This presentation summarizes our research on 40 companies from around the world that are leveraging Artificial Intelligence to improve the Healthcare Industry. They are all well-funded, have highly qualified CEOs & Boards, and are poised to achieve their product development milestones.
The I-Square Ventures proprietary rating algorithm indicates that almost all of these companies will receive more funding, and/or be acquired by larger companies.
2. “Artificial intelligence will reach
human levels by around 2029.
Follow that out further to, say,
2045, we will have multiplied the
intelligence, the human biological
machine intelligence of our
civilization a billion-fold.”
—Ray Kurzweil is a Director of Engineering at
Google heading up a team developing machine
intelligence and natural language understanding.
3. “Doctors are realizing that if they
want to make sense of massive
amounts of data, machine
learning is a way of allowing
them to learn from that data.”
—Francesca Dominici, Professor of Biostatistics at
Harvard University’s T.H. Chan School of Public Health
4. First, a quick history of Artificial Intelligence…
ARTIFICIAL
INTELLIGENCE
Academic researchers publish
AI papers based on successful
experiments.
MACHINE
LEARNING
Technology companies begin
to refine user experiences. DEEP
LEARNING
Industry-specific use cases
deliver real business results.
1950’s 1960’s 1970’s 1980’s 1990’s 2000’s 2010’s
5. This presentation summarizes our
research on 40 companies from around
the world that are leveraging Artificial
Intelligence to improve the Healthcare
Industry.
They are all well-funded, have highly
qualified CEOs & Boards, and are poised
to achieve their product development
milestones.
The I2 Ventures proprietary research
algorithm indicates that all of these
companies will receive more funding,
and/or be acquired by larger companies.
6. 1. AiCure: What’s different
USA
Adherence validated
against blood levels
Funded by the National
Institutes of Health
A patented artificial
intelligence foundation
The AiCure solution is differentiated through
science and scale
Technology that can be
accessed by billions of
smartphone users worldwide
7. 1. AiCure: What they do
USA
HIPAA-compliant
facial recognition
Automatic medication
identification
Real-time ingestion
confirmation
Fraud and duplicate
enrollment detection
No video review
is needed
Download application
onto any smartphone
Analytics allow for
pre-emptive intervention
Assistive technology for
greater patient engagement
8. AiCure: Their value
USA
• Clinically-validated artificial intelligence visually confirms medication
ingestion on smartphones to significantly increase medication adherence in
patients with schizophrenia, as measured by drug concentration levels.
• Results of Phase 2 Abbvie study: cumulative adherence, measured by study
drug concentrations above the LLOQ (minimum required therapeutic level),
were higher through 24 weeks for subjects monitored using the AiCure
platform (89.7%) compared with subjects monitored using mDOT (71.9%).
Source: AiCure Website
9. 2. Qventus (Formerly AnalyticsMD): What’s different
Transform Data Into Action that Gets Results
It’s often hard for healthcare teams to know which data to look at and
what decision to make, especially during high pressure situations.
This leads to process failures, inefficiency and missed opportunities to
provide the best possible care.
Qventus solves this problem with technology that closes the gap
between analytics and action. The platform helps hospital teams
respond to their dynamic environments and make better operational
decisions in real-time.
USA
Reduce Costs Increase Revenue Boost Efficiency
Improve Satisfaction Improve Satisfaction
10. Qventus: What they do
How it Works
Qventus uses artificial intelligence and machine learning to translate messy healthcare data into immediate, accurate, and
actionable decisions for clinical teams and staff. The platform accurately predicts issues, recommends course corrections and helps
frontline teams take the right actions, in real-time. Qventus can be used across the hospital including emergency departments,
perioperative units and inpatient wards.
USA
Computers continuously
learn from data and adapt
with minimal programming
Massive amounts of
messy healthcare data
from EMR & other
internal systems
Predict
Identifies issues
before they
occur
Prescribe
Recommends
immediate course-
corrections based on
cost, benefit and
constraints
Persuade
Engages right
team member at
right moment
Higher Utilization
of Services
Increased Patient
Satisfaction
Improved Financial
Performance
Decision
Science
Machine
Learning
Behavior
Science
11. Qventus: Their value
USA
Mercy Hospital Fort Smith (Ark.).
This 336-bed acute care hospital engaged Qventus, a software designed for staffing
optimization, patient flow and resource utilization to improve emergency
department flow. The ED receives around 50,000 visits per year and AnalyticsMD
helped to achieve a 30 percent decrease in "left without being seen" rate.
The hospital also saw a 24 minute reduction in length of stay and a 20 percent
reduction in door-to-door time. The ED was able to accommodate 1,900 additional
visits per year after the workflow update and is projected to report $1.3 million in
additional annual revenue and savings through the increased capacity and reduced
length of stay.
Source: Becker’s Hospital Review
12. 3. Apixio: What’s different
USA
ServicesApplication
Have your own coding team? Want
to bring risk adjustment in-house?
Take control of your risk adjustment
coding with your in-house team
• Reduce time required to code by
80% or more
• Complete QA of all coding
results
• Secure rich project insights on
demand
• Free your team for training and
education
No coding team?
No problem.
Gain confidence in your risk
adjustment without a large coding
team
• Increase accuracy with fewer
resources
• Use our expert coders, armed
with your coding guidelines
• Complete up to 100% of QA
yourself, in just days
• Get regular updates on the data
analysis and chart review
process
or
13. Apixio: What they do
Tackle Both Sides of Risk Adjustment with
HCC Profiler’s Two Solution Suites
USA
Find New Supported Codes
As an application or a service, your organization will
benefit from the accuracy, efficiency and deep insights
that HCC Profiler enables in your first or second pass
coding program efforts.
Confirm Current Codes
Verify that reported HCCs have sufficient evidence for CMS
submission and protect your organization from costly
audit and coding compliance failure with one of the
modules from Apixio’s compliance suite.
