Evaluating the Use of ChatGPT to Accurately Simplify Patient-centered Information about Breast Cancer Prevention and Screening.

Purpose To evaluate the use of ChatGPT as a tool to simplify answers to common questions about breast cancer prevention and screening. Materials and Methods In this retrospective, exploratory study, ChatGPT was requested to simplify responses to 25 questions about breast cancer to a sixth-grade reading level in March and August 2023. Simplified responses were evaluated for clinical appropriateness. All original and simplified responses were assessed for reading ease on the Flesch Reading Ease Index and for readability on five scales: Flesch-Kincaid Grade Level, Gunning Fog Index, Coleman-Liau Index, Automated Readability Index, and the Simple Measure of Gobbledygook (ie, SMOG) Index. Mean reading ease, readability, and word count were compared between original and simplified responses using paired t tests. McNemar test was used to compare the proportion of responses with adequate reading ease (score of 60 or greater) and readability (sixth-grade level). Results ChatGPT improved mean reading ease (original responses, 46 vs simplified responses, 70; P < .001) and readability (original, grade 13 vs simplified, grade 8.9; P < .001) and decreased word count (original, 193 vs simplified, 173; P < .001). Ninety-two percent (23 of 25) of simplified responses were considered clinically appropriate. All 25 (100%) simplified responses met criteria for adequate reading ease, compared with only two of 25 original responses (P < .001). Two of the 25 simplified responses (8%) met criteria for adequate readability. Conclusion ChatGPT simplified answers to common breast cancer screening and prevention questions by improving the readability by four grade levels, though the potential to produce incorrect information necessitates physician oversight when using this tool. Keywords: Mammography, Screening, Informatics, Breast, Education, Health Policy and Practice, Oncology, Technology Assessment Supplemental material is available for this article. © RSNA, 2023.

Assessing AI-Powered Patient Education: A Case Study in Radiology.

RATIONALE AND OBJECTIVES: With recent advancements in the power and accessibility of artificial intelligence (AI) Large Language Models (LLMs) patients might increasingly turn to these platforms to answer questions regarding radiologic examinations and procedures, despite valid concerns about the accuracy of information provided. This study aimed to assess the accuracy and completeness of information provided by the Bing Chatbot-a LLM powered by ChatGPT-on patient education for common radiologic exams. MATERIALS AND METHODS: We selected three common radiologic examinations and procedures: computed tomography (CT) abdomen, magnetic resonance imaging (MRI) spine, and bone biopsy. For each, ten questions were tested on the chatbot in two trials using three different chatbot settings. Two reviewers independently assessed the chatbot's responses for accuracy and completeness compared to an accepted online resource, radiologyinfo.org. RESULTS: Of the 360 reviews performed, 336 (93%) were rated "entirely correct" and 24 (7%) were "mostly correct," indicating a high level of reliability. Completeness ratings showed that 65% were "complete" and 35% were "mostly complete." The "More Creative" chatbot setting produced a higher proportion of responses rated "entirely correct" but there were otherwise no significant difference in ratings based on chatbot settings or exam types. The readability level was rated eighth-grade level. CONCLUSION: The Bing Chatbot provided accurate responses answering all or most aspects of the question asked of it, with responses tending to err on the side of caution for nuanced questions. Importantly, no responses were inaccurate or had potential to cause harm or confusion for the user. Thus, LLM chatbots demonstrate potential to enhance patient education in radiology and could be integrated into patient portals for various purposes, including exam preparation and results interpretation.

Use of ChatGPT, GPT-4, and Bard to Improve Readability of ChatGPT's Answers to Common Questions About Lung Cancer and Lung Cancer Screening.

ChatGPT's responses to questions about lung cancer and LCS, although deemed clinically appropriate by cardiothoracic radiologists, were difficult to read. Simplified responses from three LLMs (ChatGPT, GPT-4, and Bard) had improved reading ease and readability (in terms of U.S. grade levels). However, some simplified responses were no longer clinically appropriate.

