The aim of this post is to capture some static points on AI, catch up myself and readers on AI's history, definitions and touchpoints. Also cos the original post was getting super long.
For the latest updates and more interpretative components, you should read
this post.
Where AI was
Here's a
great article summarising the key milestones of AI and predicting where the technology is trending.
This was a particularly seminal moment: "In 2011, IBM's AI system, dubbed "Watson," won a game of Jeopardy against the top two all-time champions" but it wasn't til I played with Google Photos on a walk through Hobart that I realised this amazing potential spanned data in all its forms - not just words, but images, sound, light...
A Rose by any other name
Business Intelligence, Data Mining, Sensing, Pervasiveness, Ubiquity and Intelligent Agent
... essentially, data gathering + pattern recognition + prediction + learning / modifying rules = AI
AI in Health
From the
2015 EPIA conference in Coimbra, the organisers provided an overview of the expansive range of topics that fall under the AI banner, loosely:
Medical methodologies, architectures, environments and systems:
Agents for information retrieval;
AI in Medical Education and Clinical Management;
Wellbeing and lifestyle support;
Interoperability, Security, Pervasiveness, Ubiquity and Cloud Computing in Medicine;
Methodological, philosophical, ethical, and social issues of AI in Medicine;
Pervasive Healthcare Environments;
Software architectures.
Knowledge engineering and Decision Support Systems:
AI-based clinical decision making and Clinical Decision Support Systems;
Automated reasoning, Case-Based Reasoning or Reasoning with medical knowledge;
Business Intelligence in Health Care;
Clinical Data Mining;
Data Streaming;
Diagnostic assistance;
Expert, agent-based or knowledge-based systems;
Medical knowledge engineering;
Pervasive or Real-Time Intelligent Decision Support Systems in Critical Health Care.
Medical Applications and Devices:
Computational intelligence in bio- and clinical medicine;
Electronic Health Records (eHealth);
Image recognition and interpretation;
Intelligent devices and instruments;
Sensor-based applications;
Telemedicine and mHealth solutions;
Ubiquitous devices in the storage, update, and transmission of patient data;
Usability and acceptability.
AI in Healthcare Information Systems:
Autonomous systems to support independent living;
Healthcare System Based on Cloud Computing;
Intelligent Healthcare information systems;
Pervasive Information Systems;
Pervasiveness and Security in Clinical Systems;
Smart homes, hospitals and Intelligent Systems;
Simulation Computer systems.
Essentially: ... data drives decisions in large complex systems, and that's not a bad thing. Dan Sullivan from Strategic Coach, in a
podcast on healthcare and education, observed: "to the degree that hospitals have data scientists, those are the ones that will have great breakthroughs".
What is the impact of that?
The human element is a strength and a weakness - so the use of AI will aim to reduce the downside of that (eg fatigue, mistakes, gaps in knowledge) and augment the benefits (more patient-doctor time, comprehension / observation of complex emotional situations, comfort, human touch and compassion).
See Fastcompany's assessment on
how AI will fit into the hospital workflow:
"Robots won't steal doctors' jobs, says [Venkat] Rajan, but they will spare overworked docs some of the dangerous fatigue that can lead to mistakes. "They're stressed, they’ve got a million different things they're looking at, so [there's] stuff they might have missed."
*Venkat Rajan, Global Director, Visionary Healthcare program, at Frost & Sullivan.
