Case-based Introduction

A 50-year-old patient with a history of liver disease presents to the ED with sudden onset generalized weakness and abdominal pain. He has soft but stable blood pressures, and appears unwell. There is diffuse abdominal tenderness. His bedside ultrasound shows some free fluid in Morrison’s pouch, but you’re unsure if it is ascites or blood. His physical exam is concerning enough for you to get a CT abdomen and pelvis.

Although this patient has a history of liver disease, it’s tempting to assume the fluid is ascites. However, patients with cirrhosis are at high risk for coagulopathy, and consequently, spontaneous intra-abdominal hemorrhage (i.e. ruptured gastric varix). And, of course, a patient can have both ascites and intra-abdominal bleeding simultaneously. 

Labwork can add some pieces to this clinical picture, as would diagnostic paracentesis if an adequate “fluid pocket” can be identified. However, Hounsfield Units (HU) are built into most imaging software and can be used to measure the radiodensity of the material for simple, fast, and non-invasive identification.

Hounsfield Unit

Hounsfield units are a relative measure of radiodensity and can aid in the distinction of materials on CT, from soft tissue to blood to various fluids.

All pixels range between an assigned Hounsfield unit of -1000 (air, black) and +3000 (dense bone, white).

Attenuation is largely dependent on the iron content of blood; therefore, plasma (iron-poor) will be hypodense and be associated with a much lower Hounsfield Unit than active bleeding. Uncoagulated blood typically measures 30 to 45 HU.1,2 Clotted (or concentrated) blood measures higher at 60 to 100 HU.3 Separated serum plasma is closer to water at 0 to 20 HU. Finally, ascites also has a Hounsfield measurement of around 0 to 20 HU.4

Because the HU of plasma and ascites are in the same range, this technique may not be helpful in cases where the CT is performed more than 24 hours after the bleeding has stopped. After 24 to 48 hours, the blood separates into serum plasma and clotted blood tends to settle in the dependent portions of the abdomen. However, if bleeding has stopped for 24 hours, it also may not be as important to identify whether the intra-abdominal fluid is ascites versus blood. 

Back to the Case

Below is a side-by-side comparison of the CT image from our earlier case and a different patient’s CT scan with similar-appearing free fluid in the abdomen – one is ascites and one is hemoperitoneum. Using Hounsfield units, we can easily determine that the image from our case is likely hemoperitoneum (79.14 HU) and the image on the right is that of ascites (13.79).

 

Pitfalls

Ascitic fluid can show enhancement on contrast-enhanced CT. However, it is more apparent on delayed contrast-enhanced studies and thus is less likely to cause a false positive for hemoperitoneum on a typical CT with contrast.5 

As previously mentioned, old hemoperitoneum may be read as ascites due to the HU of serum plasma. In these instances, clotted blood may be observed in the dependent portions of the abdomen which is why it is always important to get both a CT abdomen and pelvis when suspicion is high. Furthermore, the HU technique is meant to be another piece of the puzzle and should not be the ultimate decision point on whether to further work up an unstable patient with free fluid in the abdomen. 

How To

To use Hounsfield Units in the McKesson PACS system, scroll to the image you are interested in, right-click, select Annotate, then select Elliptical ROI, then select the area you would like to view the units for.

More Examples

Browse the following images of a ruptured AAA including contrast extravasation (very hyperdense), acute bleeding (hyperdense), and soft-tissue/fat. 

Although one may not need Hounsfield Units to understand the scenario pictured, these examples capture the range and utility of the Hounsfield Unit for the Emergency Medicine physician – when it may be useful identifying ascites from blood or identifying old clot from active bleeding. Even if you don’t plan to use Hounsfield Units ever again, these images can help illustrate the concept of relative density and attenuation.

Special Thanks

Dr. Patrick Hammill, Diagnostic Radiology

References

1 DenOtter TD, Schubert J. Hounsfield Unit. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2020.

2 Levine, C D, U J Patel, P M Silverman, and R H Wachsberg. “Low attenuation of acute traumatic hemoperitoneum on CT scans..” AJR. American journal of roentgenology 166.5 (1996): 1089-93.

3 Federle, M P, and R B Jeffrey. “Hemoperitoneum studied by computed tomography..” Radiology 148.1 (1983)187.

4 Cooper, C, P M Silverman, W J Davros, and R K Zeman. “Delayed contrast enhancement of ascitic fluid on CT: frequency and significance..” AJR. American journal of roentgenology 161.4 (1993): 787-790.

5 Wise, S. W., J. H. DeMeo, and R. F. Austin. “Enhancing ascites: an aid to CT diagnosis.” Abdominal Radiology 21.1 (1996): 67-68.

The following two tabs change content below.

nicanthony

Associate Editor at County EM Blog
Nicole Anthony is a Kings County/SUNY Downstate EM Resident in the Class of 2023 whose prior life included EMS, a failed app, and a Creative Writing minor. Most of her heart is in Prague, but you can also find a part of it in the 2 Hallway column.

Latest posts by nicanthony (see all)


nicanthony

Nicole Anthony is a Kings County/SUNY Downstate EM Resident in the Class of 2023 whose prior life included EMS, a failed app, and a Creative Writing minor. Most of her heart is in Prague, but you can also find a part of it in the 2 Hallway column.

2 Comments

Robby · June 2, 2020 at 9:38 pm

Great post! If I understood you correctly, you mentioned that contrast can enhance ascites to give it a similar HU as blood. So, in your hypothetical case when hemotympanum is on the ddx, what CT would you choose? I imagine if you’re worried about hemoperitoneum that you would order a CT with IV contrast or CTA. I think there may be other clues to support ascites vs blood. For example: in a CTA you would be look for contrast extravasation to show if there is an arterial (delayed), or venous (delayed) source of bleeding, correct?

    nicanthony · June 3, 2020 at 10:48 am

    Great question! I should have clarified that ascites will only enhance 1-3 days after contrast administration, so as long as the patient doesn’t have a history of recent contrast administration, ascitic fluid enhancement shouldn’t be an issue on the patient’s current CT w/con.

Leave a Reply

Avatar placeholder

Your email address will not be published. Required fields are marked *

%d bloggers like this: