Offering the promise of more efficient reporting of findings from magnetic resonance imaging (MRI) of the brain, the Maestro Brain Model has garnered 510(k) clearance from the Food and Drug Administration (FDA), according to the software’s manufacturer ClearPoint Neuro.

Combining deformable surfaces with active shape models and machine learning, the Maestro Brain Model can provide neuroradiologists with 

automated labeling as well as shape and volumetric quantification of brain structures on MRI scans

ClearPoint Neuro said the anatomical segment analysis tool emerged from research examining volumetric and shape abnormalities caused by mild traumatic brain injuries. Noting cross-validation of the technology with more than 1,000 MRI scans, the company noted the Maestro Brain Model offers accurate and highly reproducible results.

The company says future applications of the Maestro Brain Model may facilitate targeted treatment of brain injuries

“Our plan is to quantify drug delivery using intraoperative imaging and simulate patient-specific infusions in targeted brain regions,” explained Lyubomir Zagorchev, a vice-president of Clinical Science and Applications at ClearPoint Neuro. “The unique shape representation in (the Maestro Brain Model) will provide reproducible lead placement for deep brain stimulation and micro electrode recording. Surface meshes of segmented anatomical regions will define safety zones and optimal trajectories for patient-specific laser ablations.”

Articles

The Maestro Brain Model reportedly provides automated identification, quantification and labeling of brain structures on magnetic resonance imaging (MRI).

FDA Clears AI-Enabled Software for Streamlining Brain MRI Assessment and Reporting

Could artificial intelligence (AI) enhance the ability of clinicians to detect prediabetes and diabetes?

Findings from a recently published study show that 

adjunctive use of DiaBeats, a machine learning-based algorithm, with electrocardiogram (ECG) recordings

 may provide an alternative noninvasive imaging approach to screening for prediabetes and diabetes.

“In theory, our study provides a relatively inexpensive, non-invasive, and accurate alternative (to current diagnostic methods) which can be used as a gatekeeper to effectively detect diabetes and prediabetes early in its course,” wrote study co-author Anoop R. Kulkarni, Ph.D., who is affiliated with Innotomy Consulting in Bengaluru, India, anbd colleagues. “Nevertheless, adoption of this algorithm into routine practice will need robust validation on external, independent data sets.”

As the intrinsic risk of prediabetes has become increasingly recognized by the medical community, identification of low-cost, noninvasive methods for diagnosis or risk stratification for prediabetes and future diabetes has become paramount. With this in mind, a team from the Late Medical Research Foundation in Nagpur, India and M&H Research LLC sought to assess whether leveraging machine learning could identify prediabetes or diabetes from noninvasive cardiovascular imaging as the cardiovascular system bears some of the earliest signs of damage from the diabetic process.

To do so, investigators designed their study as an analysis using data from the Diabetes in Sindhi Families in Nagpur (DISFIN) study. A cross-sectional observational study aimed at estimating prevalence of type 2 diabetes in Sindhi families of Nagpur, 1262 of the 1462 individuals enrolled in the DISFIN study were included in the current analysis. From these 1262 individuals, investigators obtained data related to 10,461 time-aligned heartbeats recorded digitally.

This cohort had a mean age of 48 years and 61 percent were female. The overall prevalence of type 2 diabetes, prediabetes, and insulin resistance was 30 percent, 14 percent, and 35 percent respectively. Investigators also pointed out hypertension, obesity, and dyslipidemia were present among 51 percent, 40 percent, and 36 percent of participants.

For the purpose of analysis, the 10,461 time-aligned heartbeats were split into a training set, a validation set, and an independent test set, which included 8892, 523, and 1,046 beats, respectively. These recordings were processed with median filtering, band-pass filtering, and standard scaling. Investigators pointed out minority oversampling was performed to balance the training dataset prior to initiation of training and extreme gradient boosting was used to train the classifier that used the signal-processed ECG as input and predicted the membership to no diabetes, prediabetes, or type 2 diabetes classes.

