Diagnostic Imaging brings you daily coverage of SIIM 2021 virtual annual meeting.

SIIM 2021 Virtual Annual Meeting: May 24 - 27, 2021

Early PET/CT imaging after transarterial radioembolization (TARE) for liver cancer could identify early who will respond to treatment and inform further treatment steps, a new study in France found.

The single-center retrospective study, published in 

, evaluated the utility of PET/CT using fluorodeoxyglucose (FDG) or 18F-FCH (18F-fluorocholine) to predict six-month treatment response and overall survival (OS) after TARE in patients with 

“Early posttreatment evaluation with PET/CT could help more reliably identify true nonresponders after TARE, which in turn could prompt early adapted therapeutic management,” the study authors, led by Edouard Reizine, MD, of Henri-Mondor University Hospital in Créteil, France, wrote.s

The study included 37 patients with HCC treated by TARE at Henri-Mondor University Hospital in Créteil, France, between February 2016 and August 2018. Those with portal vein tumor invasion or invasion or with unifocal or multifocal HCC not eligible for local therapies or transarterial chemoembolization (TACE) were included. Mean age was 67 and the majority (34) were men. FDG PET/CT and 18F-FCH PET/CT examinations were performed before treatment and 4 to 8 weeks after treatment.

Treatment response was assessed using modified response evaluation criteria in solid tumors (mRECIST).

“In patients with HCC treated by TARE, early posttreatment PET/CT response exhibited 100% sensitivity and 100% specificity for six-month mRECIST response, in comparison with 67% sensitivity and 100% specificity of one-month mRECIST response; early postteatment PET/CT response was also a significant independent predictor of OS (HR: 0.24, 95% CI: 0.08-0.76, 

On the PET/CT before treatment, 28 (76%) patients were FDG-positive, 15 (41%) FCH-positive 6 (16%) both FDG-positive and FCH-positive.

At one month after treatment, 12 of 28 (42.9%) FDG-positive HCCs exhibited early response by FDG PET/CT; and 7 of 15 (46.7%) FCH-positive HCCs exhibited early response by 18F-FCH PET/CT. At six months, 12 (52%) patients exhibited mRECIST response. 

“Given the high percentage of HCCs that were FDG-positive, metabolic evaluation before and after TARE could be conducted by FDG alone in the large majority of patients,” the authors wrote. “Nonetheless, performing pretreatment PET/CT with both FDG and 18F-FCH helps detect additional HCCs not identified by conventional imaging, and the combination of the two tracers achieves very high sensitivity for HCC detection. Hence, imaging using both radiotracers may maximize the number of patients with HCC who benefit from metabolic evaluation.”

Limitations of the study include that it was retrospective, included only 37 participants out of 73 patients, and participants received a range of additional treatments after TARE. The authors also noted that 18F-FCH is not readily available.

The 5-year survival rate for HCC, which is the leading cause of liver cancer deaths,

Predicting how patients will respond to treatment has been a focus of providers. One 

 found that applying a deep convolutional neural network improved the ability to determine how a patient will respond to Y90-RE therapy.

 found that more frequent ultrasound screening in patients who are at high risk for developing HCC was associated with earlier HCC detection, decreased mortality, and improved quality of life.

Articles

Patients with hepatocellular carcinoma may benefit from early PET/CT imaging after transarterial radioembolization to predict treatment response and survival, a new study found.

Early PET/CT Can Help Predict Treatment Response, Survival of Patients with Liver Cancer

Pranav Chitkara, M.D. is a Board Certified, Fellowship Trained Radiologist. Dr. Chitkara completed medical school at The University of Texas Southwestern and his radiology residency at State University of New York in Brooklyn. Dr. Chitkara also completed his fellowship in Musculoskeletal Imaging at Cleveland Clinic.

Infant with Prenatally Diagnosed Abnormality on Ultrasound

40-week-old male with a prenatally diagnosed abnormality on screening ultrasound. What is the diagnosis?

