Australian Dentist's Knowledge and Perceptions of Factors Affecting Radiographic Interpretation

BACKGROUND

Errors of interpretation of radigraphic images, also known as interpretive errors, are a critical concern as they can have profound implications for clinical decision making. Different types of interpretive errors, including errors of omission and misdiagnosis, have been described in the literature. These errors can lead to unnecessary or harmful treat/or prolonged patient care. Understanding the nature and contributing factors of interpretive errors is important in developing solutions to minimise interpretive errors. By exploring the knowledge and perceptions of dental practitioners, this study aimed to shed light on the current understanding of interpretive errors in dentistry.

METHODS

An anonymised online questionnaire was sent to dental practitioners in New South Wales (NSW) between September 2020 and March 2022. A total of 80 valid responses were received and analysed. Descriptive statistics and bivariate analysis were used to analyse the data.

RESULTS

The study found that participants commonly reported interpretive errors as occurring 'occasionally', with errors of omission being the most frequently encountered type. Participants identified several factors that most likely contribute to interpretive errors, including reading a poor-quality image, lack of clinical experience and knowledge, and excessive workload. Additionally, general practitioners and specialists held different views regarding factors affecting interpretive errors.

CONCLUSION

The survey results indicate that dental practitioners are aware of the common factors associated with interpretive errors. Errors of omission were identified as the most common type of error to occur in clinical practice. The findings suggest that interpretive errors result from a mental overload caused by factors associated with image quality, clinician-related, and image interpretation. Managing and identifying solutions to mitigate these factors are crucial for ensuring accurate and timely radiographic diagnoses. The findings of this study can serve as a foundation for future research and the development of targeted interventions to enhance the accuracy of radiographic interpretations in dentistry.

Utilisation of radiographer comments to reduce errors in the radiology department

INTRODUCTION

Radiographer commenting is a written account of suspected abnormalities identified on medical imaging examinations by the radiographer at the time of image acquisition. Radiographer comments were originally implemented to support emergency clinicians; however, they may also have the potential to support radiologists in reducing missed findings. Therefore, the aim of this study was to investigate if a newly implemented radiographer comment system could reduce the number of errors made in radiology reports for general X-rays. Incidental findings from multisite collaborative research led to the hypothesis that in some cases radiographer comments could accurately detect abnormal X-ray appearances that were not otherwise documented in the radiologist report, thereby enabling results to be revised and errors collaboratively reduced [1].

METHODS

This study was conducted at an 800-bed hospital, where 92% of general radiographers self-selected to participate. Radiographer comments were provided to referring physicians through the electronic medical record and could be made for any emergency or inpatient general X-ray examinations. All comments made over a 12-month period were audited against the corresponding radiologist report. Radiologists were blinded to radiographer comments at the time of reporting. Where discrepancies between the radiographer comment and radiologist report arose, additional radiologist review or subsequent imaging reports were used to determine the accurate interpretation. The number of discrepant radiographer comments that were deemed true positive (TP) and provided new and correct diagnostic information compared to the radiologist report were identified. These were converted to a percentage of total radiographer comments that were therefore able to positively influence radiologist report accuracy. The number of discrepant cases where radiographer comments were deemed false positive (FP) was also measured and converted to a percentage of the total comments. Confidence intervals for both TP and FP binomial proportions were calculated using the Wilson Score Interval.

RESULTS

Over 12 months, 282 radiographer comments were made to alert clinically significant radiographic appearances on general X-ray. Of these, 32 radiographer comments were discrepant with the report. Of these 32 comments, 24 were deemed TP meaning they correctly identified a pathological imaging appearance that was not otherwise documented in the radiology report. Therefore, 8.5% of all radiographer comments added value by correctly identifying a pathology that was not otherwise documented, 95% CI (5.8% - 12.4%). This enabled results to be promptly amended and reporting errors collaboratively reduced. Conversely, eight (2.8%) radiographer comments were discrepant with the report but deemed FP and did not add value to the investigation, 95% CI (1.4% - 5.5%). The remaining 250 non-discrepant comments did not contribute to error reduction but provided real-time abnormality detection that benefitted managing teams.

CONCLUSION

These findings are consistent with previous literature proposing radiographer comments may provide a safety net for radiologists due to factors such as direct patient contact, ability to expand on clinical history, and difference in accumulated expertise. This study demonstrates that radiographer comments may be effectively used as a multidisciplinary error-reduction tool to assist radiologists in their important role and improve clinical outcomes.