Upon analysis, investigators found the DiaBeats algorithm predicted the classes with 97.1 percent precision, 96.2 percent recall, 96.8 percent accuracy, and 96.6 percent F1 score in the independent test set. Investigators noted a low calibration error (0.06) was observed with the calibrated model and the feature importance maps suggested that leads III, aVL, V4, V5, and V6 contributed most to the classification performance. Additionally, the predictions made by the algorithm matched the clinical expectations based on biological mechanisms of cardiac involvement in diabetes, according to investigators.

“Machine-learning-based DiaBeats algorithm using ECG signal data accurately predicted diabetes-related classes. This algorithm can help in early detection of diabetes and prediabetes after robust validation in external datasets,” investigators added.

Emerging research suggests the machine learning-based DiaBeats algorithm could facilitate early detection of prediabetes or diabetes.

Study Says Combination of AI and Electrocardiogram May Enhance Detection of Diabetes Risk

Is radiology becoming more representative of the increasingly diverse population of the country? While the authors of a new study cite improved female representation among academic faculty in radiology departments, they note that the numbers of female and Black applicants to radiology residency are low and relatively unchanged over the course of a decade.

, researchers compared data from two different time periods (2010-2011 and 2019-2020) to assess gender, race and ethnic trends among medical students, applicants to radiology residency programs, radiology residents and faculty within radiology departments at academic institutions.

The study authors noted increases in the percentage of women who were associate professors (improving from 23 percent in 2010 to 28 percent in 2019); professors (improving from 17 percent in 2010 to 22 percent in 2019); and department chairs (improving from 13 percent in 2010 to 17 percent in 2019).

However, the researchers only saw a marginal percentage increase from 2010 to 2019 for female applicants to radiology residency programs despite a five percent increase for women in medical school graduates during the same time period. Out of the study’s 8,756 female graduates of medical school in 2010, 388 (4.4 percent) applied to radiology residency programs, according to the study. In 2019, the study authors noted 11,160 female medical school graduates in the study database and 656 female applicants (5.8 percent) for radiology residency programs. Researchers also pointed out a decrease in female radiology residents from 2010 (1,260) to 2019 (1,166).

“Identifying and addressing the reasons for a low proportion of female radiology residency applicants, despite a highly diverse pool of medical students, would be key to increasing their representation in the field,” wrote study co-author Ajay Malhotra, MD, MMM, a professor at the Yale University School of Medicine, and colleagues.

In regard to the racial and ethnic composition of academic radiology faculty, the study authors noted that Asians saw proportional increases in the numbers of assistant professors (improving from 26 percent in 2010 to 29 percent in 2019); associate professors (improving from 19 percent in 2010 to 28 percent in 2019); and professors (improving from 13 percent in 2010 to 21 percent in 2019). There was also a seven percent increase in the numbers of Asian educators chairing radiology departments (improving from 9 percent in 2010 to 16 percent in 2019), according to the researchers.

However, Malhotra and colleagues found the proportion of Black educators at the assistant professor (3 percent), associate professor (2 percent) and professor levels (2 percent) remained the same between 2010 and 2019. Outside of a 1 percent increase in associate professor representation (from 3 percent to four percent), educators of Hispanic, Latino or Spanish origin saw no increases in the proportion of assistant professor (5 percent) and professor (3 percent) titles between 2010 and 2019.

The study authors also noted marginal increases between 2010 and 2019 for Black radiology residents and those of Hispanic, Latino, or Spanish descent. Black residents comprised 3 percent (139) of the 4,531 radiology residents in the study database in 2010 and 4 percent (166) of the 4,312 radiology residents in 2019. People of Hispanic, Latino or Spanish origin accounted for 5 percent (219) and 6 percent (254) of the radiology residents in 2010 and 2019 respectively.

Acknowledging the lack of Black and Hispanic representation at multiple levels in academic radiology, Malhotra and colleagues suggested that recruitment targeted during the medical school and residency application periods may help address these issues.

