#4    The Trouble with Post Processing “Collimation”    2-15-2013

     Post processing “collimation”, shuttering or cropping seemed to be one of the greatest features of digital radiography.  With our first CR unit we were taking badly collimated chest x-rays and then with a couple of mouse and key strokes turning in images that looked like we were the best collimating super techs in the world!! 

     After almost two years of this one of our radiologist’s finally realized something weird was gong on and asked what we were doing.  When we told and showed him, he couldn’t believe it.  He immediately told us that what were doing was basically illegal as the radiologist is legally responsible for every bit of anatomy that was radiated and appeared on the image receptor (IR).  This rule has not changed one bit since the film days.

     The predicament we have with this is no one even knows it’s a problem.  I know for a fact that in one somewhat large U.S. city two different hospitals were sued because the radiographer cropped out anatomy that was later proven to have shown a tumor.

     Picture this:  You take a lateral C-spine and get almost the entire mandible on the image.  You don’t notice that there is a small tumor in the mandibular body so you crop out almost the entire mandible, leaving just a perfect looking lateral C-spine.  Six months later lawyers looking through every image taken on their client sees that your images would have shown the tumor half a year ago.  Six weeks ago the patient had to have a huge part of their mandible excised in surgery because of the fast growing cancer.  Your hospital is now being sued for five or ten million dollars and they are not even going to try to fight it because it would be impossible to win.  It was completely your fault and everyone knows it.

     Now I know you haven’t heard about these two cases. Also if it’s happened twice in this city, what are the odds it hasn’t happened countless times in cities all over the country?  The reason this is still a secret is because the hospital is willing to hand over this incredibly large sum of money with one stipulation, and that is a gag order is invoked whereby no one is allowed to talk about the case.  So until the day a patient decides to have their day in court, we are not going to read or hear about this.

     So there are only two ways you can post “collimate” and have it still be legal.  First is to only crop out areas that are outside the border of the body tissue (white or black areas).  Second is to make a copy of the original image and turn it in along with the cropped version.

Back to digitalradiographysolutions website

#3                                How to Properly Critique a Digital Image                                  2-1-2013                             

          We have had hundreds of digital equipment training hours at my hospital in the past 10 years, but in all of that I don’t remember any time given in learning how to correctly critique the image.  I have to imagine this is the same with everyone around the country, possibly the world.

          Last month I wrote about how reliable the Exposure Index (EI) numbers are, especially when the collimation and centering were good.  Today I am again to reiterate that these numbers have to be a huge part of the critiquing process.  Without the actual phantom lateral skull images I use in all of my talks to prove my point, we are stuck with you just having to believe me. 

          I have taken my skull phantom and shot 9 different manufacturers to prove exactly how much mAs can be used and still have a passable image.  Depending on the vendor one can use from 10 - 100 times too much mAs and still have a perfectly passable image, if a visual check is the only guideline being followed.  This is why the EI numbers must be used.

          Sometimes a radiographer will “post collimate” or “shutter” an image which depending on the vendor may change the EI number.  As a huge side note, if any of the actual body is cropped out during this process, the image is now open for a lawsuit (see upcoming blog #4 on Feb. 15^th).

          Another “tool” to use is the magnification mode button.  Quite often burn or mottle can only be seen with the image, or part of the image, being magnified.

          The last “tool” to use is Level/Windowing.  Whether your facility allows you to Level and Window the image and send it to PACS, all departments should allow a radiographer to Level and Window and then reset it before sending it to PACS.  Whenever an image does not look perfect, one should always Level and Window it and see if the image can be made to look perfect (or at least much better).  If after Level and Windowing the image does not look better, it is impossible to know if the radiologist can make it look better.