Assessing the barriers and enablers to the implementation of the diagnostic radiographer musculoskeletal X-ray reporting service within the NHS in England: a systematic literature review

INTRODUCTION

The United Kingdom (UK) government's healthcare policy in the early 1990s paved the way adoption of the skills mix development and implementation of diagnostic radiographers' X-ray reporting service. Current clinical practice within the public UK healthcare system reflects the same pressures of increased demand in patient imaging and limited capacity of the reporting workforce (radiographers and radiologists) as in the 1990s. This study aimed to identify, define and assess the longitudinal macro, meso, and micro barriers and enablers to the implementation of the diagnostic radiographer musculoskeletal X-ray reporting service in the National Healthcare System (NHS) in England.

METHODS

Multiple independent databases were searched, including PubMed, Ovid MEDLINE; Embase; CINAHL, and Google Scholar, as well as journal databases (Scopus, Wiley), healthcare databases (NHS Evidence Database; Cochrane Library) and grey literature databases (OpenGrey, GreyNet International, and the British Library EthOS depository) and recorded in a PRISMA flow chart. A combination of keywords, Boolean logic, truncation, parentheses and wildcards with inclusion/exclusion criteria and a time frame of 1995-2022 was applied. The literature was assessed against Joanna Briggs Institute's critical appraisal checklists. With meta-aggregation to synthesize each paper, and coded using NVivo, with context grouped into macro, meso, and micro-level sources and categorised into subgroups of enablers and barriers.

RESULTS

The wide and diverse range of data (n = 241 papers) identified barriers and enablers of implementation, which were categorised into measures of macro, meso, and micro levels, and thematic categories of context, culture, environment, and leadership.

CONCLUSION

The literature since 1995 has reframed the debates on implementation of the radiographer reporting role and has been instrumental in shaping clinical practice. There has been clear influence upon both meso (professional body) and macro-level (governmental/health service) policies and guidance, that have shaped change at micro-level NHS Trust organisations. There is evidence of a shift in culturally intrenched legacy perspectives within and between different meso-level professional bodies around skills mix acceptance and role boundaries. This has helped shape capacity building of the reporting workforce. All of which have contributed to conceptual understandings of the skills mix workforce within modern radiology services.

Breast Cancer Detection with an Ensemble of Deep Learning Networks Using a Consensus-Adaptive Weighting Method

Breast cancer's high mortality rate is often linked to late diagnosis, with mammograms as key but sometimes limited tools in early detection. To enhance diagnostic accuracy and speed, this study introduces a novel computer-aided detection (CAD) ensemble system. This system incorporates advanced deep learning networks-EfficientNet, Xception, MobileNetV2, InceptionV3, and Resnet50-integrated via our innovative consensus-adaptive weighting (CAW) method. This method permits the dynamic adjustment of multiple deep networks, bolstering the system's detection capabilities. Our approach also addresses a major challenge in pixel-level data annotation of faster R-CNNs, highlighted in a prominent previous study. Evaluations on various datasets, including the cropped DDSM (Digital Database for Screening Mammography), DDSM, and INbreast, demonstrated the system's superior performance. In particular, our CAD system showed marked improvement on the cropped DDSM dataset, enhancing detection rates by approximately 1.59% and achieving an accuracy of 95.48%. This innovative system represents a significant advancement in early breast cancer detection, offering the potential for more precise and timely diagnosis, ultimately fostering improved patient outcomes.

Analysis of perceptual errors in skull-base pathology

PURPOSE

Medical errors result in significant mortality and morbidity. The purpose of this study is to analyze skull-base errors at a single tertiary institution, identify common anatomic sites of errors, and offer strategies to reduce errors in this region.

METHODS

A Neuroradiology Quality Assurance Database of radiologic errors was searched for attending physician computer tomography and magnetic resonance imaging errors in skull-base pathology from 2014 to 2020. Data were limited to CT and MRI reports. Errors were separated into four subcategories (tumor, trauma, vascular, and congenital) and further divided by relevant anatomic site.

RESULTS

A total of 90 skull-based errors were identified. Most errors were perceptual (87%), with common study types including MRI Brain (39%) and CT Head (24%). Most common errors were tumors (55%), followed by trauma (24%), vascular (10%), and congenital (7%). Six anatomic sites were identified and encompassed over half of errors (58%): sella, occipital bone, cerebellopontine angle/internal auditory canal (CPA/IAC), foramen magnum and clivus, cavernous sinus, and dural venous sinus.