Accurate & Fast+
Efficient+
Deep Insights+
Compliance Auditor+
Zero Evidence+
Code Validation+
Coding Compliance
14. Apixio: Their value
USA
• Analyzed over 65 million clinical documents for more than 1.5
million patients enrolled in Medicare Advantage and Commercial
insurance plans.
• Expanded its client base of health plans and provider groups by 43
percent last year.
Source: Market Wired
15. 4. Arterys MICA: What they do
What is inside Arterys MICA?
The Arterys PHI Service™ ensures that patient data stays within the hospital yet is accessible by physicians from anywhere.
Arterys MICA brings powerfully simple solutions to complex diseases.
USA
Artificial intelligence
assistance to accelerate your
workflow
Unsurpassed
Cloud
Supercomputing
What is inside Arterys MICA?
16. Arterys MICA: Their value
USA
• Supports multiple workflows in a variety of imaging modalities, such as magnetic
resonance imaging (MRI) and computed tomography (CT).
• Works seamlessly with applications that integrate AI algorithms into workflows,
allowing clinicians to interact with and edit their output in real time.
• Ensures protected patient data remains within the hospital network, yet it can be
accessed remotely by users with the right credentials.
• Achieves scale while protecting patient data privacy.
• Complies with regional data and privacy laws in the United States, Canada and the
European Union.
Source: Imaging Technology News
17. 5. Atomwise: What they do
Merck is working with Atomwise to explore the frontiers of drug discovery
using artificial intelligence and machine learning.+
Artificial Intelligence for Drug Discovery
Atomwise is the creator of AtomNet, the first Deep Learning technology for novel small molecule discovery, characterized
by its unprecedented speed, accuracy, and diversity.
Research groups that partner with Atomwise gain powerful advantages in the scope, scale, and success rates of their drug
discovery programs.
USA
Working to Fight an Epidemic
Ebola Virus's high lethality have left thousands dead and tens of thousands orphaned since 1976. From a field of thousands of
approved medicines, Atomwise discovered a drug candidate with not previous antiviral application that blocked Ebola infectivity
across virus strains from multiple epidemics
18. Atomwise: Their value
USA
• Replacing test tubes with supercomputers during the drug development
process.
• Uses ML and 3D neural networks that sift through a database of molecular
structures to uncover therapies, helping to discover the effectiveness of
new chemical compounds on diseases and identifying what existing
medications can be repurposed to cure another ailment.
• Within one day they uncovered two new drugs which may significantly
reduce Ebola infectivity.
Source: TechCrunch
19. 6. Ayasdi: What they do
USA
SOLUTIONS OVERVIEW
The Building Blocks of Intelligence
Anti-Money Laundering
Find subtle patterns
hidden across multiple
data types, including
unlabeled data to
significantly reduce
false positives without
increasing the risk
profile.
Clinical Variation
Management
Understanding and
managing clinical
variation is foundational
for any effort to deliver
better care at lower
cost.
Regulatory Risk
The defining feature in
building regulatory risk
models isn’t speed or
even accuracy - it is
transparency. Find out
how to have all three.
Denials
Develop a deeper
understanding of
complex denial groups
to address the
persistent challenges
associated with denied
healthcare claims.
Population Health
What is profit today
may very well be cost in
the future making the
prediction of population
health one of the most
critical capabilities in
healthcare.
20. Ayasdi: Their value
USA
• Conducted a rapid systematic cluster analysis to identify which locations in the region
were most affected by "doses not reached" -- children who had received the first, but not
the follow-up, dose. The resulting visualization identified specific districts within the
region where the "doses not reached" was significantly higher, pointing to the high
fluidity of refugee populations at vaccination sites as the likely cause. Based on this
information, SumAll.org and Humanitarian Tracker enabled vaccination aid organizations
to devise a strategy to more effectively reach this critical population.
• Conducting this kind of detailed analysis with a conventional spreadsheet approach
would have taken days or weeks of tedious examination, looking for outliers, formulating
hypotheses and then testing and validation for over 100 districts. But, with Ayasdi's
topological data analysis approach and the ability to visualize the data quickly, the team
was able to eliminate the cost and time of manual analysis and the process of
formulating and testing multiple hypotheses.
Source: Market Wired
21. 7. Babylon Health: What they do
England
ASK BABYLON
Health advice in a flash
Not feeling yourself? Chat with
Babylon for fast advice on what
to do next if you’re feeling
unwell. GP appointments are
available 24/7 from your phone.
Hand-picked doctors, supported by cutting edge technology, revolutionizing how
healthcare is delivered to you.
CHECK YOUR SYMPTOMS
For the “just in case” moments
Not sure whether you need to see a doctor? Get instant health advice and
recommendations on what to do next – all in less than 60 seconds.
SEE A DOCTOR IN MINUTES 24/7
The most convenient way to see a doctor
Book appointments in seconds to video
chat with top doctors from anywhere in
the world so you can quickly and
conveniently get help and hassle-free
prescriptions.
22. Babylon Health: Their value
England
• Babylon allows doctors to work in tandem with artificial intelligence – so that
medics could focus on treating rather than diagnosing diseases.
• There are 300 million pieces of knowledge that have been collected. This might
be the largest amount of primary care clinical semantic knowledge in the world
that is held by any computer.
• Tests comparing accuracy of triage found that nurses results were accurate in
73.5 per cent of cases, while doctors achieved accuracy levels of 77.5 per cent,
while the computer reached rates of 90.2 per cent. While doctors took an
average of 3 min 12 seconds to make a diagnosis and a nurse took 2 minutes 27
seconds, the computer took 1 minute 7 seconds.
Source: The Telegraph
23. 8. Bay Labs: What they do
USA
Deep learning is a branch of artificial intelligence that follows decades of development from academic research,
industry theory, and experimentation across an increasing image-centric and data-driven world. Medicine is
emerging as one of deep learning’s biggest areas of human impact: it is used for genomic testing and
sequencing, diagnostic image analysis, drug development, and early detection of disease. Doctors in the U.S.
are already using algorithmic support to improve the speed and quality of medical decisions.