Radiomics and Deep Learning for Disease Detection in Musculoskeletal Radiology: An Overview of Novel MRI- and CT-Based Approaches.

ABSTRACT: Radiomics and machine learning-based methods offer exciting opportunities for improving diagnostic performance and efficiency in musculoskeletal radiology for various tasks, including acute injuries, chronic conditions, spinal abnormalities, and neoplasms. While early radiomics-based methods were often limited to a smaller number of higher-order image feature extractions, applying machine learning-based analytic models, multifactorial correlations, and classifiers now permits big data processing and testing thousands of features to identify relevant markers. A growing number of novel deep learning-based methods describe magnetic resonance imaging- and computed tomography-based algorithms for diagnosing anterior cruciate ligament tears, meniscus tears, articular cartilage defects, rotator cuff tears, fractures, metastatic skeletal disease, and soft tissue tumors. Initial radiomics and deep learning techniques have focused on binary detection tasks, such as determining the presence or absence of a single abnormality and differentiation of benign versus malignant. Newer-generation algorithms aim to include practically relevant multiclass characterization of detected abnormalities, such as typing and malignancy grading of neoplasms. So-called delta-radiomics assess tumor features before and after treatment, with temporal changes of radiomics features serving as surrogate markers for tumor responses to treatment. New approaches also predict treatment success rates, surgical resection completeness, and recurrence risk. Practice-relevant goals for the next generation of algorithms include diagnostic whole-organ and advanced classification capabilities. Important research objectives to fill current knowledge gaps include well-designed research studies to understand how diagnostic performances and suggested efficiency gains of isolated research settings translate into routine daily clinical practice. This article summarizes current radiomics- and machine learning-based magnetic resonance imaging and computed tomography approaches for musculoskeletal disease detection and offers a perspective on future goals and objectives.

The Radiology: Artificial Intelligence Trainee Editorial Board: Initial Experience and Future Directions.

In 2019, the journal Radiology: Artificial Intelligence introduced its Trainee Editorial Board (TEB) to offer formal training in medical journalism to medical students, radiology residents and fellows, and research-career trainees. The TEB aims to build a community of radiologists, radiation oncologists, medical physicists, and researchers in fields related to artificial intelligence (AI) in radiology. The program presented opportunities to learn about the editorial process, improve skills in writing and reviewing, advance the field of AI in radiology, and help translate and disseminate AI research. To meet these goals, TEB members contribute actively to the editorial process from peer review to publication, participate in educational webinars, and create and curate content in a variety of forms. Almost all of the contact has been mediated through the web. In this article, we share initial experiences and identify future directions and opportunities.

Artificial intelligence in orthopedic implant model classification: a systematic review.

Although artificial intelligence models have demonstrated high accuracy in identifying specific orthopedic implant models from imaging, which is an important and time-consuming task, the scope of prior works and performance of prior models have not been evaluated. We performed a systematic review to summarize the scope, methodology, and performance of artificial intelligence algorithms in classifying orthopedic implant models. We performed a literature search in PubMed, EMBASE, and the Cochrane Library for studies published up to March 10, 2021, using search terms related to "artificial intelligence", "orthopedic", "implant", and "arthroplasty". Studies were assessed using a modified version of the methodologic index for non-randomized studies. Reported outcomes included area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. The search identified 2689 records, of which 11 were included in the final review. The number of implant models evaluated ranged from 2 to 27. Five studies reported overall AUC across all included models which ranged from 0.94 to 1.0. Overall accuracy values ranged from 0.804 to 1.0. One study compared AI model performance with that of three surgeons, reporting similar performance. There was a large degree of variation in methodology and reporting quality. Artificial intelligence algorithms have demonstrated strong performance in classifying orthopedic implant models from radiographs. Further research is needed to compare artificial intelligence alone and as an adjunct with human experts in implant identification. Future studies should aim to adhere to rigorous artificial intelligence development methods and thorough, transparent reporting of methods and results.

DeepCAT: Deep Computer-Aided Triage of Screening Mammography.

