 |
IMMM,
Interactive Medical Mind Maps
|
Etienne Saliez, 2018-02-23 updated on 2018-11-27 /
Introduction
- Up to now many medical record softwares were not much more than
narrative reports of what did happen at some time in the past in some
specific context. Typically a patient record is a kind of journal
containing some free text notes for every visit.
- A large amount of medical knowledge is already available on Internet,
but the challenge is now to learn a "medical methodology" about how to
use the knowledge.
- The "Problem Oriented Medical Record" management was already
recommended by Lawrence
Weed in 1968, but at that time the computer technology was too
limited.
- Excellent data processing and telecommunication technologies are
today available but not yet understood and used in a meaningful
way.
- Modern medicine require more and more collaborations between
specialized actors, who are in different locations.
- A project of the working group https://www.isfteh.org/working_groups/category/collaborative_care_team_in_open_source
of the ISfTeH, "International Society for Telemedicine and eHealth", a
not for profit organization with members in about 100 countries, https://isft.org/ .
Objectives
- Evolution to a "task oriented model" in order to help to solve the
"health problems" of the patient. At every step what is known and what
to do.
- Visual support of "Natural Intelligence", presented in a very natural
way as graphs with focus on the relations between concepts.
- Interactive Mind synchronization between the members of "Care Teams"
across Internet. The initial implementation is intended for
collaborations between group of students and teachers.
- At the current stage an experimental prototype in order to evaluate
the advantages of graphs.
- A subsidiary objective will also be in the future to prepare tools in
order to analyze data for research purposes.
Approaches
- Evolution to a more effective work methodology:
- Old "Descriptive Model":
- Up to now most systems were based on a kind of "Descriptive
Model". Access to large archives of documentation after the
facts, about what did happen in the past. In fact most current
patient record systems do not much more than a simulation of
the old work methodology of the paper era.
- New "Task Model":
- Today the question is an evolution to a more "Task Oriented
Model", intended to help to solve Health Issues.
- In the management of tasks the relations between concepts
much matter. For example having several Observations it is
important to see to which hypothesis these facts could lead.
These "Problems" need to be clearly formulated and motivated.
- Having identified one or more Problems several Actions will
be proposed and ordered. The relations between Problems and
Action need also to be clearly documented. In both cases N to N
relations.
- A continuous iterative process. Indeed the results of Actions
(investigations and treatments) provide new Observation and the
basis for the next iteration. At any time all these relations
should be maintained up to date.
- At the current stage the goal is only to represent the
logical human reasoning in a visual way, i.e. by mean graphs
showing weighted relations.
- Information management in general:
- Graphs are a very natural way of thinking. The human brain is
made of a very large set of neurons with many connections between
these neurons. They can be navigated in a kind of N dimensional
information space. Software able to simulate these networks become
today available, in an interactive way.
Remark: From a philosophic point of view " awareness " could be
seen as the current " view " on some selected nodes out of a very
large information space ?
----- 
- Exploration of the potential of graph technologies:
- A prototype is in construction and will be set on the web for
discussions with other members of the ISfTeH Telemedicine
community.
- Graph technology is already successful in other domains research
and industries. The current experimental project is intended to
explore the potential for the healthcare domain.
- Medical Model:
- Make the medical reasoning process simply more transparent.
Experienced doctor can work efficiently and quickly in an intuitive
way. Nevertheless weighted explicit representations of the
reasoning are expected to improve the quality of care, particularly
in case of beginners. Today simply more transparency of medical
reasoning. In the future it could help to develop recommendations
and warnings.
- Focus on relation between the essential concepts of Observations
---> Health Issues ---> Actions:
- Observations: just the observed facts without any
judgement.
- Health Issues: concerns requiring attention, as well abnormal
findings, hypothesis, risk factors, as well confirmed
diagnoses.
- Actions: decisions about what to do, either more explorations
or treatments.
- Links to medical knowledge, also presented as graph
navigations.
- At any time, starting from available data up to now, the question
is "what to do next". More at http://www.chos-wg.eu/Models/iterative-care-model.html
.
- Attibutes of nodes and relations:
- "Clinical Modifiers" are essential in patient care practice. More
than "yes" or "no" information, black or white, as usual for
epidemiologic purposes.