SUMMARY

Most of the skull-base errors were perceptual. Placing a strong emphasis on both the pathology and closely examining its critical anatomic site (sella, occipital bone, CPA/IAC, foramen magnum and clivus, cavernous sinus, and dural venous sinus) could potentially reduce up to 60% of errors in these regions.

A Review of Deep Learning Techniques for Lung Cancer Screening and Diagnosis Based on CT Images

One of the most common and deadly diseases in the world is lung cancer. Only early identification of lung cancer can increase a patient's probability of survival. A frequently used modality for the screening and diagnosis of lung cancer is computed tomography (CT) imaging, which provides a detailed scan of the lung. In line with the advancement of computer-assisted systems, deep learning techniques have been extensively explored to help in interpreting the CT images for lung cancer identification. Hence, the goal of this review is to provide a detailed review of the deep learning techniques that were developed for screening and diagnosing lung cancer. This review covers an overview of deep learning (DL) techniques, the suggested DL techniques for lung cancer applications, and the novelties of the reviewed methods. This review focuses on two main methodologies of deep learning in screening and diagnosing lung cancer, which are classification and segmentation methodologies. The advantages and shortcomings of current deep learning models will also be discussed. The resultant analysis demonstrates that there is a significant potential for deep learning methods to provide precise and effective computer-assisted lung cancer screening and diagnosis using CT scans. At the end of this review, a list of potential future works regarding improving the application of deep learning is provided to spearhead the advancement of computer-assisted lung cancer diagnosis systems.

Impact of Shift Volume on Neuroradiology Diagnostic Errors at a Large Tertiary Academic Center

BACKGROUND AND PURPOSE

Medical errors can result in significant morbidity and mortality. The goal of our study is to evaluate correlation between shift volume and errors made by attending neuroradiologists at an academic medical center, using a large data set.

MATERIALS AND METHODS

CT and MRI reports from our Neuroradiology Quality Assurance database (years 2014 - 2020) were searched for attending physician errors. Data were collected on shift volume, category of missed findings, error type, interpretation setting, exam type, clinical significance.

RESULTS

654 reports contained diagnostic error. There was a significant difference between mean volume of interpreted studies on shifts when an error was made compared with shifts in which no error was documented (46.58 (SD=22.37) vs 34.09 (SD=18.60), p<0.00001); and between shifts when perceptual error was made compared with shifts when interpretive errors were made (49.50 (SD=21.9) vs 43.26 (SD=21.75), p=0.0094). 59.6% of errors occurred in the emergency/inpatient setting, 84% were perceptual and 91.1% clinically significant. Categorical distribution of errors was: vascular 25.8%, brain 23.4%, skull base 13.8%, spine 12.4%, head/neck 11.3%, fractures 10.2%, other 3.1%. Errors were detected most often on brain MRI (25.4%), head CT (18.7%), head/neck CTA (13.8%), spine MRI (13.7%).

CONCLUSION

Errors were associated with higher volume shifts, were primarily perceptual and clinically significant. We need National guidelines establishing a range of what is a safe number of interpreted cross-sectional studies per day.

Perceptual omission errors in positron emission tomography and computed tomography reporting

BACKGROUND

Omission errors in medical imaging can lead to missed diagnosis and harm to patients. The subject has been studied in conventional imaging, but no data is available for functional imaging in general and for PET/CT in particular. In this work, we evaluated the frequency and characteristics of perceptual omission errors in the PET component of oncologic PET/CT imaging, and we analyzed the hazardous scenarios prone to such modality-specific errors.

METHODS

Perceptual omission errors were collected in one tertiary center PET/CT clinic during routine PET/CT reporting over a 26-month period. The omissions were detected either in reporting follow-up PET/CT studies of the same patient or during multidisciplinary meetings.

RESULTS

Significant omission errors were found in 1.2% of the 2100 reports included in the study. The most common omissions were bone metastases and focal colon uptake. We identified six PET-specific causative factors contributing to the occurrence of omissions, and we propose solutions to minimize their influence.

CONCLUSIONS

The data presented here can help to promote the awareness of interpreting physicians to body areas that require higher attention and to implement reading strategies for improving the accuracy of PET/CT interpretation.

