Re: OCR

> Synapse - the use from anywhere EMR.

>

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

One way it would help is you would not need to open and view the document to sort it to patient's chart from a batch.Jacques Guillot, MD

Don

How would it help you? You still have to process it further...

On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

wrote:

>

> Graham,

>

> I would pay money for such a utility.

>

> dts

>

> From:

> [mailto: ] On Behalf Of Graham Chiu

> Sent: Saturday, April 05, 2008 7:45 PM

> To:

> Subject: OCR

>

> Do any imps feel that a utility that examines a scan, and extracts the

> patient name, date of test and id would be useful?

> I'd imagine that the utility would then rename the file eg:

> id_date_surname_firstname.pdf so it could then up processed further.

>

> --

> Graham Chiu

> Synapse - the use from anywhere EMR.

>

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

On Sun, Apr 6, 2008 at 4:47 PM, Jacques Guillot

wrote:

>

> One way it would help is you would not need to open and view the document to

> sort it to patient's chart from a batch.

>

> Jacques Guillot, MD

>

> >

> > Don

> >

> > How would it help you? You still have to process it further...

> >

> > On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

> > wrote:

> > >

> > > Graham,

> > >

> > > I would pay money for such a utility.

> > >

> > > dts

> > >

> > > From:

> > > [mailto: ] On Behalf Of Graham Chiu

> > > Sent: Saturday, April 05, 2008 7:45 PM

> > > To:

> > > Subject: OCR

> > >

> > > Do any imps feel that a utility that examines a scan, and extracts the

> > > patient name, date of test and id would be useful?

> > > I'd imagine that the utility would then rename the file eg:

> > > id_date_surname_firstname.pdf so it could then up processed further.

> > >

> > > --

> > > Graham Chiu

> > > Synapse - the use from anywhere EMR.

> > >

> >

> > --

> > Graham Chiu

> > http://www.synapsedirect.com

> > Synapse - the use from anywhere EMR.

> >

>

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

But someone has to view the document ?

Or are you saying that most documents are viewed before being scanned?

So, you don't need to look at them again?

>

> On Sun, Apr 6, 2008 at 4:47 PM, Jacques Guillot

> wrote:

> >

> > One way it would help is you would not need to open and view the document

to

> > sort it to patient's chart from a batch.

> >

> > Jacques Guillot, MD

> >

> > >

> > > Don

> > >

> > > How would it help you? You still have to process it further...

> > >

> > > On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

> > > wrote:

> > > >

> > > > Graham,

> > > >

> > > > I would pay money for such a utility.

> > > >

> > > > dts

> > > >

> > > > From:

> > > > [mailto: ] On Behalf Of Graham Chiu

> > > > Sent: Saturday, April 05, 2008 7:45 PM

> > > > To:

> > > > Subject: OCR

> > > >

> > > > Do any imps feel that a utility that examines a scan, and extracts the

> > > > patient name, date of test and id would be useful?

> > > > I'd imagine that the utility would then rename the file eg:

> > > > id_date_surname_firstname.pdf so it could then up processed further.

> > > >

> > > > --

> > > > Graham Chiu

> > > > http://www.synapsedirect.com

> > > > Synapse - the use from anywhere EMR.

> > > >

> > >

> > > --

> > > Graham Chiu

> > > Synapse - the use from anywhere EMR.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

The quick answer is that it would save me from typing the patient

name and test date, and if you could identify what kind of a document it was

and name it consistently, it would help me get the file into the correct part

of the patient’s chart.

My current workflow with faxed reports is to:

1.

view them from the fax inbox, then save to a “to be

attached to chart” directory and name them:

LastName_FirstName_TestOrConsultType_Date.pdf

2.

At some later time, I or my wife goes through the “to be

attached to chart” directory with a utility that allows us to save the

file to the appropriate folder in the patient’s chart

3.