Although much deep learning research has focused on mammographic detection of breast cancer, relatively little attention has been paid to mammography triage for radiologist review. The purpose of this study was to develop and test DeepCAT, a deep learning system for mammography triage based on suspicion of cancer. Specifically, we evaluate DeepCAT's ability to provide two augmentations to radiologists: (1) discarding images unlikely to have cancer from radiologist review and (2) prioritization of images likely to contain cancer. We used 1878 2D-mammographic images (CC & MLO) from the Digital Database for Screening Mammography to develop DeepCAT, a deep learning triage system composed of 2 components: (1) mammogram classifier cascade and (2) mass detector, which are combined to generate an overall priority score. This priority score is used to order images for radiologist review. Of 595 testing images, DeepCAT recommended low priority for 315 images (53%), of which none contained a malignant mass. In evaluation of prioritizing images according to likelihood of containing cancer, DeepCAT's study ordering required an average of 26 adjacent swaps to obtain perfect review order. Our results suggest that DeepCAT could substantially increase efficiency for breast imagers and effectively triage review of mammograms with malignant masses.

Saline lavage after a "dry tap".

AIMS: In patients with a "dry" aspiration during the investigation of prosthetic joint infection (PJI), saline lavage is commonly used to obtain a sample for analysis. The aim of this study was to investigate prospectively the impact of saline lavage on synovial fluid analysis in revision arthroplasty. METHODS: Patients undergoing revision hip (THA) or knee arthroplasty (TKA) for any septic or aseptic indication were enrolled. Intraoperatively, prior to arthrotomy, the maximum amount of fluid possible was aspirated to simulate a dry tap (pre-lavage) followed by the injection with 20 ml of normal saline and re-aspiration (post-lavage). Pre- and post-lavage synovial white blood cell (WBC) count, percent polymorphonuclear cells (%PMN), and cultures were compared. RESULTS: A total of 78 patients had data available for analysis; 17 underwent revision THA and 61 underwent revision TKA. A total of 16 patients met modified Musculoskeletal Infection Society (MSIS) criteria for PJI. Pre- and post-lavage %PMNs were similar in septic patients (87% vs 85%) and aseptic patients (35% vs 39%). Pre- and post-lavage synovial fluid WBC count were far more disparate in septic (53,553 vs 8,275 WBCs) and aseptic (1,103 vs 268 WBCs) cohorts. At a cutoff of 80% PMN, the post-lavage aspirate had a sensitivity of 75% and specificity of 95%. At a cutoff of 3,000 WBCs, the post-lavage aspirate had a sensitivity of 63% and specificity of 98%. As the post-lavage synovial WBC count increased, the difference between pre- and post-lavage %PMN decreased (mean difference of 5% PMN in WBC < 3,000 vs mean difference 2% PMN in WBC > 3,000, p = 0.013). Of ten positive pre-lavage fluid cultures, only six remained positive post-lavage. CONCLUSION: While saline lavage aspiration significantly lowered the synovial WBC count, the %PMN remained similar, particularly at WBC counts of > 3,000. These findings suggest that in patients with a dry-tap, the %PMN of a saline lavage aspiration has reasonable sensitivity (75%) for the detection of PJI. Cite this article: Bone Joint J 2020;102-B(6 Supple A):138-144.

Hypoalbuminemia is a risk factor for predicting early postoperative complications after proximal humerus fracture fixation.

