SIDM Tech Update: HIT for Diagnosis

Amos Cahan acahan at US.IBM.COM
Wed Mar 8 16:04:43 UTC 2017


Dear Dr. Winslow,

In a paper published today in JMIR (http://www.jmir.org/2017/3/e54/), we
review the shortcomings of current computer-aided diagnosis support systems
and suggest a framework for the development of "next generation" DSS.

The approach we advocate for involves structured representation of
knowledge, sharing of automatically-deidentified clinical data and a
knowledge base associating patient presentation patterns with the diagnosis
assigned to them. It further uses disease base rate to evaluate the
probability of a diagnosis given a compatible clinical presentation.  We
believe this approach will enable recruiting the collective clinical
experience of physicians to help diagnose new patients.

I thought this might be of interest to community members.

Best,

Amos

Amos Cahan, MD
Research scientist, Clinical Informatics
IBM T. J. Watson Research Center
1101 Kitchawan Road, Route 134
Yorktown Heights, NY 10598
1.914.945.2590
acahan at us.ibm.com





From:	Edward Winslow <edbjwinslow at GMAIL.COM>
To:	IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG
Date:	02/16/2017 03:56 PM
Subject:	[IMPROVEDX] SIDM Tech Update: HIT for Diagnosis



         Health Information Technology (HIT) Tools for Diagnosis:





In November, we posted a call for IT or ideas you have for improving
diagnosis. These tools are often referred to as either:


        ·      Medical Diagnostic Decision Support Systems (MDDSSs)


        ·      Diagnostic Clinical Decision Support Tools (DCDSTs)





Laura Zwaan in her article in our journal (Diagnosis; 2014, 1, 139-141)
suggested that, properly devised and used, health IT should help improve
the diagnostic process and that information about such tools should be
widely disseminated.





Here are the 4 tools submitted in response; all are commercially available,
and listed by date of introduction. Each of the submissions were from
representatives of the organization that created the tools.





DXplain (http://www.mghlcs.org/projects/dxplain) - Submitted by Edward P
Hoffer, MD (ehoffer at gmail.com):

DXplain, a MDDSS developed and supported by the Lab of Computer Science at
the Massachusetts General Hospital, was first released in 1987, with 500
diseases in the database (db). Over the ensuing decades it has evolved and
expanded to its current database of over 2500 diseases and some 6000
“findings,” which include symptoms, physical findings, demographics,
laboratory test results and imaging findings. Its most common use allows
the student or clinician to enter information about a patient and receive a
rank-ordered list of diseases that can explain the findings entered.
The system was designed to be transparent, and will explain why a disease
is on the list. It can guide the user to enter additional findings that
will refine the disease list. Uniquely, it allows the entry of negatives –
ie, absence of findings, which often help exclude diseases that would
otherwise be suggested. The user can interrogate the db to get disease
descriptions and diseases associated with any finding. Each disease has a
hand-selected set of good references and also facilitates PubMed and Google
searches. Many of the findings have associated descriptive text to describe
their meaning.
DXplain is available world-wide over the Internet by annual subscription
and is used by hospital, group practices and medical schools. Full details
are available at dxplain.net



Isabel (http://www.isabelhealthcare.com/ ) – Submitted by Jason Maude (
Jason.maude at isabelhealthcare.com ):

The Isabel DDx Generator is powered by statistical natural language
processing software applied to a database of disease presentations. The
system covers 10,000 conditions, all ages and all specialty areas of
medicine. Besides clinical features, Isabel utilizes additional algorithms
to only present those results relevant to the patient’s age, gender and
geographic region.
The Isabel Symptom Checker uses the medically validated Isabel system. This
enables patients to enter multiple symptoms in everyday language. Suggested
possible diagnoses are linked to further knowledge. A ‘Where now?’ (triage)
feature is also provided to help patients decide where to seek care within
a health system.

We subsequently learned that any healthcare professional can take a 10-day
free trial of the DDx Generator. There is also a “free patient version” –
“Symptom Checker” available

VisualDx (https://www.visualdx.com/)– Submitted by Art Papier, MD (
apapier at visualdx.com)
In 2016, VisualDx transformed from a dermatology app to diagnostic support
for general medicine.
VisualDx is a web-based diagnostic decision support system improving
diagnostic accuracy at the point of care. Through its medical image library
and its search technology, health care professionals can build a custom
differential. VisualDx introduced the “Sympticon,” a graphical method to
communicate the pattern of symptoms. This process encourages a logical
method of clinical reasoning and a visual approach to review, compare, and
understand variations of disease.
Available on the desktop, as a native IOS or Android mobile app, and
through the electronic health record, VisualDx fits into the clinicians’
workflow. VisualDx uses FHIR inside of Cerner Millennium EHR, and can
integrate inside of Epic and other EHR’s.
To understand the approach, please view sample differentials at the link
above
30-day free trials are available to students, residents, and any healthcare
professional.