When I am in a chart, I if the scanned images are not yet in the

chart, I open a window to the “to be attached to chart” directory,

and can open the file and view it there.

The above process takes a lot of time, and for me, importing the

scanned images into the chart itself is a relatively low priority behavior,

since I can get the files easily enough from the “to be attached to chart”

directory. However, I really would like to get all of the images into the

appropriate sections of the charts.

If all the images were consistently named, it would not be too

much trouble for me to create a program that reads the file names from the “to

be attached …” directory, parses the data from the file names, and

imports the images into the correct chart sections. Although I could do

this now if I were consistent in naming the files, having an automated way of

doing the file naming would reduce errors.

dts

From:

[mailto: ] On Behalf Of Graham Chiu

Sent: Saturday, April 05, 2008 9:24 PM

To:

Subject: Re: OCR

Don

How would it help you? You still have to process it further...

On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

wrote:

>

> Graham,

>

> I would pay money for such a utility.

>

> dts

>

> From:

> [mailto: ]

On Behalf Of Graham Chiu

> Sent: Saturday, April 05, 2008 7:45 PM

> To:

> Subject: OCR

>

> Do any imps feel that a utility that examines a scan, and extracts the

> patient name, date of test and id would be useful?

> I'd imagine that the utility would then rename the file eg:

> id_date_surname_firstname.pdf so it could then up processed further.

>

> --

> Graham Chiu

> Synapse - the use from anywhere EMR.

>

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

In my office, we do our best never to get documents that

need to be scanned, but when we do, I read them and sign off on them before they

are scanned. At txhe most, there are just a few of these a week.

But, I get 30 to 40 faxes every day, and perhaps half of these

are documents that need to be filed. The rest are prescription refill

requests. I read each of these in the fax inbox. I described the workflow

in another post.

dts

From:

[mailto: ] On Behalf Of Graham Chiu

Sent: Saturday, April 05, 2008 9:51 PM

To:

Subject: Re: OCR

But someone has to view the document ?

Or are you saying that most documents are viewed before being scanned?

So, you don't need to look at them again?

>

> On Sun, Apr 6, 2008 at 4:47 PM, Jacques Guillot

>

wrote:

> >

> > One way it would help is you would not need to open and view the

document to

> > sort it to patient's chart from a batch.

> >

> > Jacques Guillot, MD

> >

> > >

> > > Don

> > >

> > > How would it help you? You still have to process it further...

> > >

> > > On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

> > >

wrote:

> > > >

> > > > Graham,

> > > >

> > > > I would pay money for such a utility.

> > > >

> > > > dts

> > > >

> > > > From:

> > > > [mailto: ]

On Behalf Of Graham Chiu

> > > > Sent: Saturday, April 05, 2008 7:45 PM

> > > > To:

> > > > Subject: OCR

> > > >

> > > > Do any imps feel that a utility that examines a scan, and

extracts the

> > > > patient name, date of test and id would be useful?

> > > > I'd imagine that the utility would then rename the file eg:

> > > > id_date_surname_firstname.pdf so it could then up processed

further.

> > > >

> > > > --

> > > > Graham Chiu

> > > > http://www.synapsedirect.com

> > > > Synapse - the use from anywhere EMR.

> > > >

> > >

> > > --

> > > Graham Chiu

> > > Synapse - the use from anywhere EMR.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

I'm currently using

testdate_patientid_surname_firstname_originalfilename

but I guess

surname_fname_patientid_testdate_originalfilename

might be better if you need to locate a record quickly

Eventually I could also do this

surname_fname_patientid_testdate_testtype_originalfilename

On Sun, Apr 6, 2008 at 5:22 PM, T. , MD

wrote:

>

> The quick answer is that it would save me from typing the patient name and

> test date, and if you could identify what kind of a document it was and name

> it consistently, it would help me get the file into the correct part of the

> patient's chart.