BACKGROUND: Hypoalbuminemia is a marker for malnourishment and is associated with poor outcomes in the setting of hip fractures, periprosthetic joint fractures, and spine surgery. We hypothesized that hypoalbuminemia is associated with higher rates of early complication in patients undergoing surgical treatment for proximal humerus fractures. METHODS: Utilizing the ACS NSQIP database, all proximal humerus fractures treated with plate fixation were extracted from 2006 to 2015. Two cohorts were compared based on hypoalbuminemia (albumin <3.5 g/dL) versus normoalbuminemia (albumin >3.5 g/dL). Patient demographics and postoperative complications were analyzed with multivariable regression. RESULTS: Out of 916 patients undergoing PHF surgery, 290(31.7%) satisfied criteria for hypoalbuminemia (mean age: 65.9, female: 71%). Among 339 obese patients with BMI>30, 87 (25.7%) were hypoalbuminemic. Patients with hypoalbuminemia were overall at higher risk of any (29.7% vs 12.1%, p < 0.001), major (10.0% vs 2.4%, p < 0.001), and minor complications (24.1% vs 11.0%, p < 0.001) as well as readmissions (12.7% vs 5.1%, p < 0.001). Obese hypoalbuminemic patients had similar rates of complication as non-obese hypoalbuminemic patients. Multivariable regression showed that hypoalbuminemia had an odds ratio of 1.85(p = 0.003) for predicting any complication within 30 days of surgery. CONCLUSION: Hypoalbuminemia is associated with higher risk for complications and readmission after PHFs. It occurs more frequently in patients with chronic disease and is predictive of malnourishment. Paradoxically, hypoalbuminemia is not uncommon in obese patients. LEVEL OF EVIDENCE: III; Retrospective Cohort Study.

Comparison of powered drill & manual bone biopsy systems: Does the diagnostic yield justify the cost?

Bone biopsies are traditionally performed using manual drill devices. More recently, powered drill biopsy systems have been developed, ostensibly to improve diagnostic yield. We compare the powered drill biopsy system to traditional manual needle devices on the basis of diagnostic yield, specimen size and material costs. 309 consecutive bone biopsy procedures performed with imaging guidance from a single academic institution were retrospectively reviewed. Specimen diagnostic adequacy, qualitative interpretation of diagnostically inadequate specimens, aggregate specimen volume and material costs were assessed. Econometric analysis was performed to assess the relationship between materials cost and diagnostic yield. Diagnostic yield and average core specimen volume were significantly higher in the powered drill group, with 86% of cases yielding adequate biopsy specimens versus 67% of cases using the manual method. The materials cost associated with the powered drill device was higher than those of any of the manual needle devices with an average difference of $270.19 per case, however, this was offset due to higher diagnostic yield suggesting that the increased diagnostic accuracy achieved by the powered drill does not carry a significant added financial burden. The powered drill bone biopsy system results in a significantly higher yield of diagnostically adequate biopsy specimens compared to traditional manual needles, possibly attributed to larger and more intact obtained core specimen volumes and is more economically viable off-setting the higher cost.

Deep Learning and Transfer Learning for Optic Disc Laterality Detection: Implications for Machine Learning in Neuro-Ophthalmology.

BACKGROUND: Deep learning (DL) has demonstrated human expert levels of performance for medical image classification in a wide array of medical fields, including ophthalmology. In this article, we present the results of our DL system designed to determine optic disc laterality, right eye vs left eye, in the presence of both normal and abnormal optic discs. METHODS: Using transfer learning, we modified the ResNet-152 deep convolutional neural network (DCNN), pretrained on ImageNet, to determine the optic disc laterality. After a 5-fold cross-validation, we generated receiver operating characteristic curves and corresponding area under the curve (AUC) values to evaluate performance. The data set consisted of 576 color fundus photographs (51% right and 49% left). Both 30° photographs centered on the optic disc (63%) and photographs with varying degree of optic disc centration and/or wider field of view (37%) were included. Both normal (27%) and abnormal (73%) optic discs were included. Various neuro-ophthalmological diseases were represented, such as, but not limited to, atrophy, anterior ischemic optic neuropathy, hypoplasia, and papilledema. RESULTS: Using 5-fold cross-validation (70% training; 10% validation; 20% testing), our DCNN for classifying right vs left optic disc achieved an average AUC of 0.999 (±0.002) with optimal threshold values, yielding an average accuracy of 98.78% (±1.52%), sensitivity of 98.60% (±1.72%), and specificity of 98.97% (±1.38%). When tested against a separate data set for external validation, our 5-fold cross-validation model achieved the following average performance: AUC 0.996 (±0.005), accuracy 97.2% (±2.0%), sensitivity 96.4% (±4.3%), and specificity 98.0% (±2.2%). CONCLUSIONS: Small data sets can be used to develop high-performing DL systems for semantic labeling of neuro-ophthalmology images, specifically in distinguishing between right and left optic discs, even in the presence of neuro-ophthalmological pathologies. Although this may seem like an elementary task, this study demonstrates the power of transfer learning and provides an example of a DCNN that can help curate large medical image databases for machine-learning purposes and facilitate ophthalmologist workflow by automatically labeling images according to laterality.

