A novel 3D MRI-based approach for assessing supraspinatus muscle length

Rotator cuff (RC) tears are a common source of pain and decreased shoulder strength. Muscle length is known to affect muscle strength, and therefore evaluating changes in supraspinatus muscle length associated with RC pathology, surgical repair, and post-operative recovery may provide insights into functional deficits. Our objective was to develop a reliable MRI-based approach for assessing supraspinatus muscle length. Using a new semi-automated approach for identifying 3D location of the muscle-tendon junction (MTJ), supraspinatus muscle length was calculated as the sum of MTJ distance (distance between 3D MTJ position and glenoid plane) and supraspinatus fossa length (distance between root of the scapular spine and glenoid plane). Inter- and intra-operator reliability of this technique were assessed with intraclass correlation coefficient (ICC) and found to be excellent (ICCs > 0.96). Muscle lengths of 6 patients were determined before RC repair surgery and at 3- and 12-months post-surgery. Changes in normalized muscle length (muscle length as a percentage of pre-surgical muscle length) at 3 months post-surgery varied considerably across patients (16.1 % increase to 7.0 % decrease) but decreased in all patients from 3- to 12-months post-surgery (0.3 % to 17.2 %). This study developed a novel and reliable approach for quantifying supraspinatus muscle length and provided preliminary demonstration of its utility by assessing muscle length changes associated with RC pathology and surgical repair. Future studies can use this technique to evaluate changes over time in supraspinatus muscle length in response to clinical intervention, and associations between muscle length and shoulder function.

Radiology Resident Diagnostic In-Training Exam Scores: Impact of Subspecialty Imaging Volume and Rotation Scheduling

PURPOSE

To determine the relationship between resident imaging volumes and number of subspecialty rotations with Diagnostic Radiology In-Training (DXIT) subspecialty scores.

METHODS

DXIT-scaled subspecialty scores from a single large diagnostic radiology training program from 2014 to 2020 were obtained. The cumulative number of imaging studies dictated by each resident and specific rotations were mapped to each subspecialty for each year of training. DXIT subspecialty scores were compared against the total subspecialty imaging volume and the total number of rotations in a subspecialty for each resident year. A total of 52 radiology residents were trained during the study period and included in the dataset.

RESULTS

There was a positive linear relationship between the number of neuro studies and scaled neuro DXIT scores for R1s (Pearson coefficient: 0.29; p-value: 0.034) and between the number of breast studies and the number of neuro studies with DXIT scores for R2s (Pearson coefficients: 0.50 and 0.45, respectively; p-values: 0.001 and 0.003, respectively). Furthermore, a positive significant linear relationship between the total number of rotations in cardiac, breast, neuro, and thoracic subspecialties and their scaled DXIT scores for R2 residents (Pearson coefficients: 0.34, 0.49, 0.33, and 0.32, respectively; p-value: 0.025, 0.001, 0.03, and 0.036, respectively) and between the total number of nuclear medicine rotations with DXIT scores for R3s (Pearson coefficient: 0.41; p-value: 0.016).

CONCLUSION

Resident subspecialty imaging volumes and rotations have a variable impact on DXIT scores. Understanding the impact of study volume and the number of subspecialty rotations on resident medical knowledge will help residents and program directors determine how much emphasis to place on these factors during residency.

Comparison of patient characteristics and treatment approaches for femoral and inguinal hernias utilizing dynamic ultrasound at a single institution

PURPOSE

To assess the differences in management approach to femoral versus inguinal hernias and to identify patient characteristics associated with each hernia type.

METHODS

Imaging studies for patients who had undergone dynamic ultrasound evaluation for the symptom of groin pain between January 1, 2010, and March 31, 2019, at a single institution Musculoskeletal Department were analyzed. Positive femoral hernia imaging studies were compared to studies for inguinal hernias and matching medical records for imaging studies were analyzed. Association of patient characteristics (age, sex, smoking, diabetes) with hernia type was assessed. Primary outcomes were surgical versus non-surgical approach, type of surgery, number of follow-up visits, and pain resolution.

RESULTS

A total of 1319 patients presented with groin pain and were assessed with dynamic ultrasound (534 female; 785 male; mean [± SD] age 48.2 ± 16.5). While 409 (31.0%) patients had a femoral hernia detected, 666 (50.6%) had an inguinal hernia detected (p < .05). Significantly more inguinal hernias were surgically repaired than femoral hernias (65.0% vs 53.9% p = .008), and more inguinal hernias than femoral hernias were treated with open surgery (71.0% vs 57.7%; p = .014). Patients with femoral hernias had significantly more follow-up clinic visits than patients with inguinal hernias (mean [± SD] 2.65 ± 4.80 vs 1.76 ± 1.27; p = .010). No difference in the percentage of patients who had pain resolution was observed (82.2% inguinal vs 75.0% femoral; p = .13).

CONCLUSIONS

Femoral hernias were managed more conservatively than inguinal hernias at our institution.