Image IQ Quiz: Infant With Prenatally Diagnosed Abnormality on Ultrasound

In men with intermediate- to high-risk prostate cancer, 68Ga- prostate-specific membrane antigen (PSMA)-11 positron emission tomography (PET) imaging may miss small pelvic nodal metastases, according to a study recently published in 

“A PSMA PET scan negative for pelvic nodal metastasis does not indicate that a pelvic nodal dissection is not required,” wrote Thomas Hope, M.D., of the University of California in San Francisco, and colleagues.

While pelvic nodal metastases detected at radical prostatectomy are associated with biochemical recurrence after 

, improved detection of metastatic disease is needed prior to definitive therapy.

This prospective phase 3 trial included 764 patients, median age 69, with intermediate- to high-risk prostate cancer considered for prostatectomy from December 2015 to December 2019. The primary end point was the sensitivity and specificity of 68Ga-PSMA-11 PET imaging for the detection of pelvic lymph nodes compared with histopathology on a per-patient basis. Each scan was read by three blinded independent readers.

Of the 764 men who underwent a 68Ga-PSMA-11 PET scan for primary staging, 277 (36%) underwent prostatectomy with lymph node dissection. Based on pathology reports, 75 of 277 men (27%) had pelvic nodal metastasis. For pelvic nodal, extra-pelvic nodal and bone metastatic disease, 68Ga-PSMA-11 PET results were positive in 40 of 277 (14%), two of 277 (1%), and seven of 277 (3%) of patients, respectively. The sensitivity and specificity for pelvic nodal metastases were 0.40 and 0.95, respectively, compared with histopathology.

“This study is the largest prospective study using PSMA PET at time of initial staging,” the authors said. The results of this study were used to support the recent U.S. Food and Drug Administration approval of 68Ga-PSMA-11 PET at initial staging.

, Joseph Osborne, M.D., Ph.D., of Weill Cornell Medicine in New York, and colleagues wrote that the results are similar to those from the prior 

 evaluating PSMA PET diagnostic performance of 18F-DCFPyL. “There is a clear message from both trials: clinicians taking care of patients with high-risk prostate cancer being assessed for prostatectomy can use a positive PET scan as a true positive,” they wrote. “Whereas a negative scan cannot be used to exclude disease or inform nodal dissection (both studies had a diagnostic sensitivity near 40%).”

Osborne and colleagues added that the latest study “has provided a road map for how preintervention PSMA PET imaging will guide the appropriateness of radical prostatectomy for the referring urologists. As such, these results are practice changing for the nuclear medicine physicians, urologists and medical oncologists who will manage this cohort of patients.”

Clinicians taking care of patients with high-risk prostate cancer being assessed for prostatectomy can use a positive PET scan as a true positive, but a negative scan cannot be used to exclude disease or inform nodal dissection. 

PET Imaging Helps Guide Patient Selection for Radical Prostatectomy  

Can you diagnose the finding seen on this panorax?

Image IQ Quiz: What Do You See On This Panorax?

I never really liked the expression, “pay per click,” as used in radiology. It makes perfect sense in other venues, like advertising. There, a bit of revenue is generated for every click the ads receive. I don’t know the fine details, but imagine things are pretty constant; click equals $X.

In our line of work, that’s not the case…or, if it is in some practices, I’ve never seen it. The closest thing that happens is in “eat what you kill” or other such productivity-based models, where a rad’s comp increases each time he clicks to sign a report. Or uses a click-equivalent to sign off, such as a voice-command.

The issue I have with that? Long before a rad gets to that point, s/he’s clicked more than a few times: To open the case, to manipulate images, to sift through prior reports, to move around in and edit the dictation, sometimes to communicate with referrers and support-staff. It’s really “pay per click

And that’s a best-case scenario: Far from rarely, a rad will open a case and do more than a little bit of clicking around before realizing that the case can’t be read as-is. Images, priors, or history missing, the case needs to be read by someone else, or the rad just gets interrupted and someone else takes the case while s/he’s away. 

 pay-per-click going on there. Also none for clicking when a rad has to revisit a signed report for addenda, questions from clinicians, etc.