Commonly Missed Findings on Chest Radiographs: Causes and Consequences

Chest radiography (CXR) continues to be the most frequently performed imaging examination worldwide, yet it remains prone to frequent errors in interpretation. These pose potential adverse consequences to patients and are a leading motivation for medical malpractice lawsuits. Commonly missed CXR findings and the principal causes of these errors are reviewed and illustrated. Perceptual errors are the predominant source of these missed findings. The medicolegal implications of such errors are explained. Awareness of commonly missed CXR findings, their causes, and their consequences are important in developing approaches to reduce and mitigate these errors.

Missed Radiological Diagnosis of Otosclerosis in High-Resolution Computed Tomography of the Temporal Bone-Retrospective Analysis of Imaging, Radiological Reports, and Request Forms

Objectives: Several studies reported low detection rates of otosclerosis in high-resolution computed tomography (HRCT), especially when the scans were reviewed by non-specialized general radiologists. In the present study, we conducted a retrospective review of the detection of otosclerosis in HRCT by general radiologists and the impact of inadequately filled radiological request forms on the detection rate. Methods: Retrospective analysis of hospital records, HRCT reports, and radiological referral notes of 40 patients who underwent stapedotomy surgery for otosclerosis. HRCT imaging data sets were retrospectively reviewed by a blinded experienced neuroradiologist, whose reading served as the gold standard. Results: General radiologists reading HRCT scans had an overall detection rate of otosclerosis of 36.1% in this cohort (13 of 36 available HRCT reports). The neuroradiologist had a much higher detection rate of 82.5% (33 of 40 cases). Interobserver agreement between the general radiologists and the subspecialist neuroradiologist was poor (Cohen’s kappa κ = 0.26). General radiologists missed the diagnosis in 15 of the 33 CT-positive scans, corresponding to a missed diagnosis rate of 45.4%. There was a highly significant association between a missed diagnosis and the lack of an explicitly mentioned clinical suspicion of otosclerosis in the request forms (Pearson’s chi-squared test, p < 0.005). Conclusion: The diagnosis of otosclerosis is frequently missed by radiologists on HRCT scans of the temporal bone in a clinical setting. Possible reasons include a relative lack of experience of general radiologists with temporal bone imaging as well as the failure of clinicians to unambiguously communicate their suspicion of otosclerosis.

Factors affecting interpretation of dental radiographs

OBJECTIVES

To identify the factors influencing errors in the interpretation of dental radiographs.

METHODS

A protocol was registered on Prospero. All studies published until May 2022 were included in this review. The search of the electronic databases spanned Ovid Medline, PubMed, EMBASE, Web of Science and Scopus. The quality of the studies was assessed using the MMAT tool. Due to the heterogeneity of the included studies, a meta-analysis was not conducted.

RESULTS

The search yielded 858 articles, of which eight papers met the inclusion and exclusion criteria and were included in the systematic review. These studies assessed the factors influencing the accuracy of the interpretation of dental radiographs. Six factors were identified as being significant that affected the occurrence of interpretation errors. These include clinical experience, clinical knowledge, and technical ability, case complexity, time pressure, location and duration of dental education and training and cognitive load.

CONCLUSIONS

The occurrence of interpretation errors has not been widely investigated in dentistry. The factors identified in this review are interlinked. Further studies are needed to better understand the extent of the occurrence of interpretive errors and their impact on the practice of dentistry.

Diagnostic Errors in Cerebrovascular Pathology: Retrospective Analysis of a Neuroradiology Database at a Large Tertiary Academic Medical Center

BACKGROUND AND PURPOSE

Diagnostic errors affect 2%-8% of neuroradiology studies, resulting in significant potential morbidity and mortality. This retrospective analysis of a large database at a single tertiary academic institution focuses on diagnostic misses in cerebrovascular pathology and suggests error-reduction strategies.

MATERIALS AND METHODS

CT and MR imaging reports from a consecutive database spanning 2015-2020 were searched for errors of attending physicians in cerebrovascular pathology. Data were collected on missed findings, study types, and interpretation settings. Errors were categorized as ischemic, arterial, venous, hemorrhagic, and "other."