Automation of Protocoling Advanced MSK Examinations Using Natural Language Processing Techniques

Imaging examination selection and protocoling are vital parts of the radiology workflow, ensuring that the most suitable exam is done for the clinical question while minimizing the patient's radiation exposure. In this study, we aimed to develop an automated model for the revision of radiology examination requests using natural language processing techniques to improve the efficiency of pre-imaging radiology workflow. We extracted Musculoskeletal (MSK) magnetic resonance imaging (MRI) exam order from the radiology information system at Henry Ford Hospital in Detroit, Michigan. The pretrained transformer, "DistilBERT" was adjusted to create a vector representation of the free text within the orders while maintaining the meaning of the words. Then, a logistic regression-based classifier was trained to identify orders that required additional review. The model achieved 83% accuracy and had an area under the curve of 0.87.

Deep COVID DeteCT: an international experience on COVID-19 lung detection and prognosis using chest CT

The Coronavirus disease 2019 (COVID-19) presents open questions in how we clinically diagnose and assess disease course. Recently, chest computed tomography (CT) has shown utility for COVID-19 diagnosis. In this study, we developed Deep COVID DeteCT (DCD), a deep learning convolutional neural network (CNN) that uses the entire chest CT volume to automatically predict COVID-19 (COVID+) from non-COVID-19 (COVID-) pneumonia and normal controls. We discuss training strategies and differences in performance across 13 international institutions and 8 countries. The inclusion of non-China sites in training significantly improved classification performance with area under the curve (AUCs) and accuracies above 0.8 on most test sites. Furthermore, using available follow-up scans, we investigate methods to track patient disease course and predict prognosis.

COVID-19 and Radiology Resident Imaging Volumes-Differential Impact by Resident Training Year and Imaging Modality

Rationale and Objectives

The COVID-19 pandemic has greatly impacted radiology departments across the country. The pandemic has also disrupted resident education, both due to departmental social distancing efforts and reduced imaging volumes. The purpose of this study was to assess the differential impact the pandemic had on radiology resident imaging volumes by training year and imaging modality.

Materials and Methods

All signed radiology resident reports were curated during defined prepandemic and intrapandemic time periods. Imaging case volumes were analyzed on a mean per resident basis to quantify absolute and percent change by training level. Change in total volume by imaging modality was also assessed. The number of resident workdays assigned outside the normal reading room was also calculated.

Results

Overall percent decline in resident imaging interpretation volume from the prepandemic to intrapandemic time period was 62.8%. R1s and R2s had the greatest decline at 87.3% and 64.3%, respectively. Mammography, MRI and nuclear medicine had the greatest decline in resident interpretation volume at 92.0%, 73.2%, and 73.0%, respectively. During the intrapandemic time period, a total of 478 resident days (mean of 14.5 days per resident) were reassigned outside of the radiology reading room.

Conclusion

The COVID-19 pandemic caused a marked decrease in radiology resident imaging interpretation volume and has had a tremendous impact on resident education. The decrease in case interpretation, as well as in-person teaching has profound implications for resident education. Knowledge of this differential decrease by training level will help residency programs plan for the future.

Impact of the COVID-19 pandemic on radiology physician work RVUs at a large subspecialized radiology practice

PURPOSE

As the COVID-19 pandemic continues, efforts by radiology departments to protect patients and healthcare workers and mitigate disease spread have reduced imaging volumes. This study aims to quantify the pandemic's impact on physician productivity across radiology practice areas as measured by physician work Relative Value Units (wRVUs).

MATERIALS AND METHODS

All signed diagnostic and procedural radiology reports were curated from January 1st to July 1st of 2019 and 2020. Physician work RVUs were assigned to each study type based on the Medicare Physician Fee Schedule. Utilizing divisional assignments, radiologist schedules were mapped to each report to generate a sum of wRVUs credited to that division for each week. Differential impact on divisions were calculated relative to a matched timeframe in 2019 and a same length pre-pandemic time period in 2020.

RESULTS

All practice areas saw a substantial decrease in wRVUs from the 2020 pre- to intra-pandemic time period with a mean decrease of 51.5% (range 15.4%-76.9%). The largest declines were in Breast imaging, Musculoskeletal, and Neuroradiology, which had decreases of 76.9%, 75.3%, and 67.5%, respectively. The modalities with the greatest percentage decrease were mammography, MRI, and non-PET nuclear medicine.

CONCLUSION

All radiology practice areas and modalities experienced a substantial decrease in wRVUs. The greatest decline was in Breast imaging, Neuroradiology, and Musculoskeletal radiology. Understanding the differential impact of the pandemic on practice areas will help radiology departments prepare for the potential depth and duration of the pandemic by better understanding staffing needs and the financial effects.

Acute Pulmonary Embolism and COVID-19

Risk factors for pulmonary embolism in patients with coronavirus disease 2019 include obesity, an elevated d -dimer value, elevated C-reactive protein level, and a rising d -dimer value over time.