At Bay Labs, deep learning has potential to dramatically impact the leading cause of death—cardiovascular
disease. In order to serve the largest number of people, they aim to amplify the benefits of deep learning by
providing high-performance algorithmic capabilities to assist with healthcare challenges on a global scale.
Cardiovascular disease accounts for one in every four deaths
in the United States—approximately 610,000 people every
year, according to the CDC1.
24. Baby Labs: Their value
USA
• Took their technology to Africa to help identify symptoms
of Rheumatic Heart Disease (RHD) in Kenyan school children.
Their software analyzed data derived from an ultrasound to take
a good educated guess as to whether it’s seeing something
consistent with RHD.
• During the trip, medical professionals scanned 1200 children in
four days and were able to spot 48 children with RHD or
congenital heart disease.
Source: The Medical Futurist
25. 9. Butterfly Network: What they do
USA
Vision: Medical imaging should be accessible to everyone on the planet.
Dedicated to democratizing ultrasound.
At the intersection of semiconductor engineering, artificial
intelligence, and the cloud.
26. Butterfly Network: Their value
USA
• World's first Ultrasound-on-a-Chip based imaging system, the
Butterfly iQ® for iPhone.
• Covers 13 clinical applications, the broadest ever for a single
ultrasound transducer.
• Combines almost 10 thousand sensors, 40 times more than systems
costing 100 times as much.
• Ushers in a new era of accessible, high-performance medical
imaging.
Source: Market Wired
27. 10. CloudMedX: What they do
USA
CloudMedx is a health technology company that aims to streamline clinical encounters and optimize outcomes through predictive
analytics. The company utilizes evidence based algorithms, machine learning and natural language processing to generate real-
time clinical insights at all points of care.
Ingest Data
in structured and unstructured
formats from EMRs, insurance
and labs
1
Make Informed Decisions
by knowing patients' clinical and
financial risks and the best
course of action
2
Optimize Workflows
by automation along with
intelligent insights at the right
time and place
3
28. CloudMedX: Their value
USA
• Risk adjustment module automatically pulls charts out of EMRs.
• Uses natural language processing to identify the correct codes and diagnoses
• Information can be provided within 24 hours so care providers can more
properly assess patient risk, improving the course of treatment.
• Payers can use this information to more accurately reflect their patient risk
pool, potentially leading to accurate reimbursement.
Source: Market Wired
29. Time for an AI joke before
your brain gets fried.
OR
Maybe this AI joke will fry your brain?
You decide…
30. Q: How many AI people does it take to
change a lightbulb?
A: At least 81.
Source: artificial-intelligence.blog
31. The problem space group (5):
1. One to define the goal state.
2. One to define the operators.
3. One to describe the universal problem solver.
4. One to hack the production system.
5. One to indicate how it is a model of human lightbulb-
changing behavior
The logical formalism group (16):
1. One to figure out how to describe lightbulb changing
in first-order logic.
2. One to figure out how to describe lightbulb changing
in second-order logic.
3. One to show the adequacy of FOL.
4. One to show the inadequacy of FOL.
5. One to show that lightbulb logic is non-monotonic.
6. One to show that it isn’t non-monotonic.
7. One to show how non-monotonic logic is incorporated
in FOL.
8. One to determine the bindings for the variables.
9. One to show the completeness of the solution.
10. One to show the consistency of the solution.
11. One to show that the two just above are incoherent.
12. One to hack a theorem prover for lightbulb resolution.
13. One to suggest a parallel theory of lightbulb logic
theorem proving.
14. One to show that the parallel theory isn’t complete.
15. One to indicate how it is a description of human
lightbulb changing behaviour.
16. One to call the electrician.
The statistical group (1):
1. One to point out that, in the real world, a lightbulb is
never “on” or “off”, but usually something in between.
Source: artificial-intelligence.blog
32. The planning group (4):
1. One to define STRIPS-style operators for lightbulb
changing.
2. One to show that linear planning is not adequate.
3. One to show that nonlinear planning is adequate.
4. One to show that people don’t plan; they simply react
to lightbulbs.
The robotics group (10):
1. One to build a vision system to recognize the dead
bulb.
2. One to build a vision system to locate a new bulb.
3. One to figure out how to grasp the lightbulb without
breaking it.
4. One to figure out how to make a universal joint that
will permit the hand to rotate 360+ degrees.
5. One to figure out how to make the universal joint go
the other way.
6. One to figure out the arm solutions that will get the
arm to the socket.
Source: artificial-intelligence.blog
7. One to organize the construction teams.
8. One to hack the planning system.
9. One to get Westinghouse to sponsor the research.
10. One to indicate how the robot mimics human motor
behaviour in lightbulb changing.
The knowledge engineering group (6):
1. One to study electricians’ changing lightbulbs.
2. One to arrange for the purchase of the lisp machines.
3. One to assure the customer that this is a hard problem
and that great accomplishments, in theory, will come
from his support of this effort.
4. The same can negotiate the project budget.
5. One to study related research.
6. One to indicate how it is a description of human
lightbulb changing behaviour.
33. The Lisp hackers (14):
1. One to bring up the chaos net.
2. One to order the Chinese food
3. One to adjust the microcode to properly reflect the
group’s political beliefs.
4. One to fix the compiler.
5. One to make incompatible changes to the primitives.
6. One to provide the Coke.
7. One to re-hack the Lisp editor/debugger.
8. One to re-hack the window package.
9. Another to fix the compiler.
10. One to convert the code to the non-upward
compatible Lisp dialect.
11. Another to re-hack the window package properly.
12. One to flame on BUG-LISPM.
13. Another to fix the microcode.
14. One to write the fifteen lines of code required to
change the lightbulb.
