I've been getting a lot of questions on the side lately about "optimum kVp" and "how to choose what kV to use." Even though there are so many variations in x-ray producing equipment and image receptors, there are two main categories that I like to separate my answers in:
1. Optimum kVp using film/screen systems
2. Optimum kVp using digital systems
Film/Screen:
A radiographer's "old-school" medium, film/screen imaging systems require precise kVp based upon part thickness, pathology, atomic number of the part, and presence or absence of air. I would say that an "average" kVp range could be acquired for most body parts (with the exception of Chest imaging) by using calipers. We have to be precise because with film/screen systems, kVp is our primary controlling factor for contrast.
One method an x-ray tech can use to formulate a starting kV is by measuring a body part with calipers (in centimeters), then multiplying that measurement by 2, and then adding 40 to that number. Here's an example: My elbow measures about 10cm. 10 x 2 = 20. 20 + 40 = 60. Using this method, I could use 60 kVp to provide adequate part penetration and contrast using film/screen systems. My optimum mAs is another post... as you know, generator phasing and film/screen combinations will need to be addressed for proper mAs calculations.
There are also many charts available in textbooks that will display optimum kVp ranges by body part. You will need to produce some test images in the beginning phases of what kVp ranges you would like to use, and consult your Radiologists to find out what kind of "image appearance" they prefer before setting your ranges in stone. If they prefer films that are more gray (less contrast), simply add 50 instead of 40 in the example above. If they prefer higher contrast, perhaps only add 35 (recommended for extremities in an orthopedic clinic). I would refrain from using any less than 70 kVp for spine exams.
Digital Systems:
The rules change a bit here... the primary controlling factor for contrast ceases to be kVp in CR and DR systems. With these, we use the processing algorithm as the primary controller. The fundamental principles of radiation physics still apply, like the more kVp we use, the more potential difference, therefore more part penetration we will cause. But since the kVp range does not need to be as precise as film/screen systems, we have some room to play with technique a bit.
The next logical step in our reasoning is to determine if an increase or decrease in kVp will provide better images. Because the beam physics haven't changed, we can't really lower our kVp compared to optimum film ranges because no matter how magical our computed radiography and digital radiography algorithms are, they are still dependent upon adequate exposure to the image receptor. We can, however, increase our kVp.
How much should we increase? In my personal experience, I have used about 10 kVp higher than film/screen systems for small extremities, and up to 20 or 25 kVp higher than film/screen systems for thick body parts like spine. I could never use 90 kVp on a lateral c-spine with plain film because it would be too gray/washed out. Since digital radiography adjusts the scale of contrast to be ideal for a "lateral c-spine" I have two distinct advantages when using higher kVp with these systems: More uniform part penetration and less radiation dose to the patient.
For part penetration, you will be able to visualize the lower cervical and upper thoracic vertebrae at 90 kVp than at the traditional 70 kVp range. Using the 15% rule, you know that using 90 kVp will require far less mAs (about 25% than at 70 kV), which will significantly reduce the exposure to the patient. Talk about a win-win situation!
I would encourage anyone looking to make changes to their technical factors to consider a higher kV range with dose reduction. You may be tempted to use ridiculously low mAs values like 0.5 mAs on various body parts, but remember the physics... you will get quantum mottle or image noise when there are not enough x-ray photons reaching your image receptor. I would love to hear what you currently use at your facility, and if you try some of these changes, I would also love to hear what you think about them. Whatever changes you make, please keep ALARA principles in mind, and remember to consult your Radiologists.
1. Optimum kVp using film/screen systems
2. Optimum kVp using digital systems
Film/Screen:
A radiographer's "old-school" medium, film/screen imaging systems require precise kVp based upon part thickness, pathology, atomic number of the part, and presence or absence of air. I would say that an "average" kVp range could be acquired for most body parts (with the exception of Chest imaging) by using calipers. We have to be precise because with film/screen systems, kVp is our primary controlling factor for contrast.