Interobserver and Intraobserver Variability in the CT Assessment of COVID-19 Based on RSNA Consensus Classification Categories

PURPOSE

To assess the interobserver and intraobserver agreement of fellowship trained chest radiologists, nonchest fellowship-trained radiologists, and fifth-year radiology residents for COVID-19-related imaging findings based on the consensus statement released by the Radiological Society of North America (RSNA).

METHODS

A survey of 70 chest CTs of polymerase chain reaction (PCR)-confirmed COVID-19 positive and COVID-19 negative patients was distributed to three groups of participating radiologists: five fellowship-trained chest radiologists, five nonchest fellowship-trained radiologists, and five fifth-year radiology residents. The survey asked participants to broadly classify the findings of each chest CT into one of the four RSNA COVID-19 imaging categories, then select which imaging features led to their categorization. A 1-week washout period followed by a second survey comprised of randomly selected exams from the initial survey was given to the participating radiologists.

RESULTS

There was moderate overall interobserver agreement in each group (κ coefficient range 0.45-0.52 ± 0.02). There was substantial overall intraobserver agreement across the chest and nonchest groups (κ coefficient range 0.61-0.67 ± 0.06) and moderate overall intraobserver agreement within the resident group (κ coefficient 0.58 ± 0.06). For the image features that led to categorization, there were varied levels of agreement in the interobserver and intraobserver components that ranged from fair to perfect kappa values. When assessing agreement with PCR-confirmed COVID status as the key, we observed moderate overall agreement within each group.

CONCLUSION

Our results support the reliability of the RSNA consensus classification system for COVID-19-related image findings.

Chest Radiographic Appearance of Minimally Invasive Cardiac Implants and Support Devices: What the Radiologist Needs to Know

Minimally invasive implantable cardiac devices used in valve repair and replacement, cardiovascular support, and partial chamber and appendageal occlusion represent a burgeoning area of both bioengineering and clinical innovation. In addition to familiarizing the reader with the radiographic appearance of the most commonly utilized and encountered newer devices, this review will also address the relevant clinical and pathophysiological indications for usage and deployment as well as potentially encountered complications.

Order Entry Protocols Are an Amenable Target for Workflow Automation

PURPOSE

Order entry protocol selection of advanced imaging studies is labor-intensive, can disrupt workflow, and may displace staff from more valuable tasks. The aim of this study was to explore and compare the behaviors of radiologic technologists and radiologists when determining protocol to identify opportunities for workflow automation.

METHODS

A data set of over 273,000 cross-sectional examination orders from four hospitals within our health system was created. From this data set, we isolated the 12 most frequently requested examinations, which represent almost 50% of the entirety of advanced imaging volume. Intergroup comparisons were made between behavior of radiologic technologists and radiologists or residents when determining protocol. Frequencies of changes were calculated. Common parameters of changed examinations were identified.

RESULTS

The overall change rate for both radiologists and residents (4%) is very low and comparable to the overall change rate of radiologic technologists (1%). The change rates for the 12 most ordered examinations were calculated and compared individually. Most examinations that underwent change involved a patient with a low estimated glomerular filtration rate, a patient with a contrast allergy, or a provider ordering a general examination but in fact wanting an organ-specific protocol or an angiographic study.

CONCLUSION

Order entry protocol selection of the most frequently ordered advanced imaging examinations was rarely a value-added activity because these examinations are rarely changed. Changes follow predictable patterns that make order entry protocol selection of most radiology orders for advanced imaging amenable to workflow automation.

Evaluating the Effect of Unstructured Clinical Information on Clinical Decision Support Appropriateness Ratings

OBJECTIVE

To determine the appropriateness rating (AR) of advanced inpatient imaging requests that were not rated by prospective, point-of-care clinical decision support (CDS) using computerized provider order entry.

MATERIALS AND METHODS

During 30-day baseline and intervention periods, CDS generated an AR for advanced inpatient imaging requests (nuclear medicine, CT, and MRI) using provider-selected structured indications from pull-down menus in the computerized provider order entry portal. The AR was only displayed during the intervention, and providers were required to acknowledge the AR to finalize the request. Subsequently, the unstructured free text information accompanying all requests was reviewed, and the AR was revised when possible. The percentage of unrated requests and the overall AR, before and after radiologist review, were compared between periods and by provider type.

RESULTS

CDS software prospectively generated an AR for only 25.4% and 28.4% of baseline and intervention imaging requests, respectively; however, radiologist review generated an AR for 82.4% and 93.6% of the same requests. During the respective periods, the percentage of baseline and intervention imaging requests considered appropriate was 18.7% and 22.9% by prospective CDS software rating and increased to 82.4% and 88.7% with radiologist review.

CONCLUSION

Despite limited effective use of CDS software, the percentage of requests containing additional, relevant clinical information increased, and the majority of requests had overall high appropriateness when reviewed by a radiologist. Additional work is needed to improve the amount and quality of clinical information available to CDS software and to facilitate the entry of this information by appropriate end users.