Source: artificial-intelligence.blog
The Connectionist Group (6):
1. One to claim that lightbulb changing can only be
achieved through massive parallelism.
2. One to build a backpropagation network to direct the
robot arm.
3. One to assign initial random weights to the connections
in the network.
4. One to train the network by showing it how to change a
lightbulb (training shall consist of 500,000 repeated
epochs).
5. One to tell the media that the network learns “just like
a human does”.
6. One to compare the performance of the resulting
system with that of traditional symbolic approaches
(optional).
34. The Natural Language Group (5):
1. One to collect sample utterances from the lightbulb
domain.
2. One to build an English understanding program for
the lightbulb-changing robot.
3. One to build a speech recognition system.
4. One to tell lightbulb jokes to the robot in between
bulb-changing tasks.
5. One to build a language generation component so
that the robot can make up its own lightbulb jokes.
The Learning Group (4):
1. One to collect twenty light bulbs
2. One to collect twenty “near misses”
3. One to write a concept learning program that learns
to identify lightbulbs
4. One to show that the program found a local
maximum in the space of lightbulb descriptions
Source: artificial-intelligence.blog
The Game-Playing Group (5):
1. One to design a two-player game tree with the robot as
one player and the lightbulb as the other
2. One to write a minimax search algorithm that assumes
optimal play on the part of the lightbulb
3. One to build special-purpose hardware to enable 24-ply
search
4. One to enter the robot in a human lightbulb-changing
tournament
5. One to state categorically that lightbulb changing is “no
longer considered AI”
The Psychological group (5):
1. One to build an apparatus which will time lightbulb
changing performance.
2. One to gather and run subjects.
3. One to mathematically model the behaviour.
4. One to call the expert systems group.
5. One to adjust the resulting system, so that it drops the
right number of bulbs.
35. Now, its time to meet the next 10
companies that are using artificial
intelligence to improve the
Healthcare Industry…
36. 11. Digital Reasoning Systems: What they do
USA
You confront major challenges every day as you work to distill meaning from
the data coursing through your organization.
The ability to ensure compliance, maintain security, and accurately analyze
information quickly is required. You know help is needed, but you don’t have
time for “experiments.” it needs to get done – now.
37. Digital Reasoning Systems: Their value
USA
• Clients using their Synthesys platform have experienced the following
benefits:
• A double ROI in 4 months and the inefficiency of false positive
alerts halved.
• Monitoring of employees expanded from 5% to 100% coverage,
eliminating the risks of sampling while significantly increasing
accuracy.
• By eliminating the use of keywords, precision increased by 250%
while investigation times decreased by as much as 60%.
Source: Digital Reasoning Website
38. 12. Enlitic: What they do
USA
They work with a wide range of partners and data sources to develop state-of-the-art
clinical decision support products.
Their deep learning technology can incorporate a wide range of unstructured medical
data, including radiology and pathology images, laboratory results such as blood tests
and EKGs, genomics, patient histories, and electronic health records (EHRs). This
richness allows higher accuracy and deeper insights for every patient.
Their solutions integrate seamlessly into your existing health system infrastructure. For
example, their radiology solutions communicate with third party image viewers and
archiving systems.
39. Enlitic: Their value
USA
• Can interpret a medical image in milliseconds — up to 10,000 times
faster than the average radiologist.
• In a test against three expert human radiologists working together,
Enlitic’s system was 50% better at classifying malignant tumors and
had a false-negative rate of zero, compared with 7% for the
humans.
Source: The Medical Futurist
40. 13. Flatiron Health: What they do
USA
Accelerate research with the most advanced real-world evidence
platform in oncology.
Designed to meet the most rigorous quality standards for RWE.
Technology-Enabled Abstraction
Patent-pending abstraction process
carefully augments human expertise
with machine learning and other
technology-enabled optimizations.
Data Quality Control
Thorough quality assurance and
complete data provenance for every
data point generated.
Regulatory Collaborations
Active collaborations with the FDA and
NCI to define the requirements and
uses for regulatory grade RWE.
Their pioneering approach pairs clinical oncology experts with innovative
technology to rapidly generate high quality data directly from the EHR.
41. Flatiron Health: Their Value
USA
• More than 265 oncology practices, three large academic research
institutions, and other healthcare organizations, use the company’s
software to manage their daily workflows in one central system.
• Organizes cancer data and empowers care providers in their treatment
of cancer patients.
• Ensures HIPAA compliance.
• Currently analyzes more than 1.6 million cancer patients.
Source: The Oncologist
42. 14. Freenome: What they do
USA
They are technologists
They are programmers, machine learning
experts and computational biologists
developing novel computational methods to
create effective tools for screening,
diagnosis and treatment of age-associated
clinical conditions. Because in the age of
genomics, traditionally biological questions
are giving way to technological solutions.
They are scientists
They are molecular biologists, chemists,
and sequencing experts working
together to ensure the highest quality of
scientific research. They lay the
foundation necessary for next-
generation screening and treatment
methods with academic and pharma
partners.
They equip you with tools
to prevent diseases.
“Our goal is to bring accurate, accessible and non-invasive disease
screenings to doctors to proactively treat cancer and other diseases
at their most manageable stages”
- Gabe Otte, CEO
43. Freenome: Their value
USA
• Had an accuracy rate of more than 95% in detecting
the presence or absence of prostate cancer in 351
samples.
• Had a 97% average accuracy rate in detecting breast,
prostate, lung, and colorectal cancers across four
stages of progression.
• Potential to tell when a case of prostate cancer is
aggressive or low-risk.
Source: Buzzfeed News
44. 15. Gauss Surgical: What they do
Gauss Surgical provides insights
that transform patient care.
They have developed real time blood monitoring solutions to
provide an accurate and objective estimation of blood loss -
helping to optimize transfusion decisions, recognize hemorrhage
status and improve patient outcomes.