Then, there’s the little matter of different cases not having the same worth: Even if it were truly pay-per-click, clicks for XR, CT, MR, etc. understandably pay different amounts. And if you really want to get into the weeds, a given study-type will vary in its yield depending on where the patient got imaged, what payor is involved…

That doesn’t stop some radgroups from trying to bean-count and assign case-values (“work units,” for instance) for the reading rads, of course. They might simplify, and use the average value of a given case-type that the group sees from all its sources. Both for simplicity, and so the rads don’t go nuts lest they see that the chest-CT they got to read is worth less than a chest on the worklist right next to it.

Having been on the receiving-end of this for a chunk of my career, I can tell you that the values never seem to perfectly reflect the time/work balance. For instance, a conscientiously-reading rad who winds up with mostly XR on his list, even if reading efficiently and taking minimal breaks, really can’t keep work-unit pace with one who got a bunch of MR. Even if the MR-rad is far more leisurely.

That’s not all to be blamed on the radgroups in which this happens. Some give everybody access to the same worklists, instituting workflow-rules meant to curb “cherrypicking.” The idea being that, if all rads in the fullness of time get access to roughly the same case-mix, everyone will be hurt by XR and helped by MR equally.

Still, people find ways to game systems, and docs tend to be a smart bunch. If there’s a trick, they’ll find it sooner or later. Also, if the folks making the rules are reading cases like everyone else, whatever rules they make will tend to favor themselves; that’s just human nature.

Another factor: Not every rad has the same skillset. Some might appropriately beg off higher-value exams like MSK MR simply because they know they wouldn’t do a good job. Some older rads might never have gotten comfortable with 

MR. Of course, nobody’s stopping them from learning something new…but not everybody’s going to do that, especially later in their careers. Finally, sometimes subspecialty reads are requested/needed, so certain high-value exams aren’t accessible to everyone.

Another lever some radgroups can adjust is boosting the value of cheaper exams, so rads receiving more of them won’t be at such a disadvantage. Groups can finagle a little bit, robbing Peter-MR to pay Paul-XR…but that can be tricky when you don’t know how many of each type of study your group will be receiving. A month in advance, let alone a year. If you wind up with more XR and fewer MR than you expected, you might find yourself in a fiscal hole.

Every level of complexity that gets added to per-click schemes introduces new hassles in bean-counting. Ways for the unforeseen to sneak in and mess things up. Wrinkles that will create winners and losers amongst the rads (or seem to, and someone perceiving he’s being done out of something will make just as much noise as if he genuinely has). Unsurprising, then, that more than a few who’ve worked as per-click rads…or overseen them…have less-than-fond memories of the experience.

Meanwhile, a lot of the rads flocking to per-click groups do so because, philosophically, it appeals to them. A lot have previously experienced salaried jobs where everybody’s supposedly on their honor to do their fair share of the work…but that doesn’t seem to pan out. Compensation fails to be proportional to productivity. “Pay per click,” or “eat what you kill,” goes a long way towards preventing that.

Hybrid plans might be most capable of satisfying all comers. A set salary, or hourly rate, coupled with the notion that someone is keeping an eye on who’s actually getting the work done. That could be in the form of a monthly, quarterly, or even annual incentive-payment…or simply periodic performance-reviews, with outliers at both ends of the spectrum being recognized and perhaps rewarded (or sanctioned).

The nice thing about such a system is that, with less of one’s total comp on the line, keeping tabs on every little click stops being so important. Rads have less of a motivation to cherry-pick, game the system, or complain about it. And there doesn’t have to be quite as much bean-counting aggravation behind the scenes.

Still, one doesn’t want to completely lose track of who’s doing the work. Tune in next time for my “double-curve” solution.

There doesn’t have to be quite as much bean-counting aggravation behind the scenes.

All Clicks Are Not Created Equal

Despite what has already been an illustrious career in health care, now is not the time to sleep on 

 (CT). Fifty years in, CT remains a cornerstone of medical imaging, aiding not only in diagnosis of disease or trauma, but also in helping to plan and guide therapeutic interventions as well as monitor their outcomes. 

Withstanding the test of time and new technology, CT has also weathered bouts of criticism over the years. Although questions about overuse and radiation dose exposure have at times hindered its role, the risk/benefit ratio conversation nearly always tips towards the benefits of the modality, with its ability to quickly answer questions that might otherwise lead providers and patients down a road filled with a multitude of less specific—and so-called less ‘risky’—tests.

To help address some of these questions and improve education around dose, the American Association of Physicists in Medicine (AAPM) in 2010 formed the 

, which provides several educational and informational resources, including a CT Dose-Check, up-to-date CT imaging protocols for both adult and pediatric patients, as well as CT dose and automatic exposure control educational materials. 

To learn more about the history of CT and how to optimize its current use, Diagnostic Imaging spoke with Cynthia H. McCollough, PhD, Brooks-Hollern Professor and professor of medical physics and biomedical engineering at Mayo Clinic. McCollough, who is also director, CT Clinical Innovation Center and X-ray Imaging Core, discussed how CT is currently utilized and highlighted the many ways we can optimize its use in different clinical scenarios, especially when taking into account the technology that allows us to personalize the experience for each patient. 

In further commemoration of the anniversary, McCollough helped to put together an online gallery through the AAPM that illustrates the amazing history of CT. View the gallery 

Videos

Diagnostic Imaging sat down with medical physicist and CT expert Cynthia H. McCollough, PhD, to learn more about the history of CT and its expanding potential in medical imaging. 

50 Years of CT: Then, Now, and What’s to Come 

Just in time to commemorate the 50th anniversary of the first 

 (CT) scan, the US FDA announced it has granted 510(k) clearance to Siemens for its new NAEOTOM Alpha CT device, the first to utilize photon-counting.

The emergence of photon-counting detectors represents the next significant enhancement to CT use, allowing for more detailed information about the patient to be obtained while creating images with less information that is deemed not useful for review and analysis. 

“About 15 years ago, work on photon-counting and its clinical vision started at Siemens Healthineers. We always believed in the tremendous clinical value and relentlessly worked on it together with our partners. We are excited that we have received FDA 510(k) clearance,” says André Hartung, Head of Diagnostic Imaging at Siemens Healthineers, in a statement.

By measuring each individual x-ray that passes through a patient’s body, photon-counting detectors can produce detailed three-dimensional images that can ultimately aid in making a diagnosis or treatment-related decisions. 

“Computed tomography is an important medical imaging tool that can aid in diagnosing disease, trauma or abnormality; planning and guiding interventional or therapeutic procedures; and monitoring the effectiveness of certain therapies,” said Laurel Burk, Ph.D., assistant director of the Diagnostic X-ray Systems Team in the FDA’s Center for Devices and Radiological Health, in a statement. “Today’s action represents the first major new technology for computed tomography imaging in nearly a decade and underscores the FDA’s efforts to encourage innovation in areas of scientific and diagnostic progress.”

The technology ultimately processes x-rays differently; transforming the current two-step process in which X-rays are converted into visible light, which is then detected by a light sensor, into a single-step process in which X-ray photons are directly converted into digital electrical signals that are counted, ultimately producing an image with greater clarity and contrast. 

The approval represents the first major technological advance in CT in nearly a decade.

FDA Clears Siemens’ New CT Scanner With Photon-Counting Detector 

Image IQ Quiz: Patient Presents With Ankle Pain

Last time around, I dwelled on the so-called “pay per click” model of radiology-compensation. As then mentioned, I’ve seen some love it, some loathe it, and a lot of folks scattered across the spectrum in-between.

Most of the non-polar sentiments include a recognition of tradeoffs: Yes, it can be a bean-counting, headache-making, quarrel-inducing mess, but what else are you going to do? Just pay people a set amount and hope they’ll do a reasonable amount of work? Not noticing/caring if they seem to be pulling harder than others on the team who earn just as much, if not more?

A “happy medium” system would be one in which rads knew they could go about their work without fretting over the daily vagaries of their RVU tally…but nevertheless assured that their productivity would be on the record. And that slackers on the team wouldn’t go undetected.