>

> My current workflow with faxed reports is to:

>

> 1. view them from the fax inbox, then save to a " to be attached to

> chart " directory and name them:

> LastName_FirstName_TestOrConsultType_Date.pdf

>

> 2. At some later time, I or my wife goes through the " to be attached

> to chart " directory with a utility that allows us to save the file to the

> appropriate folder in the patient's chart

>

> 3. When I am in a chart, I if the scanned images are not yet in the

> chart, I open a window to the " to be attached to chart " directory, and can

> open the file and view it there.

>

> The above process takes a lot of time, and for me, importing the scanned

> images into the chart itself is a relatively low priority behavior, since I

> can get the files easily enough from the " to be attached to chart "

> directory. However, I really would like to get all of the images into the

> appropriate sections of the charts.

>

> If all the images were consistently named, it would not be too much trouble

> for me to create a program that reads the file names from the " to be

> attached … " directory, parses the data from the file names, and imports the

> images into the correct chart sections. Although I could do this now if I

> were consistent in naming the files, having an automated way of doing the

> file naming would reduce errors.

>

> dts

>

> From:

> [mailto: ] On Behalf Of Graham Chiu

> Sent: Saturday, April 05, 2008 9:24 PM

> To:

> Subject: Re: OCR

>

> Don

>

> How would it help you? You still have to process it further...

>

> On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

> wrote:

> >

> > Graham,

> >

> > I would pay money for such a utility.

> >

> > dts

> >

> > From:

> > [mailto: ] On Behalf Of Graham Chiu

> > Sent: Saturday, April 05, 2008 7:45 PM

> > To:

> > Subject: OCR

> >

> > Do any imps feel that a utility that examines a scan, and extracts the

> > patient name, date of test and id would be useful?

> > I'd imagine that the utility would then rename the file eg:

> > id_date_surname_firstname.pdf so it could then up processed further.

> >

> > --

> > Graham Chiu

> > http://www.synapsedirect.com

> > Synapse - the use from anywhere EMR.

> >

>

> --

> Graham Chiu

> Synapse - the use from anywhere EMR.

>

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

I view hard copies first (mostly fax and mail) and sign off. My staff scans them directly into patient chart in Allscripts " Impact.MD " . I got this specifically because scanning was easy. Shortcoming is it is not a full EMR with relational database to search drugs, diagnoses, screenings, etc.

I tried OCR for full documents with my first EMR. It was hard! Too much editing. Then I tried to only get key details from report, eg patient name, DOB, narrative and summary/conclusion interpretation - and cut out the fluff. I learned there is a lot of information on that report in addition to interpretation! Where done? (Left that out, oops.) What time done? That came up more than I'd expect (esp for elevated glucoses). It was superior for me to have the entire document available than to take time to process it or decide which parts of it to keep.

I am not afraid of change, but there seems to be something I trust about signing off that piece of paper, and trusting my staff will act on it (new orders, notify patient, fax to other treating physicians) when it is in a physical outbox. I do use an internal e-mail system in my EHR to document phone calls and staff orders/messages, and it works great, so I am not all paper and scanning.

I have UpDox, I've bought full Adobe Acrobat in past. I know C. has been great at implementing these. But it has resulted in more work so many times to go through too many electronic steps that were actually more streamlined with pen and paper - and scanner. And it is easier many times to have my staff handle these things than to take up more of my own time handling them in what would be a more elegant fashion.

Another thing about handling the hard copies is it is so easy to have the paper in front of me while I browse around in the EHR to look for patient history and old reports to compare. It's like having a very flexible extra touchscreen monitor on which I can write freehand on the fly!

Jacques Guillot, MD

But someone has to view the document ?

Or are you saying that most documents are viewed before being scanned?

So, you don't need to look at them again?

>

> On Sun, Apr 6, 2008 at 4:47 PM, Jacques Guillot

> wrote:

> >

> > One way it would help is you would not need to open and view the document to

> > sort it to patient's chart from a batch.