USA
45. Gauss Surgical: Their value
USA
• Performed a study in which blood lost onto laparotomy sponges during
surgical cases was measured using a tablet computer programmed with
a unique algorithm modeled after facial recognition technology.
• A significant positive linear correlation was noted between the new
system’s measurements and the rinsed Hb mass
• Provides an accurate measurement of Hb mass on surgical sponges as
compared with manual rinsing measurements and is significantly more
accurate than the gravimetric method.
Source: U.S. National Library of Medicine
46. 16. Ginger.io: What they do
Get started today and receive a free initial consultation. You'll be matched with an emotional support coach, and chatting in
under a minute.
USA
Download the app
Available on iOS and
Android.
Get matched with a Coach
Coaches come in all shapes
and sizes.
Start your free, initial
consultation
Chat with your coach in
under a minute.
Make a plan with you
Coach
Start on your path to
progress
47. Ginger.io: What they do
USA
In-the-moment Support
Receive replies to your messages just
in minutes, not hours. Chat with your
coach wherever you are, whenever
you want.
Right Care, Right Time
They offer individualized care and
proven strategies to meet you, where
you are today and in the future.
Confidential & Secure
Everything you share is kept
completely confidential and treated
as HIPAA-protected health
information.
Full Spectrum of Care
Work with Ginger.io coaches to
achieve goals, therapists to deepen
self-awareness or psychiatrists for
medication management and
treatment.
Personalized Plan
They aim to find the best type of care
for you. Using tools backed by
machine learning, we’ll tailor your
care to your unique needs and goals.
Measurement Matters
As a licensed, nationwide medical
provider, they practice measurement-
based care. Using advanced data
science, they help care teams track
your progress.
48. Ginger.io: Their value
USA
In employer settings, reports show:
• Installation of the app for up to one-quarter of the
employee population.
• Reduction of up to one-third in the relative cost of care.
• More than half showing at least a partial response (30%
drop in symptom severity) over eight weeks.
Source: World Economic Forum
49. 17. H2O.ai: What they do
USA
The World's Best Machine Learning Platform
H2O.ai makes machine learning accessible and allows business users to extract insights from data,
without needing expertise in deploying or tuning machine learning models.
DATA SCIENTISTS
Create and train machine learning
models in several environments
ENGINEERS
Operationalize data products, quickly
design and deploy AI apps
EXECUTIVES
Analyze datasets, extract insights and
tackle business problems
50. H2O.ai: Their value
USA
• Constructs risk scores for patients and aids in discharge planning.
• Enables better utilization of nurses and case managers, prioritizing
patients according to risk and complexity of the case.
• Lowered Carolina's Healthcare System readmission rate from 21 percent
to 14 percent.
Source: AI Business
51. 18. Health Fidelity: What they do
USA
They use their proven model to understand the key business characteristics of their clients, identify and assess the opportunity for risk
optimization, determine network readiness, and finally implement their solution to measure the improvement and delivered ROI.
HF360 BLUEPRINT
DISCOVERY & DATA
COLLECTION
Review current
processes
FINANCIAL
OPPORTUNITY
Search for revenue
opportunity
CLINICAL DATA
LANDSCAPE
Document data
availability
NETWORK
READINESS
Understand organization
readiness
360o
BLUEPRINT
Develop your HF360
blueprint
(1) (2) (3) (4) (5)
Evaluate current
state
ACTIVITY
Compilation of all
current activities and
processes
OUTPUT
Analyze historical
data
ACTIVITY
Financial opportunities
OUTPUT
Discuss chart retrieval
efforts/vendors
ACTIVITY
Inventory of all data
sources
OUTPUT
Reflect on internal
readiness
ACTIVITY
Gaps in organizational
alignment
OUTPUT
Identify methods of
improvement
ACTIVITY
Roadmap to
improvement
OUTPUT
52. Health Fidelity: Their value
USA
• Partnered with UPMC – the largest nongovernmental employer in
Pennsylvania.
• UPMC implemented HCC Scout to help increase accurate RAF capture in
compliance with CMS Risk Adjustment guidelines while improving overall
coding efficiency.
• Resulted in remarkable projected return on investment with a study
conducted on PY2012 members, including the following incremental results:
Source: Health Fidelity Website
53. 19. iCarbonX: What they do
Optimize your life data to learn more about yourself
Complete your digital health records by storing genomic data, medical
data and daily health data, and understand your health conditions
whenever and wherever you want.
Based on holographic health data, they provide individualized health
analysis and predictions.
Together with their partners, they plan to observe, study, guide and help
people achieve better health through their lifetime.
Multiple companies are collaborating with them, including research
institutions, pharmaceutical companies, medical examination centers,
hospitals, insurance companies and health management companies.
China
54. iCarbonX: Their value
China
• Integrates clinical data and medical analytics.
• Supported by a database with over ten dimensions of bioinformatic data,
combining genetics, molecular profile, phenotype and behavior data and
other indexes as well as their time serial interactions.
• 10 Billion data points drive their predictive algorithms.
Source: IoT M2M Council
55. 20. Imagen Technologies: What they do
USA
Imagen Technologies is a medical technology company that develops
and offers devices for the prevention of diagnostic errors.
The company combines deep-learning and computer vision technologies
to create automated diagnosis systems that are able to instantly provide
accurate detection and localization of key diagnostic criteria within
radiology images.
56. • Reads X-rays to eliminate the 30% error rate conducted by
radiologists.
• Strives to create “a world without diagnostic errors.”
• Acquired over 12 billion raw images through a major cross-hospital
partnership deal to improve their models on rich, high-quality data.
USA
Imagen Technologies: Their value
Source: NYU Center for Data Science
57. This time let’s have an AI joke
break that does not fry your
brain….
58.
59. 21. Jvion: What they do
The Cognitive Clinical Success Machine more precisely and comprehensively foresees risk and is the only
machine of its kind to deliver the recommended actions that improve outcomes.
