Healthcare + AI: Use cases & Challenges
1 like1,235 views
This document discusses the potential of artificial intelligence (AI) and emerging technologies in healthcare. It begins with brief introductions and then outlines several key AI use cases in healthcare, including data collection and management, personal health data management, diagnosis, patient management, and macro health analysis. It also discusses challenges like skilled labor shortages and lack of large data sets. Risks of AI like bias, privacy issues, and dangerous mistakes are presented. The conclusion is that AI has great potential to transform healthcare if applications are handled carefully and data is managed appropriately.
Healthcare + AI: Use cases & Challenges
- 1. Healthcare + AI , Blockchain & Emerging Technologies Srinath Perera, Ph.D. VP Research WSO2, Apache Member, ( srinath@wso2.com) @srinath_perera
- 2. Myself • I am computer Scientist and knows little about Healthcare ( I might not use the right Jargon, please bear with me) • Although I work in the Industry, I have been an active researcher ( 1000+ citations) • I have been an open-source developer (10+ years) help design a few well-known packages ( e.g., Apache Axis2, WSO2 Siddhi) • I have designed many real-life systems that run by many fortune 500 companies and governments 2
- 3. 3 • Some tasks (e.g., make a coffee) we can explain, precisely. We can write a code to do them. • Other tasks, e.g., drive a car, we can do, but can’t explain how. We can’t teach our children how to drive by writing it down. Such “intelligent tasks,” we have to give them examples and feedback and train them. • AI can learn from examples to do such “intelligent tasks.” • Andrew Ng’s rule - “AI can do anything human can do in 10 sec or less.” What is AI?
- 4. What is the big deal? • Compared to humans, Computer doing these tasks (With AI) are • Faster (1K-1000K times) • Cheaper to replicate • Reliable • AI ability to learning can be used to extrapolate the future • Outcomes • Cheaper, efficient, new possibilities4
- 5. AI Use Cases in Healthcare
- 6. Data Collection and Management 6 • Everything starts with data • Needs processes and systems to collect data and manage them. • Data need to be collected by non- technical people • We need to make data collection as possible, e.g., Collect data like images, then using AI to extract information • Data need to be managed, protected and shared across organizations
- 7. Personal Health Data (History) Management 7 • Potential Scenario • Scan and keep all your health records ( in the phone, encrypted and backed up in the Cloud) • Walked to the doctor • Scans his App • Share your data for a limited time • For full security, we will need blockchain
- 8. 8 Diagnosis • IBM Watson & Google Deepmind Health, diagnosis as a subscription? • Build on all data, unto data, learning from every mistake • AI has beaten humans in many cases (e.g., Kaggle diabetic retinopathy challenge) • In focused specific disease diagnosis likely AI will better ( and hard to beat) • Humans have a chance in broad diagnosis • Patents built to solve diagnosis problems can be very valuable • Our access to data is a significant advantage
- 9. Patients Management 9 • Using AI to manage the patient through his stay ( or through a long time) • AI Use cases • Risk assessment • Anomaly Detection (WSO2 Use case) • Keep them Healthy • Proactive Tests/ Preventive Actions & Life cycle changes • Provide connect more/ relevant information for diagnosis • Optimal resource management
- 10. Macro Health Analysis and Management 10 • Using AI & Analytics to understand broad health patterns and act proactively or reactively prevent, manage, and contain situations • e.g., Dengue hotbed identification and acting proactively • Similar work in other fields - https://lirneasia.net/big-data Photo Credit, https://lirneasia.net/wp-content/uploads/2019/02/LIRNEasia-Data-Driven-Policy.pdf
- 11. Challenges 11 • Shortage of Skilled Professionals • The data scientists, programmers, and architects are in short supply and expensive (300-500K year). • It is hard for medium and smaller organizations to attract and hire enough skilled people. • Most AI solutions will be delivered as cloud APIs • Only possible when If data formats are well known (e.g., banking or healthcare data) and key performance indicators (KPIs) are well defined. Then reusable models can be built for those use cases. • Examples are disease detection in healthcare, marketing insights, spatiotemporal models, and analytics for specific sports.
- 12. Challenges (contd.) • Lack of Large Enough Data Sets. • With 10,000 data points per day, it takes 3 years to collect 1 million data points • Labeled data sets are hard to find • Interpretability • Some models are black boxes • Tracking provenance of data used
- 13. AI mistakes are Dangerous 13 • AI weaknesses can cause much more potential harm than the same weakness in humans • Broad integration - AI will be everywhere (e.g., selecting people for an interview, picking suspicious people for further investigation). • Wide application - because the cost to replicate AI operations is small, AI is applied on a broader scale than humans • Transparent application without oversight - diversity of views among humans detect and control flowed applications, while AI will be applied transparently avoid such oversight. CC https://www.maxpixel.net/House-Of-Cards-Fragile-Sensitive- Statics-Patience-763246
- 14. Risks • Bias • AI can learn and repeat the inherent bias in the data caused by human behavior (e.g., Book “Weapons of Math Destruction” by Cathy O’Neil.) • Removing bias is hard - For example, an address in a certain neighborhood might act as a proxy for the race; the name might act as a proxy for gender, or name of the degree might act as a proxy for age. • Privacy: AI can infer sensitive data from seemingly harmless information • E.g., detect where you live using accelerometer data 14
- 15. 15 Conclusion • Unlike many hot things, I think AI is here to stay. • It enables many use cases and even potential to redefine health care • Applications are non-trivial and need careful handling • With a strong public health system, we have a real potential to manage data carefully and even convert it to a competitive advantage • AI can be a companion, not a competitor