Diagnostic and Interventional Imaging Services are Significant Sources of Medicare Revenue for Highly Reimbursed Nonradiologist Providers.

INTRODUCTION: Nonradiologist providers increasingly perform diagnostic imaging examinations and imaging-guided interventions traditionally performed by radiologists, which have raised concerns regarding appropriate utilization and self-referral. The purpose of this study was to assess the contribution of imaging studies to Medicare reimbursements for highly compensated nonradiologist providers in specialties often performing imaging studies. METHODS: The Medicare Provider Utilization and Payment Database was queried for provider information regarding overall reimbursement for providers in anesthesiology, cardiology, emergency medicine, neurology, obstetrics and gynecology, orthopedic surgery, neurology, and vascular surgery. Information regarding imaging studies reported and payment amounts were extracted for the 25 highest-reimbursed providers. Data were analyzed for relative contribution of imaging payments to overall medical Medicare payments. RESULTS: Significant differences between numbers of imaging studies, types of imaging, and payment amounts were noted based on provider specialty (p < 0.001). Highest-reimbursed cardiologists received the greatest percentage of Medicare payments from imaging (18.3%) followed by vascular surgery (11.6%), obstetrics and gynecology (10.9%), orthopedic surgery (9.6%), emergency medicine (8.7%), neurology (7.8%), and anesthesiology (3.2%) providers. Mean imaging payments amongst highly reimbursed nonradiologists were greatest for cardiology ($578,265), vascular surgery ($363,912), and orthopedic surgery ($113,634). Amongst highly reimbursed specialists, most common nonradiologist imaging payments were from ultrasound (45%) and cardiac nuclear medicine studies (40%). CONCLUSIONS: Nonradiologist performed imaging payments comprised substantial proportions of overall Medicare reimbursement for highly reimbursed physicians in several specialties, especially cardiology, vascular surgery, and orthopedic surgery. Further investigation is needed to better understand the wider economic implications of nonradiologist imaging study performance and self-referral beyond the Medicare population.

Automated detection & classification of knee arthroplasty using deep learning.

BACKGROUND: Preoperative identification of knee arthroplasty is important for planning revision surgery. However, up to 10% of implants are not identified prior to surgery. The purposes of this study were to develop and test the performance of a deep learning system (DLS) for the automated radiographic 1) identification of the presence or absence of a total knee arthroplasty (TKA); 2) classification of TKA vs. unicompartmental knee arthroplasty (UKA); and 3) differentiation between two different primary TKA models. METHOD: We collected 237 anteroposterior (AP) knee radiographs with equal proportions of native knees, TKA, and UKA and 274 AP knee radiographs with equal proportions of two TKA models. Data augmentation was used to increase the number of images for deep convolutional neural network (DCNN) training. A DLS based on DCNNs was trained on these images. Receiver operating characteristic (ROC) curves with area under the curve (AUC) were generated. Heatmaps were created using class activation mapping (CAM) to identify image features most important for DCNN decision-making. RESULTS: DCNNs trained to detect TKA and distinguish between TKA and UKA both achieved AUC of 1. Heatmaps demonstrated appropriate emphasis of arthroplasty components in decision-making. The DCNN trained to distinguish between the two TKA models achieved AUC of 1. Heatmaps showed emphasis of specific unique features of the TKA model designs, such as the femoral component anterior flange shape. CONCLUSIONS: DCNNs can accurately identify presence of TKA and distinguish between specific arthroplasty designs. This proof-of-concept could be applied towards identifying other prosthesis models and prosthesis-related complications.

Postoperative Spinal CT: What the Radiologist Needs to Know.