USA
Hospitals
Helping hospitals
reduce target illnesses,
prevent patient
deterioration, reduce
the cost of care, and
improve community
health efforts.
IDNs
INTEGRATED DELIVERY
NETWORKS
Helping IDNs manage
utilization, network
performance, and
financial risk while
providing the patient-level
insights that drive
improvements in health
outcomes and care.
ACOs
ACCOUNTABLE CARE
ORGANIZATIONS
Enabling ACOs to better
serve the covered
population, determine
market opportunities, track
performance against Value
Based Purchasing
measures, and manage
Medicare spend per
beneficiary.
PACs
POST-ACUTE CARE
FACILITIES
Reducing complications,
infections, and
readmissions for patients
regardless of care setting.
60. Jvion: Their value
USA
• Increase in core/Surgical Care Improvement Project (SCIP) scores from the
50th percentile to 100th percentile.
• 73 percentage point improvement in Hospital Consumer Assessment of
Healthcare Providers and Systems (HCAHPS) physician communication
scores.
• Projected 30% drop in instances of catheter associated urinary tract
infections (CAUTIs).
Source: Jvion Website
61. 22. Lucina Health (formerly CerviLenz): What they do
USA
• Improve early identification and
intervention with predictive risk scoring and
alerts for patients who need immediate
triage
• Increase impact with more time for nurses
to support high risk patients along with
targeted education for all patients
• Improve efficiency and allocate resources
more effectively
• Improve and customize care with a direct
channel to collect patient data collection
and deliver personalized pregnancy plans
• Increase communication between patient
and care manager with "click to call"
• Educate pregnant moms with quick, on-
demand access to important information
specific to her needs
62. Lucina Health: Their value
USA
• A measurement of 30 mm was equivalent to the fetal
fibronectin test (Hologic, Inc.) in predicting preterm birth in
women with signs and symptoms of preterm labor.
• Clinicians gain a quantifiable measurement and can make
confident decisions faster.
Source: MedGadget
63. 23. Lumiata: What they do
USA
ACOs
Code more accurately, and capture
and report your Quality Measures
earlier, to operationalize value from
risk-sharing agreements.
Health Plans
Improve your top and bottom lines
with precise disease predictions that
pinpoint your highest risk members
for risk management.
Partners
Integrate the Lumiata Matrix Suite to help your
clients contain costs, minimize risk, improve
quality and deliver more targeted care.
64. Lumiata: Their value
USA
• Services over 1 million patients.
• Provided a 50 percent increase in
revenue for its clients.
Source: TechEmergence
65. 24. Medalogix: What they do
USA
TOUCH
Increase care with your at-risk
patients using automated
touchpoints to avoid unnecessary
readmissions and provide value-
based care.
BRIDGE
Bridge the right patients to hospice
at the right time by identifying
patients at-risk and increasing
clinical collaboration.
NURTURE
Nurture former patients who are at-
risk for decline by streamlining your
post discharge calling program.
66. Medalogix: Their value
USA
• Increased review efficiency and accuracy.
• Increased confidence in appropriate patient venue.
• Increased care coordination and accountability.
• Increased total hospice days by nearly 300 percent.
Source: Medalogix Website
67. 25. Medasense: What they do
Novel, patented technology
Addresses wide clinical needs
Flagship product PMD-200: CE marked
Clinically validated
Proved to be superior to existing measures
Led by highly qualified scientific team
Israel
68. Medasense: Their value
Israel
• Changes proportionately with patients’ response to various clinical and
experimental noxious stimuli and discriminates noxious from non-noxious
stimuli with high sensitivity and specificity.
• Responds progressively to increasing stimuli intensity and is appropriately
blunted by analgesic administration.
• Superior to other compared measures and appears to accurately
characterize nociception during general anesthesia.
Source: Medasense Website
69. 26. NextHealth Technologies: What they do
WHO
Predictive analytics uncover the
populations that can most impact
the targeted use case based on a
rich data set. The platform
pinpoints avoidable and
impractical behavior vs. just
finding the highest risk patients.
This approach means you can
focus resources on fewer total
members while concurrently
driving better (and bigger)
outcomes.
USA
WHAT
Prescriptive analytics define the
best outreach strategy based on
the unique attributes of each
population and the available
programs - the “next best action”.
The consumer engagement
element in the platform manages
the messaging and execution of
interventions across all channels,
including call center, direct mail,
email, SMS texts, and websites.
HOW WELL
On the platform dashboard, dive
into the details of each campaign
to gauge what worked for which
members, how well your delivery
channels performed, and where
your ROI was greatest. Machine
learning optimizes subsequent
campaigns to deliver only those
interventions that worked well to
the populations most likely to
respond to them, saving you time
and money - and maximizing
outcomes.
70. NextHealth Technologies: Their Value
USA
• Saves tremendous time and money by quickly scaling to
multiple use cases.
• Reduce costs in ER overuse then pivot to reducing out-of-
network visits without having to rework your models or
data sets.
Source: NextHealth Website
71. 27. Numerate: What they do
USA
THE POWER
Numerate has mastered the
conflicting demands of complex
signal processing by applying
advanced, proprietary AI, evolved
to use all available data – literally
everything the industry currently
knows, against multiple,
simultaneous drug design
objectives. They model real
biology and apply our models at a
massive scale, sifting through
noise to follow signals that elude
other approaches.
THE PLATFORM
Numerate has created an
algorithm-centric process that
drives all preclinical decisions,
from hit identification through
the nomination of a clinical
candidate. Their platform is
battle-tested and refined by a
decade of experience tackling the
industry’s hardest problems. It
goes beyond modeling protein
binding to impact the PK, ADME,
and toxicity challenges of
discovery.