Study limitations included the faculty data being drawn from academic radiology practices, which may curtail extrapolation of the study findings to practices outside of academic settings, according to the study authors. They also acknowledged a lack of clarity as to whether race and sex were self-reported. In regard to residency program applicants and residents, Malhotra and colleagues noted the data did not distinguish between race and ethnicity, which may have led to inconsistencies in the reporting of statistics involving Hispanic and Latino populations.

In a new study, researchers examined trends with diversity in the radiology workforce, offering a closer look at the gender, race and ethnic makeup of radiology residency programs and academic faculty.

Diversity and Radiology: Where Things Stand

Expanded eligibility doesn’t necessarily translate to expanded access to lung cancer screening. In a recent study looking at the impact of 2021 recommendations from the United States Preventive Services Task Force (USPSTF) to lower screening thresholds from 55 to 50 years of age and cumulative smoking exposure from 30 to 20 pack years, researchers found that racial and socioeconomic disparities still play significant roles in limiting access to preventive lung cancer screening.

Comparing lung cancer screening eligibility estimates after 2013 recommendations from the USPSTF to eligibility estimates after the 2021 recommendations from USPSTF, the study authors noted an overall 81.4 percent increase in eligibility for lung cancer screening. Breaking the increases down by race and ethnicity, the researchers noted a 242 percent increase in eligibility for non-Hispanic Asian people; a 113 percent increase for the Hispanic population; a 107.5 increase for non-Hispanic Black study participants; and a 77.9 percent increase for non-Hispanic White study participants, according to the study.

significant socioeconomic disparities that didn’t appear to change after the lowered screening thresholds from the 2021 USPSTF recommendations

. In 2013, people with less than a high school diploma (20 percent) were four times less likely than those with a high school diploma or higher educational level (80 percent) to have lung cancer screening. After the 2021 recommendations, researchers estimated 19.1 of people of those with less than a high school diploma would be eligible for screening in comparison to 80.9 percent of those with a high school diploma or higher education.1

For people under the age of 65, the study authors also estimated that in 2013, people with no health care insurance were over 5.5 times less likely to have lung cancer screening in comparison to people with private insurance (10.9 percent versus 57.6 percent). Researchers noted similar percentages in the under 65 population after the 2021 recommendations with 57.1 percent of those with private insurance receiving lung cancer screening in comparison to 11 percent of those with no insurance.

“Despite the increases in the proportion of individuals who are now eligible for (lung cancer screening) based on the updated (2021 USPSTF recommendations), further work is needed to mitigate inequities in access,” wrote study co-author Robert J. Volk, Ph.D., a Hubert L. and Olive Stringer Distinguished Professor in Cancer Research at the University of Texas MD Anderson Cancer Center, and colleagues.

In a subgroup analysis, Volk and colleagues examined racial disparities in lung cancer screening by comparing the impact of socioeconomic factors among participants after the 2021 USPSTF recommendations. Comparing non-Hispanic White study participants to non-Hispanic Black study participants, the researchers found a higher percentage of private health insurance (59.4 percent versus 44.5 percent) for those under 65 years of age; a lower percentage of those without a high school diploma (17 percent vs. 28.7 percent); and a lower percentage of people with income below the poverty level (14.9 percent vs. 26.2 percent).

“Our findings show potential gaps, particularly among non-Hispanic Black individuals who have lower levels of education, income, and employment. These factors are important to consider within the context of exacerbating disparities that may compound upon one another,” noted Volk and colleagues.

The lack of transportation is another potential contributing factor to disparities in lung cancer screening, according to the study. The study authors said 11.1 percent of non-Hispanic Black study participants cited a lack of transportation as a reason for delaying medical care in comparison to 3.6 percent of non-Hispanic White study participants.

Citing other research, Volk and colleagues said longer driving distances in combination with socioeconomic disparities can be a deterrent to timely screening with computed tomography (CT) lung cancer screening.