Leadership would, similarly, have an idea who was shouldering how much of the group’s burden…without having to constantly tinker with complex schemes of differing exam-values, deterrence-mechanisms for “cherrypicking” of easier or higher-value cases, etc.

Tracking RVUs, or even RVU-related “work units,” doesn’t really get at this. You certainly wind up with a bunch of numbers to apply to your rads, numbers which can be put on tables and charts for periodic review. You can make sophisticated-looking curves to show a rad’s productivity over time, and stack them up against other rads, throw in standard-deviation ranges, etc.

Most rads who’ve been working more than a little while have seen their performance tracked. Even if you’re not in a pay-per-click situation, the folks in charge need to know if you’ve been generating enough revenue to warrant whatever salary or hourly rate you receive. Also how to fairly apportion bonuses and consider requests for raises.

But as long as all of that is based on curves, bar-charts, etc. derived from your RVU total, the per-click issues are a big part of the picture. Somewhat paradoxically, I think adding a 

 curve would simplify things, for groups preferring to operate via salaries or hourly rates:

By all means, keep tracking RVUs. After all, if that’s how your group is receiving its revenue, it doesn’t make sense to pretend otherwise. Rads who are doing the heavy lifting to make that revenue happen should definitely be recognized, if not rewarded.

But alongside the RVU curve, have another measurement, truer to the simplicity that “per click” implies: A curve of the raw number of exams a rad has read.

This way, the rad(s) who are dealing with the neverending onslaught of low-paying studies like XR…work that has to get done one way or another…can be tracked for the service they are providing. One important facet of which is running interference so the RVU-sluggers can focus on their MR.

Maybe the rads who would look better on the second curve don’t have the skills to read many of the higher-ticket scans. Maybe they prefer lower-tech modalities. Or maybe their other responsibilities (fluoro, diagnostic mammos, fielding phone calls, etc.) interrupt them so much, it’s just easier to squeeze “bite-sized” studies like XR in between.

If such rads know that they will be valued for taking care of the mountain of XR—despite that mountain’s reimbursement-worth being unimpressive next to a reasonable daily haul of MR and CT—they can go about their work without constantly being tempted to sift through the list for brainscans to shore up their RVU-numbers.

When it’s time for performance reviews, each rad’s two curves can then be seen as equally-important parts of the big picture. Total exams in bottom quartile, but you’re tops RVU-wise? You did your part. RVUs low, but study-volume high? You get praise, not browbeating.

Sort of like being on a track-n-field team: A sprinter is used/valued for what he can do in the 100-meter dash. A distance-runner is regarded for his performance in the 5k. You wouldn’t put either guy in the other’s event and expect good results.

Of course, rads aren’t quite as pigeonholed as all that. Some might do well if told, each day they came into work, whether they should be going for case-volume or RVUs. Or self-select which suits them at any given moment, depending on their mood or circumstances around them.

Obviously, not everybody can be focused on the same curve; the work all needs to be done. Some coordination of efforts via assignment of roles might be needed to avoid all the XR getting gobbled up early in the day, lest there be dozens of RVU-heavy scans that only a fraction of the group can take care of. Or the opposite: Nothing but XR left on the lists at 4pm, and while some of the volume-centric rads industriously plug away, the “I wish I had some MR” rads reluctantly poke through a few cases at a snail’s pace.

 of a rad’s curves are relevant, rather than just choosing one to be his personal yardstick. A top-quartile RVU-rad with bottom-quartile volume, for instance, might be considered as being on the same level as a top-volume/bottom-RVU member of the team…as well as a rad who’s dead-center for both RVU and case-number.

Meanwhile, a rad in the 75th percentile on both of his curves might be considered something of a hero, and bonused appropriately. And a rad whose two curves were in the bottom 10% might be approached for remediation…unless, of course, that rad had other roles of value such as in administration.

Because not everything shows up on curves, no matter how many you’re willing to draw.