One method an x-ray tech can use to formulate a starting kV is by measuring a body part with calipers (in centimeters), then multiplying that measurement by 2, and then adding 40 to that number. Here's an example: My elbow measures about 10cm. 10 x 2 = 20. 20 + 40 = 60. Using this method, I could use 60 kVp to provide adequate part penetration and contrast using film/screen systems. My optimum mAs is another post... as you know, generator phasing and film/screen combinations will need to be addressed for proper mAs calculations.
There are also many charts available in textbooks that will display optimum kVp ranges by body part. You will need to produce some test images in the beginning phases of what kVp ranges you would like to use, and consult your Radiologists to find out what kind of "image appearance" they prefer before setting your ranges in stone. If they prefer films that are more gray (less contrast), simply add 50 instead of 40 in the example above. If they prefer higher contrast, perhaps only add 35 (recommended for extremities in an orthopedic clinic). I would refrain from using any less than 70 kVp for spine exams.
Digital Systems:
The rules change a bit here... the primary controlling factor for contrast ceases to be kVp in CR and DR systems. With these, we use the processing algorithm as the primary controller. The fundamental principles of radiation physics still apply, like the more kVp we use, the more potential difference, therefore more part penetration we will cause. But since the kVp range does not need to be as precise as film/screen systems, we have some room to play with technique a bit.
The next logical step in our reasoning is to determine if an increase or decrease in kVp will provide better images. Because the beam physics haven't changed, we can't really lower our kVp compared to optimum film ranges because no matter how magical our computed radiography and digital radiography algorithms are, they are still dependent upon adequate exposure to the image receptor. We can, however, increase our kVp.
How much should we increase? In my personal experience, I have used about 10 kVp higher than film/screen systems for small extremities, and up to 20 or 25 kVp higher than film/screen systems for thick body parts like spine. I could never use 90 kVp on a lateral c-spine with plain film because it would be too gray/washed out. Since digital radiography adjusts the scale of contrast to be ideal for a "lateral c-spine" I have two distinct advantages when using higher kVp with these systems: More uniform part penetration and less radiation dose to the patient.
For part penetration, you will be able to visualize the lower cervical and upper thoracic vertebrae at 90 kVp than at the traditional 70 kVp range. Using the 15% rule, you know that using 90 kVp will require far less mAs (about 25% than at 70 kV), which will significantly reduce the exposure to the patient. Talk about a win-win situation!
I would encourage anyone looking to make changes to their technical factors to consider a higher kV range with dose reduction. You may be tempted to use ridiculously low mAs values like 0.5 mAs on various body parts, but remember the physics... you will get quantum mottle or image noise when there are not enough x-ray photons reaching your image receptor. I would love to hear what you currently use at your facility, and if you try some of these changes, I would also love to hear what you think about them. Whatever changes you make, please keep ALARA principles in mind, and remember to consult your Radiologists.
My facility has Agfa CR 75 system, and I normally add 10k-15kVp for table/wall Bucky exams. For table top, 5-7kVp (e.g., 60kVp @ 2mas for a foot/hand), and for PCXRs, I normally start off with 80kVp for non-grid exams. For example, I routinely use 85kVp for LAT C-Spines unless the PT has a real skinny neck. PA Chests in the Dept. are taken with a kVp of 115/120. I have used 125kVp on morbidly obese PTs for adequate penetration to visualize, somewhat, the spine.
ReplyDeleteFWIW, I don't like the Agfa system due to LgM Exposure range (1.7-2.3). Moreover, the image will be "distorted" and/or show lack of detail (especially S/T) if its over-exposed, and attempting to Window/Leveling it for an optimum image doesn't usually correct problem.
My facility has Agfa CR 75 system, and I normally add 10k-15kVp for table/wall Bucky exams. For table top, 5-7kVp (e.g., 60kVp @ 2mas for a foot/hand), and for PCXRs, I normally start off with 80kVp for non-grid exams. For example, I routinely use 85kVp for LAT C-Spines unless the PT has a real skinny neck. PA Chests in the Dept. are taken with a kVp of 115/120. I have used 125kVp on morbidly obese PTs for adequate penetration to visualize, somewhat, the spine.
