Saturday, September 27, 2008

Efficient PACS Reading

Radiologists are reading more cases every year, and anything that makes our day more efficient is a good thing. In a computer-centric environment, the workday revolves around digital imaging, the monitor, mouse, and keyboard.

A great scripting program called Autohotkey can save you thousands of repetitive mouse clicks and keystrokes a year. If you are a heavy computer user like me, it can make your computing experience significantly better, and save you from repetitive strain injury.

It's simple to learn, if you let your inner geek run wild and free. It can be run off a USB stick, so administrative privileges are not required.

Learn more about it:

Autohotkey in Radiology

Using AHK


Vic David MD

Sunday, September 14, 2008

Heart Surgeons, Abscesses, and Lytic Lesions

Michael DeBakey M.D. died at the age of in July 2008. He was best known as a pioneering heart surgeon, operating on celebrities and commoners alike, saving thousands of lives.

Debakey was a creative surgeon even early in his career. With his mentor Alton Ochsner, he devised a new way to drain subphrenic abscesses, through a transthoracic approach. An abscess is a collection of pus (infected fluid), and is typically found within the soft tissues. In some cases, however, abscesses can also occur within the bone.

In 1832 the surgeon B.C. Brodie described three cases of a chronic abscess within the tibia:

The patients were young adults, and each presented with chronic tibial pain and swelling. Brodie was a good writer, and his description is compelling:

"The lower extremity of the left tibia was considerably enlarged; the skin covering it was tense, and adhered closely to the parts below. The patient complained of a constant aching pain, which he referred to the enlarged bone. Once in two or three weeks there was an attack of pain more severe than usual, during which his sufferings were excruciating, lasting several hours, and sometimes one or two days, and rendering him altogether incapable of following his usual occupations. The pain was described as shooting or throbbing, worse during the night, and attended with such exquisite tenderness of the parts in the neighborhood of the ankle that the slightest touch was intolerable."

Recently, I saw a case of a 12 year-old girl with a lytic lesion in the tibia. Of course, the clinical history on the prescription was a little less eloquent than the description above: "Pain, MRI ankle".

Sagittal T1-weighted and STIR images depict an oval lesion (red arrows) in the distal tibial metaphysis. The lesion crosses the physis (yellow arrow), to involve the epiphysis as well. There is a great deal of marrow edema surrounding the lesion. A close-up of the lesion reveals that the lesion is heterogeneous, with a rind of T2 hyperintensity representing granulation tissue (red arrow) surrounding a relatively hypointense core (blue arrow):

The appearance is highly suggestive of a chronic bone abscess (Brodie's abscess), which was confirmed at surgery. In some cases, the lesion can be difficult to distinguish from a tumor, such as osteoid osteoma. Conventional radiographs and CT are often helpful, and depict a lytic lesion with surrounding sclerosis. This sclerotic response typically has a sharp interface with the lesion, but merges gradually with the surrounding bone (Musculoskeletal Imaging: A Teaching File. Felix S. Chew, Catherine C. Roberts; Lippincott Williams & Wilkins, 2005)

A Brodie's abscess develops when osteomyelitis is contained by the host immune response, but is not cured. The infection is walled off, but remains active. The lesion usually is within the metaphysis, but can occur anywhere. When it occurs in the epiphysis, it can be mistaken for a chondroblastoma.

Nearly 200 years have flashed by since Brodie's original description, but chronic bone abscess remains an important medical diagnosis. These lesions are eminently treatable, once the diagnosis is established.

Vic David MD

Friday, September 12, 2008

Diving and the Thumb CMC Joint

Platform diving is acrobatic and beautiful:

Photo by shutterhack

Inevitably, there are some serious g forces exerted on the human body, especially if the rapid deceleration is less like a knife-like entry into the water, and more like one of my dives, the classic belly-flop:

Photo by j. cliss

Recently, a hand surgeon called me and told me about a set of patients who are platform divers, with intractable pain at the base of the thumb, at the thumb carpometacarpal (CMC) joint. She remarked that it would be nice to get some imaging information about the ligamentous structures supporting this joint, particularly along the dorsal aspect of the joint, a region that is difficult to visualize at arthroscopy.

There is not a great deal written about MR imaging of the thumb CMC joint. One good article was written by Connell et al.; there is also a Wandering Radiologist blog post.

There is some variability in how these ligaments are named, but if we follow the convention of Connell et al., there are four main ligaments:

1. Anterior oblique (beak) ligament- extends from the trapezium to the first metacarpal base, on the volar side of the joint.