Clicks and Curves

Michelle Alexander has over 20 years experience as a registered sonographer and greater than 10 years in the development of strain elastography, point shear wave elastography and real-time shear wave elastography. Michelle has done extensive global scientific research with all types of elastography in the areas of liver, prostate, breast, thyroid, kidney, cervical and musculoskeletal applications alongside luminaries in Europe, China, Canada and the United States. In addition, she holds a patent in the United States and Europe for elastography visualization

Breast cancer is more common than most people would like to admit. In the U.S., a woman has a 1 in 8 chance of developing breast cancer during their lifetime.

That’s why the American Cancer Society recommends annual mammograms for women beginning at age 40, and even earlier for those with more risk factors.

 But scheduling and receiving a mammogram can be anxiety provoking, especially if a woman has a family history of breast cancer. If the medical professional should discover something during the exam that warrants further examination, the next step is usually an ultrasound of the breast to help clarify what is seen on mammography and if a biopsy is needed or not. The days spent waiting for biopsy results can be terrifying, so if it can be avoided it can greatly improve a patient’s wellbeing.

Luckily advancements in image quality and advanced technologies in ultrasound devices such as shearwave imaging can help. Shearwave elastographic features provide qualitative and quantitative information that can be added to mammographic BIRADS analysis, allowing sonographers and radiologists to improve specificity of breast ultrasound mass assessment.

 These tools provide the physician with additional information, and possibly eliminate the need for a biopsy.

I’ve seen these benefits first-hand. In my experience, patients will go straight from the first ultrasound scan to a biopsy if suspicious abnormalities are detected within the breast tissue on mammography and ultrasound. In order to change the patient care cycle, giving a patient a second ultrasound prior to a biopsy may help them avoid un-needed tests. I was once performing a pre-biopsy ultrasound with a Samsung RS85 Prestige® and the device’s 2D and S-Shearwave Imaging™ feature showed that the lump was just an intramammary lymph node. This meant there was no reason to biopsy it, so it was canceled.

Another imaging tool that simplifies and enhances the screening process is a feature called S-Detect™. The semi-automated reporting feature tool analyzes selected lesions and is able to define margins clearly so the user knows a more precise distance from the nipple and depth of the lesion. This is very helpful for documentation and provides clear information on follow-up visits. There is also an ergonomic benefit. In many cases, a sonographer is reaching across the patient’s body to perform a scan. The automated nature of S-Detect reduces the time a sonographer spends on the breast scanning, which therefore reduces time spent in a very uncomfortable position, causing shoulder and arm fatigue. Lastly, it may also give those who are questioning their judgement an extra layer of confidence.

As time goes on, ultrasound manufacturers are moving beyond 2D and 3D imaging, and into the world of advanced intelligence (AI) to construct new ways to display, interpret and understand what is going on inside the human body. S-Detect is a good example of this. Breast ultrasound has become an increasingly common and crucial tool used by medical professionals to determine abnormalities faster and with more confidence. In fact, because image quality and software has advanced to where it has, practitioners have even begun using ultrasound as a first modality to avoid exposing a woman to small amounts of radiation.

When choosing an ultrasound device, medical professionals should consider not only the image quality but also the software and hardware features available in order to improve both their own experience, but, more importantly, the experience of the patient.

https://www.cancer.org/cancer/breast-cancer/about/how-common-is-breast-cancer.html

 Recommendations for the Early Detection of Breast Cancer 

https://www.cancer.org/cancer/breast-cancer/screening-tests-and-early-detection/american-cancer-society-recommendations-for-the-early-detection-of-breast-cancer.html

3. Berg WA et. al; BE1 Investigators. Shear-Wave Elastography Improves the Specificity of Breast US: The BE1 Multinational Study of 939 Masses. 

. 2012 Feb;262(2). doi: 10.1148/radiol.11110640. PMID: 22282182.

Advancements in image quality and advanced technologies in ultrasound devices such as shearwave imaging can help.

Image IQ Quiz: Patient Presents With Ankle Pain

Clicks and Curves

The Importance of Ultrasound When Determining Suspicious Abnormalities in Breast Tissue