“An analysis of census tract-level socioeconomic characteristics and distance to (American College of Radiology)-accredited computed tomography (CT) facilities showed increased driving distances to CT facilities for census tracts with higher proportions of individuals who are uninsured, have Medicaid, and have less than a high school degree,” pointed out Volk and colleagues.

The study authors conceded a limitation of the study was using data from the 2015 National Health Interview Survey in order to assess a variety of socioeconomic factors that may impact lung cancer screening. Clinical considerations that could affect screening eligibility were not included in the study, according to Volk and colleagues.

Maki KG, Talluri R, Toumazis I, Shete S, Volk RJ. Impact of U.S. Preventive Services Task Force lung cancer screening update on drivers of disparities in screening eligibility. 

 2022 Jul 24. doi: 10.1002/cam4.5066. Online ahead of print.

2. Tailor D, Tong BC, Gao J, Choudhury KR, Rubin GD. A geospatial analysis of factors affecting access to CT facilities: implications for lung cancer screening. 

While recent recommendations from the United States Preventive Services Task Force (USPSTF) to lower lung cancer screening thresholds significantly expanded eligibility for screening tests such as low-dose computed tomography (CT), differences in education, health-care insurance and proximity to health-care facilities continue to be key drivers of racial and socioeconomic disparities limiting access to appropriate preventive care. 

Lung Cancer Screening Study Says Significant Disparities Persist Despite Expanded Eligibility

Case submitted by Saeed Rad, M.D. via the 

Image IQ Quiz: Patient with History of Excessive Drinking Vomits Blood

An emerging artificial intelligence (AI) software program that may enhance the efficiency and accuracy of prostate cancer diagnosis has received 

510(k) clearance from the Food and Drug Administration (FDA)

Through computer-aided detection (CAD), ProstatID™ (Bot Image) offers a colorized translucent overlay of two-dimensional axial T2 magnetic resonance imaging (MRI) to identify soft tissue lesions that may warrant closer inspection for potential malignancy. The ProstatID algorithm also utilizes the Prostate Imaging Reporting and Data System (PI-RADS) to score the cancer probability of suspicious lesions, according to Bot Image. The company noted that two internal studies showed improved accuracy and a reduced false positive rate with the use of ProstatID.

While other technologies have enhanced formatting and segmentation of MRI images of the prostate, Bot Image said ProstatID is a significant advance for the diagnosis of prostate cancer.

“Prostate cancer screening and detection methods adoption (have) changed little over the past 30 years despite the mountain of evidence pointing to the efficacy of superior technologies and the futility of the old methods,” noted Bot Image founder and CEO Randall W. Jones, Ph.D. “Sadly, this has resulted in the unnecessary and premature deaths of countless numbers of men in the (United States) alone. ProstatID represents an exciting step in the fight to save lives.”

Bot Image noted that ProstatID is available to radiologists as a software as a service (SaaS) tool, which provides secure, HIPAA compliant connectivity between a cloud-based ProstatID server and a facility’s radiology department server.

The ProstatID, an adjunctive artificial intelligence software that radiologists can utilize with traditional magnetic resonance imaging (MRI), reportedly measures prostate gland volume, and suggests PI-RADS scoring of suspicious lesions.

FDA Clears New AI Software for MRI Detection of Prostate Cancer

Eric Postal, MD, is a diagnostic radiologist with the University of Pittsburgh Medical Center. 

I just had some extended family over for a pool party/cookout. Once upon a pre-pandemic time, I threw larger annual, non-familial summertime blowouts but you can imagine what the past few years have done to such plans.

This year, however, I decided that there had been enough familial isolation, and cast a wide net of invites to various cousins. Everybody involved had taken COVID-19 seriously, gotten their shots, etc. so it wasn’t exactly a risky affair. Anybody who felt otherwise could simply choose to not attend. It was a great time, and I do enjoy hosting.