- In this project the challenge is to represent these attributes in
3 ways:
- ( A ) Internally as numeric values, in order to allow weigted
pattern recognition.
- ( B ) Rounded to free speech style (as for example "not to be
excluded", "very probable", "obvious", etc...) in order to
remain compatible with the traditional way of human spoken
communication.
- ( C ) As visual graph, where the numeric values may be
converted to color, shape, size, thickness, etc... Visual
representations are expected to provide easier understanding of
the patient situation by the other members of the Care Team of
the patient. Several options are to be experimented in function
of the user preferences.
- Main attributes of items, to be explored:
- Probability:
- Precision, range of probabilities:
- Degree of belief which may depend on several actors.
- Severity:
- Handling of both positive and negative assertions:
- Completeness of information:
- Reliability of the sources:
- Current status of Health Issues: Active, Chronic, History,
Risk factor.
- Expected utility of a recommendation.
- Costs: a multidimensional concept including risk,
disagreement, delay before conclusion, money, availability of
the resource.
- etc...
- Navigation:
- The first screen of a patient record is intended to be a kind of
dash board providing a global overview of the most important Health
Issues. The current situation, as far as possible, according to the
latest available versions of nodes and relations.
- When the patient arrive in a new environment where he is not
known, it is important to have a first screen showing the most
important points about the patient regardless of any
specialism.
- Starting from a given Health Issue, navigation to more details.
Zoom to deeper related details, taking account that a screen can
only show a limited number of information at a time.
- Navigation with Filters on one or more types of nodes as:
- Health Issues,
- Observations,
- Actions
- Anatomical domain,
- Related Medical Knowledge,
- Author. Remark: in a collaborative team some authors could
make different assessment based on the same Observations.
- Encounter time and location, in fact as the traditional paper
records.
- Current Care Plan, scheduled Actions.
- Navigation in the time: at a more advanced stage of software
development, the time could be seen as a 3th dimension of the
graphs.
- etc...
- Access to medical knowledge:
- At lot of medical knowledge sources are available. The challenge
is to find the most relevant information and to make that
information available in a graph minded environment.
- As far as possible make compatibility with existing official
standards, classifications standards and terminologies.
- Education issues:
- On-line exercises on real medical records, but made
anonymous.
- MOOC approach:
- Knowledge:
- Nearly no lectures any more. The student are required to
learn the pure knowledge by mean guided books and
electronic documents.
- Training:
- Training is what most matter and what require much
attention from the teaching staff. Interactive graphs are
expected to be very useful for the training of small groups
of students: "Given all what is know up to now, what to do
next ? "
- Health education in developing regions are an important goal of
the ISFTEH.
- Migration from traditional patient records to
graphs:
- At the initial experimental stage, a few realistic patient
records are created from scratch in a graph way.
- The next step is that graphs could be used as a kind of synthesis
of Global Table of Content of a patient records, while the content
of nodes could still be traditional documents.
- Pattern detection:
- Pattern detection in function of weighted factors.
- Recommendations presented as coming from the experts behind the
system at the same level as other opinions from any care team
Actor.
- For research purposes, graphs can help to discover new relations
to some "patterns", using a "Big Data" approach.
- Technical challenges:
Current experimental prototype:
- Objectives:
- Intended to explore the medical advantages and the technical
feasibility of the graph approach.
- Warnings:
- Experimental and still very basic with bugs. The medical content
of the initial version is stupid and only intended for technical
test purposes on only one patient.
- Not to be used for real patients. Security measures not yet
installed.
- Prototype on Internet:
- IMMM-preliminary-prototype/immm.html
.
- User guide:
- (The initial reactivation of the experimental server may take
about 10 to 20 seconds).
- To view the content of a node = move the mouse on that
object.
- To move to related nodes = left click on the node.
- To create or edit a node = right click..
- To create a relation = drag and drop.
Call for partners in Open Source:
- Healthcare professionals:
- Interested in medical information management and experimental
graph technologies.
- Informatics professionals:
- Interested in the challenges of the development on interactive
graphs.
- Sponsors:
- Sponsors willing to support the social goals of the not for
profit International Society for Telemedicine.
-
Contacts: