant term also known as intercept. B is the slope coefficient attained via regression of the estimated equation. Interaction of PE and QE and their statistical importance is highlighted with the help of this model. 

Beta can be found out with the help of regression done on standard coefficients which are the values obtained from the participants of the research. 

Clinical Test Ranges for Diabetes Diagnosis

Results                                  A1C                                            FPG                                      OGTT Normal                     Less than 5.7%            Less than 100 mg/dl         Less than 140 mg/dl Prediabetes                  5.7% to 6.4%              100 mg/dl to 25 mg/ dl         140 mg/dl to 199 mg/ dl Diabetes                      6.5% or higher                126 mg/dl or higher                   200 mg/dl or higher

Multilevel Objective Function Blocks used in Optimization for Optimal Solution

Predication

 

log (riskF) = ΣβiXi,F + βweightXweight,F

hF*t1/hF(t0) = h0(t1) exp (βiXi,F + βweightXweight,F*) /h0(t0) exp (βiXi,F + βweightXweight,F) =  h0(t1)/h0(t0)   exp (βweight(Xweight,F* - Xweight,F)=h0(t1)/h0(t0) exp (βweightΔXF) 

In the above graph, A,B,C,D,E,F, and G are the seven patients whose data is considered in the research. Y denotes output and X denotes input at any point in the graph. The above computational logic is adopted for the optimization and the enhancement of the evaluation for clinical data.

Multi-Level Research Model for Multi-Criteria Optimization

 

Gantt Chart

Collection of Data Evaluation of Data/ Data Analysis Integration with CGM and Online Applications Successful Optimization of the System Enhanced Findings and Contribution 

June

July

August

September

Challenges

While doing research on Multi-Criteria Optimization Technique, there are three most prominent challenges that are faced throughout the research. The first challenge is standardization of measurements and scales while using transdisciplinary approach for data handling. Second challenge is patient’s education and familiarity with the technological usage and outcomes of the results that are monitored via CGM. Third challenge is the analysis of multivariable outcomes and their autoregressive nature for repeated data sets [6].

The above challenges highlighted the importance of data and standardization for the Fitbit wearable device to stay in market and compete with other brands. Cost effectiveness and battery life are also the major parameters that are to be considered for the successful launch of a wearable device to fulfill the future needs of the patients. Accuracy and burden of the gadget must be considered as an active parameter. Fitbit allows people to live healthier by empowering them with inspiration and guidance by monitoring and control over their goals. The track and motivation provided by Fitbit allows patients to live healthier and enjoy their daily activities with complete record over their wrist. 

The Fitbit health package includes guidance via leading software and interactive platform, there are several applications that include Fitbit Coach Applications and Operating Systems for Smart Watches as well as paid subscriptions such as Fitbit Premium for advanced analytical and guidance for the fitness goals. One on one virtual coaching is provided on health coaching services and Fitbit premium as to design personal plans with help of data available via Fitbit applications and plans. Positive health outcomes and the increased engagement with Fitbit health solutions allow design of positive health outcomes and health systems.   

Integrating Fitbit Wearable device and its Results

A clinical study in Taiwan named as Health2Sync found out that there are significant improvements in blood glucose and HbA1C with the help of proper care and monitoring of diabetes via Fitbit Wearable Devices. The proper care and monitoring provided by Fitbit allowed improvements in the health of patients and they now feel more relaxed due to constant check and balance of their blood glucose level. Type 2 diabetes can be easily monitored and controlled with the help of glucose control application provided by Health2Sync that works in collaboration with Wearable Fitbit Devices. The results also highlighted that with improved physical activity and continuous monitoring the health condition of the patient improved and the HbA1C was reduced. The higher blood glucose measurement frequency allows improvements in the condition of the patient and his glucose level is now easily maintained with the help of proper monitoring as provided by Fitbit. The consultative advice can be taken on time with proper monitoring being provided by Fitbit. The Average Glycated Haemoglobin (HbA1C) was decreased to a percentage of 0.33 while the percentage of HbA1C decrease for those patients that moderated their activity for 150 minutes/week was 0.66 percent.The BG (Average Fasting Blood Glucose) decreased to 10.92 mg/dl. Average low-density lipoprotein cholesterol (LDL-C) decreased 11.55 mg/dL. Weight reduction of up to 2 kilograms among some patients. Increased frequency in moderate to high intensity activity to 7.03 times a week among some patients [3].

Contribution

With the help of increased evaluation of CGM Systems that are available, telemedicine can be empowered for the ease of patients. This progress in the existing CGM systems and their integration with latest technology will allow to proceed in the domain of research to deal with the problems of diabetic patients. Time and lives of the patients can be saved with the help of betterments that are proposed in this research [5]. Additionally, this research will increase the data security and accountability of the patient’s useful information by use of secure communication with the help of advanced technological methods. Evaluation and access of patient’s current situation and his previous results will be made rapid and fast with the proposed system.

References

Dwivedi, R., Mehrotra, D., & Chandra, S. (2022). Potential of Internet of Medical Things (IoMT) applications in building a smart healthcare system: A systematic review. Journal Of Oral Biology And Craniofacial Research, 12(2), 302-318. https://doi.org/10.1016/j.jobcr.2021.11.010 

. . . . . . . . . . . . . . . . . .  …  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 

Schwartz, F., Marling, C., & Bunescu, R. (2018). The Promise and Perils of Wearable Physiological Sensors for Diabetes Management. Journal Of Diabetes Science And Technology, 12(3), 587-591. https://doi.org/10.1177/1932296818763228 

. . . . . . . . . . . . . . . . . .  …  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Integrating Fitbit wearable devices into diabetes care leads to significant improvements in blood glucose and HbA1C, finds Health2Sync clinical study in Taiwan. Investor.fitbit.com. (2020). Retrieved 7 June 2022, from https://investor.fitbit.com/press-releases/press-release-details/2020/Integrating-Fitbit-wearable-devices-into-diabetes-care-leads-to-significant-improvements-in-blood-glucose-and-HbA1C-finds-Health2Sync-clinical-study-in-Taiwan/default.aspx. 

. . . . . . . . . . . . . . . . . .  …  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Innovative Wearable Devices Offer Patients a New Way to Manage Their Conditions. Prnewswire.com. (2020). Retrieved 7 June 2022, from https://www.prnewswire.com/news-releases/innovative-wearable-devices-offer-patients-a-new-way-to-manage-their-conditions-301056619.html. 

Usoh, C., Kilen, K., Keyes, C., Johnson, C., & Aloi, J. (2022). Telehealth Technologies and Their Benefits to People With Diabetes. Diabetes Spectrum, 35(1), 8-15. https://doi.org/10.2337/dsi21-0017 

Wan, T., Matthews, S., Luh, H., Zeng, Y., Wang, Z., & Yang, L. (2022). A Proposed Multi-Criteria Optimization Approach to Enhance Clinical Outcomes Evaluation for Diabetes Care: A Commentary. Health Services Research And Managerial Epidemiology, 9, 233339282210891. https://doi.org/10.1177/23333928221089125 

Gupta, K., Roy, S., Poonia, R., Kumar, R., Nayak, S., Altameem, A., & Saudagar, A. (2022). Multi-Criteria Usability Evaluation of mHealth Applications on Type 2 Diabetes Mellitus Using Two Hybrid MCDM Models: CODAS-FAHP and MOORA-FAHP. Applied Sciences, 12(9), 4156. https://doi.org/10.3390/app12094156 

. . . . . . . . . . . . . . . . . .  …  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .