4.1 KiB
4.1 KiB
Smart healthcare
Intro
- History:
- Patient centric
- Precision medicine
- Atrial digital twins
Reasons
- Current problems:
- Global Health Disparities
- Unequal shortage of healthcare workers
- Increasing rate of burnout in healthcare
- Mental health
- Sustainable Development Goal 3
- What smart healthcare can help
- Remote doctor consult
- Reduce un-necessary hospital visits
- Democratize healthcare access
- Remote emergency
- 5G powered emergency channel give doctor real time view
- Real time access of vital signs like ECG, ultrasound image, and blood pressure
- Access to medical history
- Doctors guide paramedics in ambulance
- Drug discovery
- Cost saving and agility methods
- All fancy words including (ML, DL, ANN)
- Offset a shortage of specialists
- AI can help in aiding diagnosis
- Before smart hospitals
- Data presented is overwhelming
- hindering the ability of using it effectively
- People trip and die in hospitals
- Too much administration work
- Shortage of qualified staff
- Spending wasted in managing assets
- Nurses waste time in searching for medical equipment
- Data presented is overwhelming
- Benefits of smart hospitals
- Save money on asset searching
- Reminder
- Home-based health monitoring reduces re-admission
- Eliminate emergency
- Warning of deterioration
Implementation
- EMR: Electronic medical records
- Diagnosis, Image Data, Unstructured text data, Medications, Procedures, Lab tests
- Data is challenging to use: high dimensionality, sparse, noisy, irregular, biased
- IoMT: lots of data, lots of money going in
- Healthcare technology: Better quality of life, with less money required
Research: data driven health accessment
Background
- Role of AI
- Detection of aliment
- Improve decision making
- Help in treatment
- Superior experience
- Tech revolution
- Automated steps in diagnostic process
- Can acquire different types of data, representing the physiology of humans
- Images: MRI, CT
- Wearable devices
- Electronic health records
- Robotics
- ML learn from data, so that it can make decisions like human brain
Data
Wearables
- Versatile monitoring
Patient Health Records
- Personal demographics: age, gender, ethnicity, weight, height, BMI
- Medical parameters: blood, pressure, fat, blood sugar, plasma
- Non-medical parameters: physical activity, diet plan
- Questionnaires filled
- Text mining can be performed
Structured vs. unstructured
- Structured: Time, date and code in diagnosis logs, can be stored in RDBMS
- Semi-structured: medication taken, procedures, allergic reactions
- Unstructured: Clinical notes, need NLP tools
Smart digital health platform
Diagram
Use Cases
- Heart disease detection
- Computer aided retinal disease diagnosis
Conclusion
- Future of healthcare technology involves:
- Self monitoring
- AI for health management
- Research is being developed to integrate real-time data with technology
- Require future clinical trails
- Explainable artificial intelligence:
- Need model transparency for clinicians and stakeholders
- Need collaboration in all fields:
- Academics
- Clinicians
- Industries
- Stakeholders