Xebra (https://www.physiciancognition.com/) Submitted by Vipindas Chengat,
MD (syncopesystem at gmail.com)
Physician Cognition’s decision-support and clinical education tools are
driven by its core technology, Xebra™, an engine designed to mimic the
cognitive strategies of a clinician. Xebra combines a variety of
statistical and data analysis strategies to replicate the problem-solving
approach taught by many medical schools. Xebra can process any number and
combination of symptoms, signs, labs, medications, and past medical history
and present a differential list in the order of probability, but its goal
is not to offer a definitive diagnosis Rather it is to support the
clinician’s decision-making process with a checklist approach. Part of
Xebra’s core capability is its workup guide, which can help the clinician
refine the diagnosis through an evidence-based step-wise process.
Physician Cognition’s products are targeted toward clinician students,
residents, nurse practitioners and physician assistants to facilitate
further learning and critical thinking development. While all products
provide pre-searched and direct access to relevant educational resources,
XebraPro and XebraED also incorporate both public and private clinician
social media platforms to facilitate the discussion of unique and
challenging cases, as well as workup guides.
The company also has ongoing R&D projects with various academic
institutions to investigate integration of Xebra decision-support tools
with EHRs.


 On a Google and Pub Med search we were able to identify several other
systems with potential for being used as a DCDST:


            1.     The Open Clinical Website has links to some diagnostic
            tools


            2.     LeXMED (Learning EXpert Systems for MEdical Diagnosis)


                  a.     Primarily for Diagnosis of abdominal pain.


            3.     AideDiag, in French


            4.     CACDDST (Computer-Aided Clinical Diagnostic Decision
            Support Tool)


            5.     Up to Date: Available on most University Library
            Websites





This is an evolving field. Some DDSS’s that were available in 2011[1] no
longer have active websites.





To date, it is not yet clear if any of these DCDSTs can be imbedded into
existing EMRs/EHRs for seamless use by clinicians desiring an integrated
DCDS.





We welcome any comments on this topic, which will go through the usual SIDM
ListServe review process.






[1] Bond, WF, et al: Differential Diagnosis Generators: an Evaluation of
Currently Available Computer Programs: J Gen Intern Med, 2011, 000027,
213-9




--
Edward B, J. Winslow, MD, MBA
Home 847 256-2475; Mobile 847 508-1442
edbjwinslow at gmail.com
winslowmedical.com

"The only thing new in the world is the history that you don't know"
       Harry S. Truman, 33rd President of US (1945-1953)



"... it can be argued that underinvestment in assessing the past is likely
to
lead to faulty estimates and erroneous prescriptions for future action."
        Eli Ginzberg, 1997





Address messages to: IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG

To unsubscribe from IMPROVEDX: click the following link:
http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1



or send email to: IMPROVEDX-SIGNOFF-REQUEST at LIST.IMPROVEDIAGNOSIS.ORG

Visit the searchable archives or adjust your subscription at:
http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?INDEX



Moderator:David Meyers, Board Member, Society for Improving Diagnosis in
Medicine

To learn more about SIDM visit:
http://www.improvediagnosis.org/






To unsubscribe from the IMPROVEDX:
mail to:IMPROVEDX-SIGNOFF-REQUEST at LIST.IMPROVEDIAGNOSIS.ORG
or click the following link: IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG

Address messages to: IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG

For additional information and subscription commands, visit:
http://www.lsoft.com/resources/faq.asp#4A

http://LIST.IMPROVEDIAGNOSIS.ORG/ (with your password)

Visit the searchable archives or adjust your subscription at:
http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?INDEX

Moderator: David Meyers, Board Member, Society to Improve Diagnosis in Medicine

To unsubscribe from the IMPROVEDX list, click the following link:<br>
<a href="http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1" target="_blank">http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1</a>
</p>

HTML Version:
URL: <../attachments/20170308/7273bd72/attachment.html> ATTACHMENT:
Name: graycol.gif Type: image/gif Size: 105 bytes Desc: not available URL: <../attachments/20170308/7273bd72/attachment.gif>


More information about the Test mailing list