THE PROMISE
Numerate is applying the
platform to unlock valuable
biology not easily translated by
the traditional approaches. Their
focus is on building a valuable
pipeline– on our own and through
innovative collaborations with
academic, government and
industry partners. After all, the
ultimate promise of AI-guided
drug design is not just a faster,
cheaper discovery process, but
the delivery of better treatments
for human disease.
72. USA
Numerate: Their value
• Scales up to 10,000 nodes in the US-East zone of EC2 (Elastic Cloud
Compute).
• Tackles large-scale computational problems cost effectively.
• Cost 2.7¢ per core-hour, or $270 per hour for a 10,000-core cluster
– just one-third of the on-demand cost.
Source: Numerate Website
73. 28. Nuritas: What they do
Republic of
Ireland
Targeted
approach
Their targeted or on-demand
discovery process enables true
white space innovation, leading
to the creation of new markets
and solutions to both old and new
problems.
High speed
prediction
The speed of our prediction
process is unrivalled. By
leveraging the power of machine
learning, the speed and accuracy
of our predictions improves with
each peptide they discover.
100% naturally sourced
ingredients
Their bioactive peptides are
sourced from food and thus are
natural with extraordinary health
benefits.
Cost effective
innovation
Driven by the speed and accuracy
of our platform, Nuritas™
accomplishes in months what
previously took years, saving tens
of millions in research and
development costs.
Product
differentiation
Each of our bioactive peptides is
unique with powerful therapeutic
benefits. The peptides that they
discover are the first of their kind
and patented.
Sustainable
These molecules are found in
everyday foodstuffs, meaning that
they have the ability to deliver
sustainable, cost-effective and
scalable natural ingredients to
many industries.
74. Nuritas: What they do
Republic of
Ireland
• Discovered peptides up to 10 times faster than traditional research
techniques and at a fraction of the cost.
• This peptide prevents people who are pre-diabetic progressing to full
blown Type 2 diabetes.
Source: The Irish Times
75. 29. Pathway Genomics: What they do
1. Collect Sample
Collect your DNA at home with
our easy-to-use cheek swab.
2. Analyze DNA
They analyze your sample and build
your genetic profile.
3. Get Your Results
Read your results and begin your
journey.
USA
76. Pathway Genomics: Their value
USA
• Neither invasive nor time-consuming.
• Predicts estimated lifetime risk for 120 diseases including asthma, gallstones,
rheumatoid arthritis, and scoliosis.
• Predicts how you'll react to everyday prescription drugs.
• Delivers personalized wellness insights to the user.
Source: Men's Journal; IBM
77. 30. PhysIQ: What they do
PhysIQ applies artificial intelligence and machine learning to the data streaming from wearable or implantable devices to learn
one’s unique vital sign patterns and then detect subtle deviations that may indicate a change in health. The result is a data- and
device-agnostic platform that can be applied across multiple clinical use cases, hardware platforms, and patient populations to
generate unprecedented personalized insight.
For health systems, payers, and
post-acute
providers…continuous real-
world personalized patient
insight
For pharmaceutical and med
device clinical
trials…unprecedented data
collection for clinical product
differentiation
For wearable and implantable device
companies…licensable and scalable
Personalized Physiology Analytics
USA
78. PhysIQ: Their value
USA
• 150,000 hours of data captured through the Vital Connect device.
• Demonstrates patient compliance rates in excess of 80%.
• Automated data cleansing.
• Delivers a high-powered triage tool.
Source: PhysIQ Website
81. 31. Sentrian: What they do
Identify:
By analyzing a client’s claims and/or clinical data, Sentrian’s risk assessment engine identifies patients who are at the greatest risk
of hospital admissions (and would benefit most from remote monitoring and analytics), while eliminating those that can’t be
impacted.
Monitor:
Once the right patients are identified, Sentrian’s clinical team uses our advanced analytics engine to parse them by disease state,
co-morbidities, and other clinical indicators to determine the appropriate mix of biometric sensors for their condition. Sentrian’s
turnkey fulfillment solution manages the sensor delivery and activation process end-to-end for clients.
Analyze:
Their cloud-based Clinician-Directed Machine Learning engine continually runs the sensor data through a disease deterioration
model that looks for personalized patterns predictive of hospitalization.
Act:
Care managers and clinicians are only notified when a patient’s risk status has changed and requires their attention. These
notifications are designed to provide care managers and clinicians with enough time to intervene with safer, lower cost, treatment
options that keep patients healthy and happy at home.
USA
82. Sentrian: Their value
USA
• Represents an exponential increase in the pace with which they create
and deploy remote disease deterioration models.
• "Current remote patient monitoring only gives us hours to act, the
Sentrian product has the potential to give us days or weeks to act". -
Chief Medical Officer of a Medicare Advantage Health Plan
Source: Sentrian Website
83. 32. Wellframe: What they do
USA
Innovative Mobile
Platform
Evidence Based
Clinical Portfolio
Advanced Analytics
& Impact Evaluation
Strategy &
Supplementary Services
84. Wellframe: Their value
USA
80%
Sustained weekly
care plan engagement
5x
Increase in care
management efficiency
$3,000+
Average per patient
medical cost saving
Increased member to care manager
communications significantly
decreases time per contact and
associated costs for care.
Source: Wellframe Website
85. 33. WellTok: What they do
USA
Health Plans
Support bending
healthcare trends
across all lines of
business
Employers
Provide a unified, highly
personalized and
rewarding employee
experience
Government Programs
Engage hard-to-reach
Medicare and Medicaid
beneficiaries
Providers
Improve employee health
and provide a community
health solution
86. WellTok: Their value
USA
Highlights from WellTok's CaféWell survey of Colorado DPA’s employees include:
• Employees (60 percent) have a better perception of the State of Colorado as an
employer with the availability of CaféWell;
• Nearly three quarters (71 percent) of respondents feel their health has improved
since joining the program, including:
•feeling less stressed (41 percent),
•losing weight (38 percent),
•feeling happier (32 percent) and
•having more energy (28 percent); and
• Employees also feel more educated about their health (77 percent) and their
benefits (38 percent).