          So as a quick synopsis, the 3 things that should be done to properly critique a digital image are:

1-    Check the EI number

2-    Use the magnification mode to check for burn and mottle

3-    Level and Window the image and either reset it or send it depending on your facility’s protocol

Back to DRS Website

#2                                Problems with Exposure Index (EI) Numbers                             01-15-2013

     When Fuji first developed CR back in the early 80’s, they were also the first company to invent an Exposure Index (EI) number system.  As most of us know, they went with S numbers and decided to have those numbers correspond with the speed of film –screen systems so that the higher the number the less the dose.  Unfortunately they didn't ask one radiographer anywhere in the world who would have told them that this is was not a good idea as it’s not intuitive and it’s backwards from the way we think.  Pretty much every CR vendor after them followed suite as they decided to use Logarithms.  Finally DR came into the picture and all the vendors were able to develop EI numbers that were more user friendly.  Almost all of them came up with systems that had an EI number that would double when the mAs was doubled.
     Pretty much all digital radiography equipment have some sort of EI number.  There are S, LgM, DEI, EXI, REX, EI and probably others.  These numbers were developed to tell us how much radiation made it through the patient and exposed the image receptor (IR).  The problems with the numbers are twofold:
1-      They are easily corrupted or skewed by poor collimation and/or centering (up to 75% I've found in my research).
2-      Unless one is good with math the EI numbers are difficult to use to figure out how to fix a bad technique.
     Because of these two reasons, many radiographers do not use the EI numbers when critiquing their images.  Instead they use the same methods that were used with film, which are incorrect (see my next blog on Feb 1^st that will cover how to properly critique a digital image).  These EI numbers are quite reliable, even more so when the collimation and centering are good.  This is why everyone should be trying their hardest to collimate as tightly as possible.
     There is no doubt that the EI system is not a great system by any stretch of the imagination, but it is the ONLY system we have to help us critique our technique at this point in time.  Hopefully in the near future the manufacturers will be able to create better and truer EI numbers, so all radiographers will truly know if they have used the correct technique.
     I'm not a big government guy, but I think I would be OK with Big Daddy telling the vendors that they have to all play nice with each other and figure out 1 EI system that will be universal.  It's all going to be too much hassle (and costly) for them to do it, so unless they are forced into it, I don't  think it will ever happen.  And that's just a cryin' shame.


Back to DRS Website

#1                           The Problems of “Creeping mAs/Dose” in America                             01/01/13     

     As this is my first blog, I decided to jump right in and tackle what I believe is one of the biggest problems in radiology today; and that is the over radiation that is occurring everywhere with digital x-ray (CR and DR).  As of January 1st, 2013, well over half of the radiology departments in the US are now digital.  

     The first CR equipment came out in 1983, which means it has now been around for 30 years.  That's thirty years!!  The problem is most of us never had proper training when it came to the new techniques which implements using at least 15% more kV and cutting the mAs in half.  By doing this the patient’s entrance dose is cut by 33%.  Most vendors either had facilities utilize the same techniques that were used with film, a slightly higher kV with the same or even more mAs.

     Most training consisted of being shown how to correctly set up and get an exposure and post process the image, and that was about it.  The whole concept of a “light” image on your monitor no longer exists but if the patient is under radiated the image could have mottle/noise.  To take no chances, many radiographers slowly started to increase the mAs little by little to guarantee a perfect “looking” image every time. The problem is the computer has the ability to automatically rescale the image and make it look perfect even though 5, 10, 50 times too much mas/dose/radiation is used.  The Exposure Index (EI) numbers show that the patient has been over radiated but these numbers are not utilized nearly enough (see my next blog on the EI numbers later this month).

     To make matters even worse, students don’t really learn to be as discriminating in determining the exact size of a patient so that they can use the perfect technique.  Also there is so much latitude with how much mAs can be used in digital radiography, it’s impossible for any radiographer or student to be as good with manual techniques as we had to be with film.  So each year the mAs slowly and insidiously creeps up a little bit at a time but over a 10-20 year period departments are now using double and triple the mAs that is needed    

     Finally, how do we combat this problem? 

1.     Everyone needs to be better educated about using higher kV and lower mAs.

2.     Radiographers need to teach the students how to visually access the correct size of a patient (or use calipers).

3.     Everyone needs to be better trained in using the EI numbers.