2. Posterior oblique ligament- extends from dorsoulnar trapezium to the first metacarpal base.

3. Dorsoradial ligament- extends from the dorsoradial aspect of trapezium to the first metacarpal base. This ligament is reinforced by the abductor pollicis longus tendon.

4. Intermetacarpal ligament- runs from the radial base of the index metacarpal to the ulnar aspect of the thumb metacarpal base.

The combination of direct coronal and
oblique sagittal images of the CMC joint visualize the first three of ligamentous structures quite well, at 1.5 Tesla. The intermetacarpal ligament can also be seen, but is more wispy in character.

Consecutive oblique sagittal intermediate-weighted images of a normal volunteer:

The anterior oblique ligament (red arrows) is well seen, as is the posterior oblique ligament (green arrows). The dorsoradial ligament is depicted on this next set of images:

(A) Oblique sagittal and (B) direct coronal images delineate the normal dorsoradial ligament (pink arrows).

Direct coronal images are usually the best way to visualize the anterior oblique (beak) ligament:

The MRI technologists I work were instrumental in determining the optimal imaging planes for us to visualize these structures. This post illustrates once again the vital role the MR technologist plays in producing high-quality MR images.

Vic David MD

Friday, September 5, 2008

Movies and ACL Tears

Movement is essential to life. Our brain is hard-wired to detect motion, and we respond to video much more than static pictures. Consider these two different depictions of a horse galloping:

Now, the video version:

Clearly, the video version is more appealing. Moreover, the video version contains important information that the still picture does not. This video, made by Eadweard Muybridge in 1878, was the first motion picture ever made. Muybridge was commissioned by Leland Stanford (California governor/ Stanford University) to answer a popularly debated question of this era— are all four of a horse's hooves ever off the ground at the same time while the horse is galloping? Muybridge's time-motion photography proved they were off the ground simultaneously. Video analysis had answered a scientific question for the first time in history.

Although cross-sectional imaging has revolutionized radiology, video representations of anatomic detail are the exception in MRI and CT. Volume-rendering techniques enable us to depict 3D data sets, but these techniques are diffusing into clinical practice in a sluggish fashion.

Almost everyone reads off PACS now, which is clearly superior to reading from film for many reasons. One advantage of PACS is the ability to rapidly scroll through a stack of images. This video-like representation of data sets will sometimes enable you to quickly recognize abnormalities that are subtle and easily missed on static images.

One example is the analysis of the anterior cruciate ligament (ACL), a key stabilizer of the knee. It is important to recognize tears of this structure, as the presence of an ACL tear will often change the treatment algorithm. There have been many excellent articles written about the MRI analysis of ACL tears, and these tears are usually easy to recognize on MRI.

The key word here is usually. These tears can be subtle in some patients. The ACL should be examined in the sagittal, axial, and coronal planes to maximize your accuracy. One common error is to rely solely on the sagittal plane to determine the status of the ACL. The axial and coronal planes will often yield additional information about this critical structure.

In the coronal plane, the normal ACL has an oblique course, arising from the lateral femoral condyle, and coursing anteriorly and medially before inserting on the anterior tibia. In the following figure, the green arrow depicts the normal oblique course of the ACL:

Any significant deviation from this normal oblique course of the ACL on coronal images is abnormal.

Here is a movie of a stack of coronal images, depicting a normal ACL. Note how the ACL tracks from the top left of the figure to the bottom right, following its normal oblique course:

Next, let's look at a sagittal image from a 37 year-old patient with knee pain:

The ACL is visualized (red arrow) throughout most of its course. The femoral origin (green arrow) is hazy , but on the adjacent image, this part of the ACL can be seen:

The ACL looks "funny", but it would not shock me if a busy reader passed this off as volume averaging, or a partial tear of the ACL.

A complete analysis of the ACL, however, requires scrutiny of the axial and coronal planes as well, particularly in complex cases. Here is a movie of a stack of coronal images from the same patient. T
he ACL (red arrow) no longer follows its normal oblique course through the intercondylar notch. Its path is now curvilinear, and the inferior aspect of the ACL is deviated laterally:

With these additional images, we can confidently diagnose a complete tear of the ACL, which was subsequently confirmed at arthroscopy.

One of the less-advertised benefits of PACS is its ability to give us video-like representations of the imaging data.

Vic David MD