> >

> > Jacques Guillot, MD

> >

> > >

> > > Don

> > >

> > > How would it help you? You still have to process it further...

> > >

> > > On Sun, Apr 6, 2008 at 4:19 PM, T. , MD

> > > wrote:

> > > >

> > > > Graham,

> > > >

> > > > I would pay money for such a utility.

> > > >

> > > > dts

> > > >

> > > > From:

> > > > [mailto: ] On Behalf Of Graham Chiu

> > > > Sent: Saturday, April 05, 2008 7:45 PM

> > > > To:

> > > > Subject: OCR

> > > >

> > > > Do any imps feel that a utility that examines a scan, and extracts the

> > > > patient name, date of test and id would be useful?

> > > > I'd imagine that the utility would then rename the file eg:

> > > > id_date_surname_firstname.pdf so it could then up processed further.

> > > >

> > > > --

> > > > Graham Chiu

> > > > http://www.synapsedirect.com

> > > > Synapse - the use from anywhere EMR.

> > > >

> > >

> > > --

> > > Graham Chiu

> > > Synapse - the use from anywhere EMR.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

I'm doing all of those things - I store the original document, the

ocr'd version, and a synopsis I write.

I do try to avoid looking at the original ever again.

On Sun, Apr 6, 2008 at 5:40 PM, Jacques Guillot

wrote:

> I tried OCR for full documents with my first EMR. It was hard! Too much

> editing. Then I tried to only get key details from report, eg patient name,

> DOB, narrative and summary/conclusion interpretation - and cut out the

> fluff. I learned there is a lot of information on that report in addition

> to interpretation! Where done? (Left that out, oops.) What time done?

> That came up more than I'd expect (esp for elevated glucoses). It was

> superior for me to have the entire document available than to take time to

> process it or decide which parts of it to keep.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

When I scan a hard copy, I am finished, I don't have to look at it again.I have found it quicker to look at the original when needed for later reference than to perform the work that would make looking a tweaked or processed version of the original almost always unnecessary. Of course, I do update my problem list with major updates, so in that regard, I won't have to review a report for any big-picture information.

Jacques

April 6, 2008

I'm in a similar situation with stuff that gets faxed to me

that I want to keep. I view all faxes on the computer, and if it is

something significant that I want to remember, I make an entry

in a health history summary page that comes up on each patient and

try to file it electronically at that time. I spend a fair amount

of time making the entries in the health history summary page, but

it's worth it because if the pt. calls me on the phone or comes in, I

can refer to that page usually to remember what they've had done

without referring or looking through the original documents.

Like for an MRI or an important procedure such as a colonoscopy,

a stress test I enter a brief note by hand in the health summary

page and try to refer to that note rather than pulling up or

looking through the original documents from then on. I do file these

under each patient with a filename that tries to make it easier

to find things if I do need to look at the original or send a copy

of it to somebody else. The filenames are in this format:

YYYY_MM_DD_TypeInfo_pagenumber.g3 and are mostly generated by the

computer.

The date gets added automatically by the computer but is the date

filed, not the date of the test, but usually they are not too far

apart. The " TypeInfo " I pick from a dropdown list. I add the

specialty and consultants last name if it is a consult. The page

number is for multipage

faxes that you want to keep together as a group. The .g3 is for

..g3 tif which is the image format faxes are done in. I need the .g3

because I am using one fax viewer program on Linux that needs it

to know it's a fax format. Later when you look at a list of these

files on the computer in a listbox, the computer will sort them

out by date, type of document, consultants name if present and

finally by page number. This makes it easier to find things in scanned

documents by using some clue as to when it was done, the type of

document, the consultants name etc.

On stuff that comes in as paper, I sign and date it and make the same

entry in the past history summary page and then later my wife scans it

files the paper in alphabetical order by patient last name in a big

bin. We have one bin for each year we've been in practice. When we

have time, we eventually file the scanned in stuff electronically

under the patient in the EMR, but this process is time consuming and

the last thing on our priority list, so we are chronically behind on

it.

> > I tried OCR for full documents with my first EMR. It was hard!

Too much

> > editing. Then I tried to only get key details from report, eg

patient name,

> > DOB, narrative and summary/conclusion interpretation - and cut

out the

> > fluff. I learned there is a lot of information on that report in

addition

> > to interpretation! Where done? (Left that out, oops.) What

time done?

> > That came up more than I'd expect (esp for elevated glucoses).

It was

> > superior for me to have the entire document available than to

take time to

> > process it or decide which parts of it to keep.

>

> --

> Graham Chiu

> Synapse - the use from anywhere EMR.

>

April 6, 2008

Yes, this is very time consuming and is the part I am attempting to automate.

>

> On stuff that comes in as paper, I sign and date it and make the same

> entry in the past history summary page and then later my wife scans it

> files the paper in alphabetical order by patient last name in a big

> bin. We have one bin for each year we've been in practice. When we

> have time, we eventually file the scanned in stuff electronically

> under the patient in the EMR, but this process is time consuming and

> the last thing on our priority list, so we are chronically behind on

> it.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

I think that utility would be wonderful, Graham -I'm fascinated to think it would be possible - I find that even looking at all the reports I get, I have trouble scanning for that information. The names, DOB's, DOS, all scattered differently through the material. And Jacques-I used to do the same thing, feeling that paper was faster (and just safer, easier). I had my paper faxes also stored simultaneously on the computer fax inbox and they were sorted from there in a way similar to Don's method. I paid someone just to sit and sort those documents.My staff doesn't write the letters, though, I do, so another step seemed to be involved. See fax, sign, think.Then later in the day (or days later) finally get around to writing a note quickly, but thinking about it again, using my EMR and templates. It took at least 2-4 weeks to get used to UpDox, but it did accomplish one thing that is efficient. I only look at the document once and react then. Open file, read labs, assign patient, move to EMR file at which point an email letter opens automatically. Type a few lines about the labs, off they go. No stamps. Also, the way I sort labs and results is different, the files are very accurately named, so they are easier to use later.My pharmacy faxes still go to my land line fax to be handled in paper. Only lab and radiology go to the box directly to UpDox. GuinnI view hard copies first (mostly fax and mail) and sign off. My staff scans them directly into patient chart in Allscripts "Impact.MD". I got this specifically because scanning was easy. Shortcoming is it is not a full EMR with relational database to search drugs, diagnoses, screenings, etc. I tried OCR for full documents with my first EMR. It was hard! Too much editing. Then I tried to only get key details from report, eg patient name, DOB, narrative and summary/conclusion interpretation - and cut out the fluff. I learned there is a lot of information on that report in addition to interpretation! Where done? (Left that out, oops.) What time done? That came up more than I'd expect (esp for elevated glucoses). It was superior for me to have the entire document available than to take time to process it or decide which parts of it to keep. I am not afraid of change, but there seems to be something I trust about signing off that piece of paper, and trusting my staff will act on it (new orders, notify patient, fax to other treating physicians) when it is in a physical outbox. I do use an internal e-mail system in my EHR to document phone calls and staff orders/messages, and it works great, so I am not all paper and scanning. I have UpDox, I've bought full Adobe Acrobat in past. I know C. has been great at implementing these. But it has resulted in more work so many times to go through too many electronic steps that were actually more streamlined with pen and paper - and scanner. And it is easier many times to have my staff handle these things than to take up more of my own time handling them in what would be a more elegant fashion. Another thing about handling the hard copies is it is so easy to have the paper in front of me while I browse around in the EHR to look for patient history and old reports to compare. It's like having a very flexible extra touchscreen monitor on which I can write freehand on the fly!Jacques Guillot, MDOn Sat, Apr 5, 2008 at 11:51 PM, Graham Chiu <compkarorigmail> wrote:But someone has to view the document ?Or are you saying that most documents are viewed before being scanned?So, you don't need to look at them again?On Sun, Apr 6, 2008 at 4:49 PM, Graham Chiu <compkarorigmail> wrote:>> On Sun, Apr 6, 2008 at 4:47 PM, Jacques Guillot> <jacques.guillotgmail> wrote:> >> >> >> >> >> >> > One way it would help is you would not need to open and view the document to> > sort it to patient's chart from a batch.> >> > Jacques Guillot, MD> >> >> >> > On Sat, Apr 5, 2008 at 10:23 PM, Graham Chiu <compkarorigmail> wrote:> >> > >> > >> > >> > >> > >> > >> > > Don> > >> > > How would it help you? You still have to process it further...> > >> > >> > > On Sun, Apr 6, 2008 at 4:19 PM, T. , MD> > > <DonSsammamishdiabetesandlipid (DOT) org> wrote:> > > >> > > >> > > > Graham,> > > >> > > >> > > >> > > > I would pay money for such a utility.> > > >> > > >> > > >> > > > dts> > > >> > > >> > > >> > > >> > > >> > > > From: > > > > [mailto: ] On Behalf Of Graham Chiu> > > > Sent: Saturday, April 05, 2008 7:45 PM> > > > To: > > > > Subject: OCR> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > > Do any imps feel that a utility that examines a scan, and extracts the> > > > patient name, date of test and id would be useful?> > > > I'd imagine that the utility would then rename the file eg:> > > > id_date_surname_firstname.pdf so it could then up processed further.> > > >> > > > --> > > > Graham Chiu> > > > http://www.synapsedirect.com> > > > Synapse - the use from anywhere EMR.> > > >> > >> > > --> > > Graham Chiu> > > http://www.synapsedirect.com> > > Synapse - the use from anywhere EMR.-- Graham Chiuhttp://www.synapsedirect.comSynapse - the use from anywhere EMR.