Source: WellTok Website
87. 34. Your.MD: What they do
Hello, I’m Your.MD Let
me help improve your
health
Your.MD uses Artificial Intelligence
to guide you to the best and most
relevant health information for
you to get better and stay healthy.
You can use Your.MD whenever,
wherever and always for free.
Clear health info when
you need it most
When you need medical
information you want the best
answers and clear information,
not a million links from a search
engine! Their Health Library has
the most accurate and simplest to
understand medical information.
Find doctors, pharmacies,
chronic illness solutions
and more on OneStop
Health™
With OneStop Health™ you get the
best healthcare services for your
problem. All the services are
continuously checked by their
team for your peace of mind.
UK
88. Your.MD: Their value
UK
8x
Better conversion rates than advertising.
20%
Monthly increase of partners joining the marketplace.
Source: Your.MD Website
89. 35. Zebra Medical Vision: What they do
EMPOWER RADIOLOGY
They work with millions of imaging and
correlated clinical records to create high-
performance algorithms that automatically
detect medical conditions faster, for numerous
findings in parallel. Over the next few years they
expect to release dozens of automated findings
and insights to help radiologists provide more
comprehensive, accurate outcomes – faster,
without compromising quality of care. TRANSFORM PATIENT CARE
Identify high risk patients earlier
Facilitate disease prevention programs
Optimize work list prioritization for urgent
cases
Manage population health programs to
reduce cost of care
Israel
90. Israel
Zebra Medical Vision: Their value
• AI1 offers all its current and future algorithms to healthcare providers
globally for only $1 dollar per scan.
• Automatic detection of liver, lung, cardiovascular, and bone disease all while
new capabilities are constantly being released.
Source: Forbes; Medical Product Outsourcing
91. 36. Zephyr Health: What they do
Connected global health data
Comprehensive physician,
institution and treatment data
for every major disease area
Machine learning-based data
linkage
Proprietary algorithms link
disparate data sources delivering
actionable insights
Faster time to insights
Cloud-based software designed for
business users to easily extract
insights
USA
92. Zephyr Health: Their value
• Provides a clearer picture of the market landscape, including key
customers (patients, prescribers, experts, payers, and other
influencers) and competitor activity.
• Better product positioning and market access.
• More effective segmentation of customer groups and allocation of
resource.
• Deeper and more meaningful engagement with all customers.
USA
Source: Zephyr Health Website
93. 37. Benevolent.ai: What they do
Enabling 21st Century scientific innovation
BenevolentAI is the world leader in the development and application of artificial
intelligence for scientific innovation. They are the largest private AI company in
Europe and one of the top five private AI companies in the world.
UK
Discovering new medicines
They began with a focus on bioscience – increasing the understanding of human
biology by ingesting the entire compendium of information on human health and
biological systems. They are applying this knowledge to pharmaceutical
development to bring better medicines to patients faster.
94. UK
Benevolent.ai: Their value
• Could speed up the process of finding new drug candidates for
ALS therapy.
• Predicts blood-brain barrier penetration and other
physicochemical properties of compounds.
• Trains itself on the data accumulated and takes feedback from
scientists to increase system accuracy.
Source: Labiotech
96. Israel
DreaMed Diabetes: Their value
• Continuously monitors glucose levels, and defines precisely
when and how to adjust insulin levels.
• Enables improvement in metabolic control and the quality of
life for patients with diabetes.
Source: Jvion Website
97. 39. Picwell: What they do
A SaaS-based solution to enhance
the entire benefits lifecycle
From open enrollment to ongoing population management, their product offers a broad range
of applications for commercial insurance, Medicare, and other benefits:
+ Plan selection and decision support
+ Consumerism and HDHP adoption
+ Consumer engagement
+ Risk management
+ Local market optimization and more
USA
98. USA
Picwell: Their value
• Reduced time spent building new models.
• Increased confidence in model accuracy.
• Highly Reliable API.
Source: PromptWorks
99. 40. Roam Insight: What they do
COMMERCIAL EXCELLENCE
Through machine learning and
advanced analytics, Roam
applications provide context to vast
amounts of healthcare data.
Whether it's planning a new
product launch or working to
improve the performance of an in-
market treatment, Roam helps life
sciences companies solve their
most critical business challenges by
making complex multidimensional
data analysis accessible to all.
VALUE BASED CARE
Success in the healthcare industry
today depends on patient outcomes
and high-value care. That's why the
Roam Health Knowledge Graph creates
a rich, unified mapping of internal
company data, hospital data, and
relevant external data to facilitate
patient-level insights and drive
improvements in care.
REAL-WORLD EVIDENCE
Real-World Evidence (RWE) has the
potential to transform how
biopharmaceutical and medical device
companies manage healthcare data. By
algorithmically connecting, aggregating,
and reconciling this data, Roam's
platform leverages RWE to inform clinical
development, patient outcomes, and
risk-sharing agreements.
USA
100. USA
Roam Insight: Their value
• Enables customers to confidently navigate outcome-based compensation.
• Organizes data collected from disparate sources, enriches data with machine
learning-derived insights, augments that data with human judgment and
presents recommendations in a compelling visual story.
• Links treatments from drugs and devices to outcomes.
• Accelerates FDA approval and enable sales and marketing teams to target
doctors and treatment centers most capable of demonstrating the desired
outcomes.
Source: Roam Insight Website
101. That was the last of our 40 companies using artificial intelligence to
improve the Healthcare Industry.
I am sure there are other great companies that our filtering
algorithm missed this time round. Think of it as just another
example of how AI algorithms need to be continuously refined.
102. Thank you for enjoying this presentation.
I2 Ventures focuses on helping innovative companies expand their
businesses globally. For a free download of our methodology, please visit
www.isquare.ventures