During the past 2 decades, the number of spinal surgeries performed annually has been steadily increasing, and these procedures are being accompanied by a growing number of postoperative imaging studies to interpret. CT is accurate for identifying the location and integrity of implants, assessing the success of decompression and intervertebral arthrodesis procedures, and detecting and characterizing related complications. Although postoperative spinal CT is often limited owing to artifacts caused by metallic implants, parameter optimization and advanced metal artifact reduction techniques, including iterative reconstruction and monoenergetic extrapolation methods, can be used to reduce metal artifact severity and improve image quality substantially. Commonly used and recently available spinal implants and prostheses include screws and wires, static and extendable rods, bone grafts and biologic materials, interbody cages, and intervertebral disk prostheses. CT assessment and the spectrum of complications that can occur after spinal surgery and intervertebral arthroplasty include those related to the position and integrity of implants and prostheses, adjacent segment degeneration, collections, fistulas, pseudomeningoceles, cerebrospinal fluid leaks, and surgical site infections. Knowledge of the numerous spinal surgery techniques and devices aids in differentiating expected postoperative findings from complications. The various types of spinal surgery instrumentation and commonly used spinal implants are reviewed. The authors also describe and illustrate normal postoperative spine findings, signs of successful surgery, and the broad spectrum of postoperative complications that can aid radiologists in generating reports that address issues that the surgeon needs to know for optimal patient management.©RSNA, 2019.

Insurance Status Affects Complication Rates After Total Hip Arthroplasty.

INTRODUCTION: Previous studies have examined the relationship between total hip arthroplasty (THA) and insurance status in small cohorts. This study evaluates the effect of patient insurance status on complications after primary elective THA using the Nationwide Inpatient Sample. METHODS: All patients undergoing primary elective THA from 1998 to 2011 were included. Patient demographics, comorbidities, and complications were collected and compared based on insurance type. Multivariable logistic regression and a matched cohort analysis were performed. RESULTS: About 515,037 patients (53.7% Medicare, 40.1% private insurance, 3.9% Medicaid/uninsured, and 2.2% other) were included, who underwent elective THA. Privately insured patients had fewer medical complications (odds ratio, 0.80; P < 0.001), whereas patients with Medicaid or no insurance demonstrated no notable difference (odds ratio, 1.03; P = 0.367) compared with Medicare patients. Similar trends were found for both surgical complications and mortality, favoring lower complication rates for privately insured patients. Furthermore, patients with private insurance tend to go to higher volume hospitals for total hip replacement surgery compared to those with Medicare insurance. DISCUSSION: Patients with government-sponsored insurance (Medicare or Medicaid) or no insurance have higher risk of medical complications, surgical complications, and mortality after primary elective THA compared with privately insured patients. Insurance status should be considered an independent risk factor for stratifying patients before THA procedures.

Deep-Learning-Based Semantic Labeling for 2D Mammography and Comparison of Complexity for Machine Learning Tasks.

Machine learning has several potential uses in medical imaging for semantic labeling of images to improve radiologist workflow and to triage studies for review. The purpose of this study was to (1) develop deep convolutional neural networks (DCNNs) for automated classification of 2D mammography views, determination of breast laterality, and assessment and of breast tissue density; and (2) compare the performance of DCNNs on these tasks of varying complexity to each other. We obtained 3034 2D-mammographic images from the Digital Database for Screening Mammography, annotated with mammographic view, image laterality, and breast tissue density. These images were used to train a DCNN to classify images for these three tasks. The DCNN trained to classify mammographic view achieved receiver-operating-characteristic (ROC) area under the curve (AUC) of 1. The DCNN trained to classify breast image laterality initially misclassified right and left breasts (AUC 0.75); however, after discontinuing horizontal flips during data augmentation, AUC improved to 0.93 (p < 0.0001). Breast density classification proved more difficult, with the DCNN achieving 68% accuracy. Automated semantic labeling of 2D mammography is feasible using DCNNs and can be performed with small datasets. However, automated classification of differences in breast density is more difficult, likely requiring larger datasets.