April 6, 2008

The name of the entry (if I am the person naming sorted faxes) becomes the equivalent of those health history entries.I just go to the Imports section in a person's chart, scan a list that reads:Radiology:MRI brain normal dateXray ankle fracture dateDexa dateLabs:TSH 4.5 dateCMP HBza1c 6.5 U micro dateLipid dateConsultCardio meds visit and echo dateOrtho knee postopI can make the tags say whatever I want because I'm writing them when I'm looking at the results.The date is automatically attached when it's imported into the EMR.I'm in a similar situation with stuff that gets faxed to methat I want to keep. I view all faxes on the computer, and if it issomething significant that I want to remember, I make an entry in a health history summary page that comes up on each patient andtry to file it electronically at that time. I spend a fair amountof time making the entries in the health history summary page, butit's worth it because if the pt. calls me on the phone or comes in, Ican refer to that page usually to remember what they've had done without referring or looking through the original documents. Like for an MRI or an important procedure such as a colonoscopy,a stress test I enter a brief note by hand in the health summarypage and try to refer to that note rather than pulling up orlooking through the original documents from then on. I do file these under each patient with a filename that tries to make it easierto find things if I do need to look at the original or send a copyof it to somebody else. The filenames are in this format:YYYY_MM_DD_TypeInfo_pagenumber.g3 and are mostly generated by the computer.The date gets added automatically by the computer but is the datefiled, not the date of the test, but usually they are not too farapart. The "TypeInfo" I pick from a dropdown list. I add the specialty and consultants last name if it is a consult. The page number is for multipage faxes that you want to keep together as a group. The .g3 is for.g3 tif which is the image format faxes are done in. I need the .g3because I am using one fax viewer program on Linux that needs itto know it's a fax format. Later when you look at a list of thesefiles on the computer in a listbox, the computer will sort themout by date, type of document, consultants name if present and finally by page number. This makes it easier to find things in scanneddocuments by using some clue as to when it was done, the type of document, the consultants name etc.On stuff that comes in as paper, I sign and date it and make the same entry in the past history summary page and then later my wife scans itfiles the paper in alphabetical order by patient last name in a big bin. We have one bin for each year we've been in practice. When we have time, we eventually file the scanned in stuff electronically under the patient in the EMR, but this process is time consuming and the last thing on our priority list, so we are chronically behind on it.> > I tried OCR for full documents with my first EMR. It was hard! Too much> > editing. Then I tried to only get key details from report, eg patient name,> > DOB, narrative and summary/conclusion interpretation - and cut out the> > fluff. I learned there is a lot of information on that report in addition> > to interpretation! Where done? (Left that out, oops.) What time done?> > That came up more than I'd expect (esp for elevated glucoses). It was> > superior for me to have the entire document available than to take time to> > process it or decide which parts of it to keep.> > -- > Graham Chiu> http://www.synapsedirect.com> Synapse - the use from anywhere EMR.>