ReplyDeleteFWIW, I don't like the Agfa system due to LgM Exposure range (1.7-2.3). Moreover, the image will be "distorted" and/or show lack of detail (especially S/T) if its over-exposed, and attempting to Window/Leveling it for an optimum image doesn't usually correct problem.
Those ranges sound like they would work well with CR. I think a lot of facilities are going with grids for PCXR's for all patients, but the bucky range should be appropriate that you listed.
ReplyDeleteWhat don't you like about the LgM exposure? Also, are you attempting to window/level overexposed images at the QC workstations or on the reading room workstations? It makes a huge difference depending on which you are doing.
Those ranges sound like they would work well with CR. I think a lot of facilities are going with grids for PCXR's for all patients, but the bucky range should be appropriate that you listed.
ReplyDeleteWhat don't you like about the LgM exposure? Also, are you attempting to window/level overexposed images at the QC workstations or on the reading room workstations? It makes a huge difference depending on which you are doing.
Not so convinced about how effective increasing kVp is with digital, and it also varies considerably between systems and detector types. Yes, you can get away with higher kVp's than film screen for many projections, but as kVp is still the primary factor affecting contrast you need to be careful that you are not loosing subtle detail.
ReplyDeleteFor torso imaging, I would recommend adding beam filtration before considering increasing kVp. This will remove more of the photons at energies that make little contribution to the image, and will increase the average beam energy (beam quality). After this, then increase kVp's until just before radiologists start to squeal.
For orthopaedic extremities, a different approach may be required. My efforts at increasing kVp's and decreasing mAs for extremity imaging resulted in noisier images and worse contrast. A recent study from a major DR vendor has shown that for extremity images, if you decrease kVp, and increase mAs, whilst keeping effective dose (not entrance skin dose) to the patient constant, then you can massively increase the contrast to noise ratio. This is the opposite to current thinking!
Not so convinced about how effective increasing kVp is with digital, and it also varies considerably between systems and detector types. Yes, you can get away with higher kVp's than film screen for many projections, but as kVp is still the primary factor affecting contrast you need to be careful that you are not loosing subtle detail.
ReplyDeleteFor torso imaging, I would recommend adding beam filtration before considering increasing kVp. This will remove more of the photons at energies that make little contribution to the image, and will increase the average beam energy (beam quality). After this, then increase kVp's until just before radiologists start to squeal.
For orthopaedic extremities, a different approach may be required. My efforts at increasing kVp's and decreasing mAs for extremity imaging resulted in noisier images and worse contrast. A recent study from a major DR vendor has shown that for extremity images, if you decrease kVp, and increase mAs, whilst keeping effective dose (not entrance skin dose) to the patient constant, then you can massively increase the contrast to noise ratio. This is the opposite to current thinking!
Sure it has its limitations, but raising kVp will reduce dose while still giving you an optimum image. If you're diagnosing a hairline fracture on a finger, you probabaly should be using over 65 kVp (if your generator allows for mAs that low). This is probably one reason you are experiencing image noise with your extremities - you still need adequate exposure to the plate (enough photons) to avoid mottle/noise.
ReplyDeleteThe best way this helps you is in large body parts that usually require a lot of mAs... you should give it a try - use 90 kVp for lateral c-spine or 95-100 for a lateral l-spine (especially for a large patient). Let me know what you think.
As far as beam filtration, that would be great, except that I don't know of any place still using added filtration. I haven't seen that for 10 years or so, and other than a compensating filter, I'm not sure I have seen it used much. It would be interesting to try though.
Sure it has its limitations, but raising kVp will reduce dose while still giving you an optimum image. If you're diagnosing a hairline fracture on a finger, you probabaly should be using over 65 kVp (if your generator allows for mAs that low). This is probably one reason you are experiencing image noise with your extremities - you still need adequate exposure to the plate (enough photons) to avoid mottle/noise.
ReplyDeleteThe best way this helps you is in large body parts that usually require a lot of mAs... you should give it a try - use 90 kVp for lateral c-spine or 95-100 for a lateral l-spine (especially for a large patient). Let me know what you think.
As far as beam filtration, that would be great, except that I don't know of any place still using added filtration. I haven't seen that for 10 years or so, and other than a compensating filter, I'm not sure I have seen it used much. It would be interesting to try though.
Jeremy,
ReplyDeleteYour comments,"Also, are you attempting to window/level overexposed images at the QC workstations or on the reading room workstations?"
Do you mean CR workstation? The answer is yes. Our Agfa Digital Reader has a tendency to overexpose (LgM number is usually 2.4-2.6) during the processing of skull,sinus, occasional UPRT ABD (slender PTs), and other exams. Agfa service technicians have investigated this problem along with coordinating with GE (multiple rooms with Proteus XR/a equipment)to determine if there were equipment issues. Both claimed their respective machines were operating within specified parameters. With that said, the RTs have learned to use -2 density on the aforementioned exams so they're processed without being over exposed (LgM greater than 2.3). Normally, this correction works, but at times, it doesn't so the RT will adjust the images' window/level for best diagnostic appearance at the CR workstation before transmitting to PACS.
Jeremy,
ReplyDeleteYour comments,"Also, are you attempting to window/level overexposed images at the QC workstations or on the reading room workstations?"
Do you mean CR workstation? The answer is yes. Our Agfa Digital Reader has a tendency to overexpose (LgM number is usually 2.4-2.6) during the processing of skull,sinus, occasional UPRT ABD (slender PTs), and other exams. Agfa service technicians have investigated this problem along with coordinating with GE (multiple rooms with Proteus XR/a equipment)to determine if there were equipment issues. Both claimed their respective machines were operating within specified parameters. With that said, the RTs have learned to use -2 density on the aforementioned exams so they're processed without being over exposed (LgM greater than 2.3). Normally, this correction works, but at times, it doesn't so the RT will adjust the images' window/level for best diagnostic appearance at the CR workstation before transmitting to PACS.
It is possible that you may need your Agfa rep to come out and adjust the algorithm a little bit more to your liking. Is this something that's included in your service contract? If not, it still might be worth it to pay to have them come on out and make adjustments. You should be able to 1/2 your mAs to reduce the LgM by 0.3 instead of setting the density to -2.
ReplyDeleteYou could also try avoiding any window/level at the CR workstation. If you adjust it before sending it to PACS, you are removing data from the original image. Even though it might look worse at the CR workstation, reset any window/level that is done prior to sending. Doing this will allow the Radiologist to be able to window/level a lot more if they have the entire data set to manipulate with their superior software and monitors in the reading room.
You may want to try these things out on test phantoms before paying someone to come out to adjust your algorithms.
It is possible that you may need your Agfa rep to come out and adjust the algorithm a little bit more to your liking. Is this something that's included in your service contract? If not, it still might be worth it to pay to have them come on out and make adjustments. You should be able to 1/2 your mAs to reduce the LgM by 0.3 instead of setting the density to -2.
ReplyDeleteYou could also try avoiding any window/level at the CR workstation. If you adjust it before sending it to PACS, you are removing data from the original image. Even though it might look worse at the CR workstation, reset any window/level that is done prior to sending. Doing this will allow the Radiologist to be able to window/level a lot more if they have the entire data set to manipulate with their superior software and monitors in the reading room.
You may want to try these things out on test phantoms before paying someone to come out to adjust your algorithms.
Hi Mr Anonymous,
ReplyDeleteI work for Agfa and I can tell you if you're getting 2.4-2.6 for your LgM then you are OVEREXPOSING!! You should be aiming for 2.1 and a bit lower for chests (say 1.9). No wonder your images aren't too good. To reduce the LgM from 2.4 to 2.1, then half your mAs.
Good luck.
Hi Mr Anonymous,
ReplyDeleteI work for Agfa and I can tell you if you're getting 2.4-2.6 for your LgM then you are OVEREXPOSING!! You should be aiming for 2.1 and a bit lower for chests (say 1.9). No wonder your images aren't too good. To reduce the LgM from 2.4 to 2.1, then half your mAs.
Good luck.