A new era of cognitive computing
Rob Smith
Program Director
Software CTO team, Europe
[email protected]
@laserfunrob
Watson is ushering in a new era of computing
Tabulating
Systems Era
Programmable
Systems Era
Cognitive
Systems Era
1900
1950
2011
2014 International Business Machines Corporation
Cognitive systems expand the problems we can address
Programmatic Systems
Leverage traditional data sources
Follow pre-defined rules (programs)
Provide the same output to all users
Cognitive Systems
Are taught, not programmed.
Learn and improve based on experience
Interpret sensory and non-traditional
data
Relate to each of us as individuals
Allow us to expand and scale our own
thinking
Watsons intelligence and capabilities are expanding
Understands natural
language
Explores
Reasons
Generates and
evaluates hypotheses
Adapts and learns
Visualizes
2014 International Business Machines Corporation
Watson enables three classes of cognitive services
ASK
DISCOVER
DECIDE
User has a question
and answer
requirement, with
questions posed in
natural language
Seek answers and
insights from a defined
data repository (i.e.
corpus) comprised
largely of unstructured
data
Provide transparency
and supporting
evidence for
confidence weighted
responses to questions
and queries
2014 International Business Machines Corporation
Cognitive computing will transform Healthcare
Care is determined by a
proactive plan to meet health
needs, with or without visits
We measure our quality with
analytics and make rapid
changes to improve it
Care is standardized according
to evidence based guidelines
and advance cognitive systems
Engage Patients
Improve Outcomes
Control Costs in uk,
think about referrals
2014 International Business Machines Corporation
Watson Oncology: helping oncologists treat cancer patients
Attacking the cause
of one in four deaths
IBM Watson
Oncology
Built with Memorial Sloan Kettering
2014 International Business Machines Corporation
Business problem:
Need better individualized cancer treatment plans
Solution:
Suggestions to help inform oncologists decisions
based on 600K+ pieces of evidence and 2M pages of
text from 42 publications
Analyzes patient data against thousands of historical
cases and trained through 5000+ Memorial SloanKettering MD and analyst hours
Evolves with the fast-changing field
Watson Oncology is a Cognitive System that differs from
conventional programmatic systems
2
It is taught by example, not programmed. This gives it the ability to learn and adapt
as new information is presented and outcomes and actions are taken. Ground truth
is communicated through training / test cases as well as interpretation of those
cases by physicians
It generates and evaluates a set of hypothesis appropriate to the case being
presented rather than being confined to only a set of rules, or simple decision tree
logic (e.g. if this, not that), giving it a much wider range of alternatives to consider
It delivers a set of probabilistic treatment responses, which are confidence
weighted into categories
It provides the underlying rationale for its recommendations and points to evidence
that support and refute those recommendations
For physicians, Watson Oncology provides ranked treatment options
with supporting evidence rather than just listing valid guideline choices
2014 International Business Machines Corporation
Watson Oncology helps medical oncologists and their care
teams address these challenges
2
Use those attributes to
find candidate treatment
options as determined by
consulting NCCN
Guidelines
Patient Case
61 y/o woman s/p
mastectomy is here to
discuss treatment options
for a recently diagnosed
4.2 cm grade 2 infiltrating
ductal carcinoma
Guidelines
3
Search a corpus of
evidence data to find
supporting evidence for
each option
Evidence
Candidate Treatment
Options
Key Case
Attributes
Watson Oncology
Supporting
Evidence
1
Extract key attributes
from a patients case
2014 International Business Machines Corporation
Prioritized Treatment
Options
+
Evidence Profile
Inclusion / exclusion
criteria
Co morbidities
Contraindications
FDA risk factors
MSK preferred
treatments
Other guidelines
Published literature studies, reports, opinions
from Text Books,
Journals, Manuals, etc.
4
Use Watsons analytic
algorithms to prioritize
treatment options based
on best evidence.
Watson Genomics: Bridging between sequencing and personalized
medicine.
Ampli-seq panel
Gain consent
Mutation(s) identified
Whole Genome Seq
Gene expression,
Tumour Heterogeneity analysis
RNA seq
In absence of genomic
information the patient
would be a candidate
for chemotherapy or
radiation
Chemotherapy or
Radiotherapy
Case specific analysis
of mutation patterns,
copy number analysis,
differential gene
expression, DNA repair
defects, SNVs, WGSs
Cross reference against
a system that can
identify confidence
levels against treatment
regimen
No mutation
Pharma
Cancer Gene
Specific Genome
Catalogue of
Cogenomics
corpus
databases
annotated tumour
variants
Output is a highly
personalised set of
targetted treatments
Ranked list of
actionable targets
with confidence level
List of drug targets
Drug option
summarisation
On demand
evidence
retrieval
Drug
sensitivity
prediction
analysis
Patient
molecular
profile
analysis
IBM recently joined the
NYGC Consortium as the
sole Founding Technology
Member
Together IBM and NYGC will
work on applying core skills
and cognitive technologies to
rapidly ramp sequencing to
Oncologists
Bob Darnell, President, NYGC
http://www.nygenome.org/
Watson facilitating medical school problem-based learning methods
Cognitive systems in
a classroom-based
setting
WatsonPaths and
Watson EMR Assistant
Research Projects
Business problem:
Give medical students and doctors easier insight into
data to inform their diagnoses and decisions
Solution:
Intuitive, new user interface to Watsons power
revealing chains of evidence to support clinical
reasoning
Analysis of whole EMRs to extract and visually present
summarized knowledge with semantic understanding
of context
IBM Watson Engagement Advisor
What it does:
How it does it:
Transforms client engagement by
knowing, engaging and empowering clients
where they are
Answers questions and guides
users through processes with plainEnglish dialogue
Develops client relationships by reaching
out to clients who do not leverage traditional
channels
Leverages natural language to
interact with users and build
knowledge and expertise
Empowers consumers and contact center
agents to take informed action with
confidence
Utilizes evidence evaluation and
learning to provide informed and
effective responses to users
13
Re-imagining medicine in 2020
Watson is changing the way medicine is:
Researched
Connecting clinicians & researchers at
MD Anderson
Taught
Students learning from Watson at the
Cleveland Clinic
Practiced
Memorial Sloan Kettering, Community
Cancer Care Centers
Paid for
WellPoint & Utilization Management
2014 International Business Machines Corporation
Delivering evidenced
based personalized
medicine
14
ibmwatson.com
2014 International Business Machines Corporation
facebook.com/ibmwatson
@ibmwatson
15