April 6, 2008

On Mon, Apr 7, 2008 at 2:39 AM, scottclemensen@...

wrote:

>

> This is a great workflow discussion! A couple of comments:

> 1. If you are processing your lab reports on the same day you get them, and

> part of your automated filing system puts a date on them automatically, than

> how much benefit is there to place the actual date of the blood draw? I

I think it's the time factor. If you don't have time to file them

inside your EMR, then having them datestamped in the name is a

temporay fix.

> don't bother. All imported items into Amazing Charts are time stamped

> automatically and that is good enough for me.

> 2. As for putting a patient name on the scanned document, UpDox minimizes

and same for the name.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

I can now take a directory of scanned PDFs ( not previously OCR'd ),

and apply my recognition engine to them. My test suite was mainly lab

results because they are forms, and I can get close to 100% accuracy.

The files are then renamed surname_fname_id_testdate_filename.ext but

actually there is enough information now for Synapse to insert them

directly into the Synapse database.

I asked the question to see whether there was enough interest in

pulling this functionality out as a stand alone utility.

It would run either on Windows or Linux. And possibly OSX. And it

requires a decent scanner preferably with auto deskewing.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 6, 2008

I would love to try/test the Windows version.

dts

From:

[mailto: ] On Behalf Of Graham Chiu

Sent: Sunday, April 06, 2008 12:32 PM

To:

Subject: Re: Re: OCR

I can now take a directory of scanned PDFs ( not previously OCR'd ),

and apply my recognition engine to them. My test suite was mainly lab

results because they are forms, and I can get close to 100% accuracy.

The files are then renamed surname_fname_id_testdate_filename.ext but

actually there is enough information now for Synapse to insert them

directly into the Synapse database.

I asked the question to see whether there was enough interest in

pulling this functionality out as a stand alone utility.

It would run either on Windows or Linux. And possibly OSX. And it

requires a decent scanner preferably with auto deskewing.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 7, 2008

That's how it works in Synapse too, but we also have to deal with non

faxed paper records that need to be scanned in .. and that's time

intensive.

>

> In EMD's the lab and radiology faxes (not to mention faxed records) come

> directly into the program-no paper- you can electronically refile them to

but to refile them, you have to open it up and look at it

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 7, 2008

>

> I think that utility would be wonderful, Graham -

> I'm fascinated to think it would be possible - I find that even looking at

> all the reports I get, I have trouble scanning for that information.

> The names, DOB's, DOS, all scattered differently through the material.

Well, the utility does require that the names, dob and dos are

consistently in the same place for a particular provider. If the same

provider varies where they put that information, well, that is a much

harder task.

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 7, 2008

I think you should market the code to all the EMR companies.

Automating the filing all the way into the EMR would be huge.

Haresch

>

> I can now take a directory of scanned PDFs ( not previously OCR'd ),

> and apply my recognition engine to them. My test suite was mainly lab

> results because they are forms, and I can get close to 100% accuracy.