There were some capable negotiators in that crowd. One does it for a living. His background was in physics, but his livelihood has been in the business world for as long as I can remember. He prides himself on getting the best possible deal and leaving no stone unturned. Another couple of them are seasoned car traders and always seem to wind up with deals almost too good to be true.

I’m not such an individual. Just the thought of hard bargaining exhausts me to the point of outweighing any potential gains I might be leaving on the table. Even when I have been in marketplaces where you’re supposed to haggle over everything, I find myself unmotivated to buy rather than dive into that process. Give me your best faith offer. I’ll accept or reject it and let that be that.

Still, I know that there’s a certain societal expectation of a bargaining dance, and I’d be a fool to imagine myself somehow immune. I go ahead with it when I must do so. At least it can be said that I have improved over the years. Every now and then, hindsight makes it clear that I could have done this or that better, and I can wind up feeling embarrassingly sucker-like.

That embarrassment can be tinged with anger at myself. It’s one thing to look back and wish that you’d known then what you know now. However, sometimes you can tell that you probably had the wisdom to avoid whatever pitfall you fell into.

There’s more than just a tinge of anger when I know that my suboptimal deal wasn’t 100 percent my own fault. That is, I didn’t completely make a sucker out of myself. Someone else made a sucker out of me. Maybe there is blame to be shared. Perhaps the other party saw that I was overlooking something important and let me proceed for instance. Perhaps the other guy was more of a villain, giving me incomplete information or outright lying.

Sometimes the deal I made was fine but subsequently, the other party deviates from its terms (or just the spirit of the agreement), essentially with the attitude of “What are you going to do about it? Take me to court?” In these cases, I can still wind up feeling like a sucker, a particularly angry one, because I should have built more defenses into the deal for myself, or simply because I misread the other party’s character as being better than it was.

In the final analysis, though, it matters relatively little whether I am 100 percent to blame for being a sucker, the other guy was totally guilty, or whether we were both to blame.

Feeling like a sucker makes me want to take action. Perhaps it is just to convince myself that I have learned enough not to make the same mistake again. Perhaps it is to get myself out of the situation rather than staying in it and remaining a sucker. Possibly, there is an opportunity to get some payback if I think somebody else is to blame.

Not only does it matter relatively little who is to blame for being a sucker, I also don’t much care which side of the deal he or she is on. I don’t want any suckers at my table. If I could magically transform myself into a consummate negotiator like those relatives I mentioned, and always get the best deal even if it meant making the other person a sucker, I wouldn’t want to.

Maybe I would feel otherwise if I were in a different line of work, in which I could have transactions with people and never see them again. But the diagnostic radiology field is a small world, and even the health-care sphere is a limited space. Heck, with social media being ubiquitous as it is, anyone could cross your path again someday and just might make a point of it if he or she remembers you making a sucker out of them.

If there is even a shred of a chance someone I dealt with is going to be in my life again at some point, why on Earth should either of us prioritize out dealing the other to the point that bad blood may be engendered or try to bend the rules of an agreement to one’s own benefit after the fact?

To my way of thinking, the potential gain would have to be monumental to sacrifice that much goodwill. It would be better to let the other person come away with something, even if you don’t have to, so he or she can feel like there was a “win,” and be happy at the thought of dealing with you again in the future.

I think back to a couple of former employers who chose to mess around with the deals I had with them while still trying to maintain a good working relationship. I wonder if things might have gone differently if I had the presence of mind and gumption to say, “Maybe my expectations were too high when you hired me, but the way things are panning out has me feeling like a sucker. Do you want me to feel like a sucker for being on your team? How can we fix this?”

Consistently leveling with colleagues, patients and others can promote a relatively straightforward path to optimal goodwill.

Emphasizing Negotiation-Free Candor in and Outside of Your Radiology Circles

's Weekly Scan is your chance to catch up on the latest radiology content, from new research and device news to the latest case quizzes and conference reports. Review the slideshow below to see this week's top content:

Diagnostic Imaging's Weekly Scan: July 31-August 6

Slideshows

Catch up on the top radiology content of the past week.

