Multi-functional Biosensors and Advanced Low Laser Therapy with Computer-aided Disease Diagnosis for Smart Healthcare Application
- Abstract
- The advances in physiological monitoring, and analytics is particularly important for the prevention of cardiovascular problems. Our research focuses on developing a body-worn biosensor patch for continuous and longitudinal measurement of vital signs and other health-related outcomes. The main contribution of the research includes: (1) A hardware design of a wireless, and wearable biosensor device with high stability and flexibility for long-term monitoring was introduced. (2) An internet of medical things (IoMT) ecosystem with edge devices, a cloud database, a smartphone app, and a management website was developed in this study. (3) The practical application of the multimodal wearable biosensor device was tested under different exercise conditions for physiological response monitoring and self-monitoring compliance. The integration of the body-worn biosensor patch with IoMT is necessary for the development of cardiovascular management in near future.
A secondary line of research is in the area of computer-aided diagnosis via machine learning/ deep learning applications. Early diagnosis and treatment of many dermatological diseases are essential to minimize infection and negative side effects. However, manual approaches depend on the clinical appearance and experience of doctors for dermatologic diagnosis, which may lead to misdiagnosis, high prices, and low treatment efficiency. To overcome these problems, our research proposes a smart low laser therapy (LLLT) system for automatic facial dermatological disease diagnosis based on deep fully convolutional networks and IoMT applications. The overall contributions and advantages of this study can be summarized as follows: (1) A detailed hardware and software design of a smart LLLT system with automatic phototherapy configuration settings is developed. (2) A synthetic data generation process is proposed to solve the issue of the small dataset and enhance the robustness of the proposed deep learing (DL) models. The performance of different DL models, including Yolo, U2-Net, U-Net, and Segnet-5 was assessed and compared to identify the most efficient DL model for dermatological disorder diagnosis. Finally, the practical application of automatic facial skin disorder diagnosis with LLLT was demonstrated. Moreover, the feasibility of IoMT platforms for phototherapy care and management was verified in this study. The integration of artificial intelligence and IoMT-based healthcare platforms plays important role for the development of medical assistant tools in the near future. Besides, our research also has worked extensively and collaborated with other experts to develop a real-time closed-loop phototherapy system for controlling thermal dose during photothermal therapy using a non-contact infrared thermal sensor array and a deep neural network (DNN). The combination of real-time thermal control with DNN in predicting surface temperature is a promising method for minimal thermal injury on the surrounding tissue and improved the efficiency of papillary thyroid microcarcinoma treatment.
- Issued Date
- 2023
- Awarded Date
- 2023-02
- Type
- Dissertation
- Publisher
- 부경대학교
- Affiliation
- Pukyong National University, Graduate School
- Department
- 대학원 4차산업융합바이오닉스공학과
- Advisor
- Junghwan Oh
- Table Of Contents
- Chaper 1 Introduction 1
1.1. Motivation 1
1.2. Goal and outline of thesis 2
1.3. References 4
Chaper 2 A flexible, wearable, and wireless biosensor patch with internet of medical things applications 5
2.1. Introduction 5
2.2. Materials and Methods 9
2.2.1 A biosensor patch with a healthcare IoT application 9
2.2.2 Device structure and mechanical performance 10
2.2.3. Hardware Architecture 12
2.2.4. IoT-Connected Healthcare Platform 15
2.3. Results 17
2.3.1. Health status monitoring 17
2.3.2. Advanced Use Cases in BP Estimation 19
2.4. Discussion 22
2.5. Conclusions 25
2.6. References 27
Chaper 3 Noninvasive, Wearable Multi Biosensors for Continuous, Long-term Monitoring of Blood Pressure via Internet of Things Applications 31
3.1. Introduction 31
3.2. Materials and Method 35
3.2.1 Multimodal body sensors for BP estimation with IoT application 35
3.2.2 Noninvasive, cuff-less BP device hardware architecture 36
3.2.3 BP-connected IoT healthcare platform 39
3.2.4 Blood pressure estimation 40
3.2.5 Subject and experiment details 45
3.3. Simulation Results and Analysis 47
3.3.1 Evaluation of impedance measurement accuracy 47
3.3.2 Correlation between extracted PTT features and BP 48
3.3.3 BP estimation using different PTT-BP models 51
3.3.4 Test of the BP device with different exercise tasks 54
3.3.5 Performance of the BP device in long-term monitoring cases 56
3.4 Discussion 57
3.4.1 PTT-BP estimation models 58
3.4.2 Blood pressure response to exercise and hypertension prediction 59
3.4.3 BP estimation based on PTT and impedance plethysmography 60
3.4.4 Limitations 60
3.5 Conclusions 61
3.6 References 62
Chaper 4 A smart LED therapy device with an automatic facial acne vulgaris diagnosis based on deep learning and internet of things application 65
4.1. Introduction 65
4.2 Smart LED light therapy system 70
4.2.1 LED therapy device hardware architecture 70
4.2.2 Deep learning-based autonomous acne detection 72
4.2.3 LED therapy control program 83
4.2.4 IoT-connected healthcare platform and smartphone application 86
4.3 Experiment and result 90
4.3.1 Acne detection with YOLOv2 90
4.3.2 LED therapy device with an automatic facial acne detection 93
4.4 Discussions and conclusions 95
4.5 References 99
Chaper 5 A flexible, and wireless LED therapy patch for skin wound photomedicine with IoT-connected healthcare application 103
5.1 Introduction 103
5.2 Results 107
5.2.1 Design of the flexible wireless LED therapy device 107
5.2.2 Structure of the flexible wireless LED therapy device 109
5.2.3 Electrical and Optical Properties 112
5.2.4 Thermal Stability 115
5.2.5 IoT-Connected Healthcare platform 117
5.2.6 Photomedical Application 119
5.3 Discussion 127
5.4 Conclusion 131
5.5 References 133
Chaper 6 Enhanced precision of real-time control photothermal therapy using cost-effective infrared sensor array and artificial neural network 136
6.1 Introduction 136
6.2. Materials and methods 139
6.2.1 Schematic diagram and experimental setup 139
6.2.2 Photothermal control and temperature prediction 142
6.3 Experiment and Results 153
6.3.1 Network training and selection 153
6.3.2 Laser control with feedback-controlled temperature 156
Chaper 7 Conclusions 169
7.1 Findings 169
7.2 Future work 171
Abstract in Korean 173
Publications 175
Acknowledgments 178
- Degree
- Doctor
- Appears in Collections:
- 대학원 > 4차산업융합바이오닉스공학과
- Files in This Item:
-
-
There are no files associated with this item.
-
Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.