> The files are then renamed surname_fname_id_testdate_filename.ext but

> actually there is enough information now for Synapse to insert them

> directly into the Synapse database.

>

> I asked the question to see whether there was enough interest in

> pulling this functionality out as a stand alone utility.

>

> It would run either on Windows or Linux. And possibly OSX. And it

> requires a decent scanner preferably with auto deskewing.

>

> --

> Graham Chiu

> Synapse - the use from anywhere EMR.

>

April 7, 2008

The response from one vendor was " why bother when really users want HL7 " ...

>

> I think you should market the code to all the EMR companies.

> Automating the filing all the way into the EMR would be huge.

>

> Haresch

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 7, 2008

We will probably be able to get radiology reports in HL7 in the

next 5-10 years, but I would be surprised if we got consult notes in that

format in the near future.

dts

From:

[mailto: ] On Behalf Of Graham Chiu

Sent: Sunday, April 06, 2008 6:05 PM

To:

Subject: Re: OCR

The response from one vendor was " why bother when really users want

HL7 " ...

>

> I think you should market the code to all the EMR companies.

> Automating the filing all the way into the EMR would be huge.

>

> Haresch

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 7, 2008

One would wonder if they have actually talked to their users. Real

life is that we're spending a ton of time getting paper into EMRs, and

apparently will be for a long time.

>

> The response from one vendor was " why bother when really users want

HL7 " ...

April 7, 2008

Lynn

I'll add you to my list of people interested in testing

>

> Graham

> Sounds like a great utility.

> I have a bunch of folders of patient related documents that do not get put

> into the EMR - prior auths, physical therapy, outgoing forms, VNA,

> nutrtion, faxed old records- low information yield documents (a polite way

> of saying that these documents are not that meaningful to me). Although it

> is not a huge number, I do sort them by last name, first names, type, date

> and then stick them in the proper folder, not in the EMR. They are low

> priority and so they sit for a while. Hmm. I suppose I could sort them

> using Updox into the charts but it seems less cumbersome to keep them in the

> folders and out of the EMR, maybe slightly less time (14 seconds less) than

> to put them into the EMR with Updox where you have to find the name and

> click to assign, and also I feel like I don't want to adulterate what weak

> real information is in my EMR with what is in these folders, so I've kept

> them out so far.

> I would try it out for scanned documents if you are looking for people to

> try it.

> Lynn

>

--

Graham Chiu

Synapse - the use from anywhere EMR.

April 7, 2008

I have tried a couple of open source OCR programs on Linux to

see if I could extract the data and automate lab test filing. Our hospital

can generate the lab reports in a fax format and send them. They

refuse to give me an HL7 feed. Since the fax file is actually produced

by the computer, rather than being hand scanned, they are never

skewed and are very consistent in size and type of font and positioning

of the data. I found that the open source OCR programs " GOCR "

and " Tesseract " could do a pretty good job on the computer generated

faxes. I'm pretty sure you could

write a little program that would be able to extract the name, date, and

even the results from those (I haven't done it). But the faxes that were

hand scanned, and not computer generated as a fax without scanning,

gave very poor results with

the open source OCR programs. I imagine some of the commercial

OCR program have better ability to deal with skewed hand scanned

faxes with multiple sizes and type of fonts etc. What OCR " engine "

are you

using in your project? Did you write it yourself or are you using

an open source or commercial OCR program?

> > >

> > > I think that utility would be wonderful, Graham -

> > > I'm fascinated to think it would be possible - I find that even

looking at

> > > all the reports I get, I have trouble scanning for that information.

> > > The names, DOB's, DOS, all scattered differently through the

material.

> >

> > Well, the utility does require that the names, dob and dos are

> > consistently in the same place for a particular provider. If the same

> > provider varies where they put that information, well, that is a much

> > harder task.

>

> --

> Graham Chiu