C) Progressive multifocal leukoencephalopathy 

Day 1 images

Day 2 images

Image IQ Quiz: 36-Year-Old Patient with Six-Month History of 'Brain Fog'

Recognizing the potential benefits and daunting challenges of incorporating artificial intelligence (AI) models into radiology workflows, researchers have published a guidance that examines four key questions on the process for evaluating and implementing AI-enabled modalities into practice.

, the authors discuss the makeup of an AI governance and management structure, pertinent factors for assessing AI-powered imaging modalities, parameters for implementation into practice as well as monitoring of AI tools for effectiveness.

Here are seven key takeaways from the guidance.

1. Radiology AI algorithms currently comprise the biggest amount of AI models that have been cleared by the Food and Drug Administration (FDA), according to the article authors. Not only are radiologists represented in organizational AI governance structures, the study authors also pointed out that radiology groups are leading the implementation of AI imaging at many health-care institutions.

2. In describing the makeup of an AI governing body, the authors said those defining priorities, strategies, and the scope of evaluating and implementing AI models at larger institutions may include leadership from multiple imaging departments, electronic health record (EHR) managers, information technology (IT) management, a legal representative, AI experts as well as institutional administration liaisons. However, the involvement of end users of the AI software is particularly critical, according to the article authors.

“Incorporating end users into the governance structure is of utmost importance to consider their needs and concerns about an (AI) algorithm and to include (them) into the decision-making process,” wrote study co-author Curtis Langlotz, M.D., Ph.D., the director of the Center for Artificial Intelligence in Medicine and Imaging at Stanford University, and colleagues.

(Editor’s note: For related video content, see “

Assessing and Implementing Artificial Intelligence in Radiology

Essential Questions for Assessing Artificial Intelligence Vendors in Radiology

3. Determining who pays for the installation and maintenance of AI models is an important consideration for radiologists in community hospital settings. While the AI models would be monitored by the radiology department, the authors maintained that radiologists would need to emphasize how the given AI model would benefit the overall health system as well as radiologists in order to prevent the costs from being fully absorbed by the radiology department.

“If a health system bears the financial burden for AI, radiologists must develop the value proposition for each model. If models are seen as only improving radiologist efficiency or accuracy, then the radiology group may be asked to bear some or all of the financial cost” noted Langlotz and colleagues.

4. Does the data set used in the development of the AI model align with clinical use in your practice’s patient population? The article authors emphasize assessment of inclusion and exclusion criteria in the model development for potential bias. Another important consideration is whether there has been external validation of the AI model. In order to gauge the accuracy and reproducibility of the AI model, Langlotz and colleagues suggested that institutions perform a test of the model prior to incorporating it into daily use.

5. For risk assessment of AI tools, the study authors recommend a risk categorization framework issued by the International Medical Device Regulators Forum (IMDRF), which considers the nature of intended use cases as well as the potential impact of the software upon clinical decision-making. For lower-risk AI algorithms, such as those that aid in worklist prioritization, the IMDRF system can also be beneficial, according to Langlotz and colleagues.

6. Ensuring appropriate data security is a key aspect of implementing AI tools into practice. Noting the use of data maps to convey data location, flow and encryption status for projects, the article authors emphasized appropriate data security and sharing practices when vetting AI tools. Langlotz and colleagues added that those assessing the AI tools should be especially vigilant for models that are cloud-based or those developed internally at an institution for clinical use.

“This detailed analysis and accountability for data security and robust governance of data use are vital steps toward ensuring patient protection and public trust,” explained Langlotz and colleagues.

7. Acknowledging that post-AI implementation monitoring can be challenging for smaller practices and those without dedicated informatics support, the article authors suggest the use of automatically populated registries to ease monitoring burden and looking for AI vendors that can provide monitoring data capture in their products.

Researchers discuss key parameters for the assessment, implementation and post-implementation monitoring of emerging artificial intelligence (AI) tools in radiology practices large and small.