RESULTS

A total of 245,762 CT and MR imaging neuroradiology examinations were interpreted during the study period. Vascular diagnostic errors were present in 165 reports, with a mean of 49.6 (SD, 23.3) studies on the shifts when an error was made, compared with 34.9 (SD, 19.2) on shifts without detected errors (P < .0001). Seventy percent of examinations occurred in the hospital setting; 93.3% of errors were perceptual; 6.7% were interpretive; and 93.9% (n = 155) were clinically significant (RADPEER 2B or 3B). The distribution of errors was arterial and ischemic each with 33.3%, hemorrhagic with 21.8%, and venous with 7.5%. Most errors involved brain MR imaging (30.3%) followed by head CTA (27.9%) and noncontrast head CT (26.1%). The most common misses were acute/subacute infarcts (25.1%), followed by aneurysms (13.7%) and subdural hematomas (9.7%).

CONCLUSIONS

Most cerebrovascular diagnostic errors were perceptual and clinically significant, occurred in the emergency/inpatient setting, and were associated with higher-volume shifts. Diagnostic errors could be minimized by adjusting search patterns to ensure vigilance on the sites of the frequently missed pathologies.

Artificial Intelligence for the Analysis of Workload-Related Changes in Radiologists' Gaze Patterns

Around 60-80% of radiological errors are attributed to overlooked abnormalities, the rate of which increases at the end of work shifts. In this study, we run an experiment to investigate if artificial intelligence (AI) can assist in detecting radiologists' gaze patterns that correlate with fatigue. A retrospective database of lung X-ray images with the reference diagnoses was used. The X-ray images were acquired from 400 subjects with a mean age of 49 ± 17, and 61% men. Four practicing radiologists read these images while their eye movements were recorded. The radiologists passed a series of concentration tests at prearranged breaks of the experiment. A U-Net neural network was adapted to annotate lung anatomy on X-rays and calculate coverage and information gain features from the radiologists' eye movements over lung fields. The lung coverage, information gain, and eye tracker-based features were compared with the cumulative work done (CDW) label for each radiologist. The gaze-traveled distance, X-ray coverage, and lung coverage statistically significantly (p < 0.01) deteriorated with cumulative work done (CWD) for three out of four radiologists. The reading time and information gain over lungs statistically significantly deteriorated for all four radiologists. We discovered a novel AI-based metric blending reading time, speed, and organ coverage, which can be used to predict changes in the fatigue-related image reading patterns.

Artificial intelligence system for identification of false-negative interpretations in chest radiographs

OBJECTIVES

To investigate the efficacy of an artificial intelligence (AI) system for the identification of false negatives in chest radiographs that were interpreted as normal by radiologists.

METHODS

We consecutively collected chest radiographs that were read as normal during 1 month (March 2020) in a single institution. A commercialized AI system was retrospectively applied to these radiographs. Radiographs with abnormal AI results were then re-interpreted by the radiologist who initially read the radiograph ("AI as the advisor" scenario). The reference standards for the true presence of relevant abnormalities in radiographs were defined by majority voting of three thoracic radiologists. The efficacy of the AI system was evaluated by detection yield (proportion of true-positive identification among the entire examination) and false-referral rate (FRR, proportion of false-positive identification among all examinations). Decision curve analyses were performed to evaluate the net benefits of applying the AI system.

RESULTS

A total of 4208 radiographs from 3778 patients (M:F = 1542:2236; median age, 56 years) were included. The AI system identified initially overlooked relevant abnormalities with a detection yield and an FRR of 2.4% and 14.0%, respectively. In the "AI as the advisor" scenario, radiologists detected initially overlooked relevant abnormalities with a detection yield and FRR of 1.2% and 0.97%, respectively. In a decision curve analysis, AI as an advisor scenario exhibited a positive net benefit when the cost-to-benefit ratio was below 1:0.8.

CONCLUSION

An AI system could identify relevant abnormalities overlooked by radiologists and could enable radiologists to correct their false-negative interpretations by providing feedback to radiologists.

KEY POINTS

• In consecutive chest radiographs with normal interpretations, an artificial intelligence system could identify relevant abnormalities that were initially overlooked by radiologists. • The artificial intelligence system could enable radiologists to correct their initial false-negative interpretations by providing feedback to radiologists when overlooked abnormalities were present.

Faster and Better: How Anomaly Detection Can Accelerate and Improve Reporting of Head Computed Tomography

Background: Most artificial intelligence (AI) systems are restricted to solving a pre-defined task, thus limiting their generalizability to unselected datasets. Anomaly detection relieves this shortfall by flagging all pathologies as deviations from a learned norm. Here, we investigate whether diagnostic accuracy and reporting times can be improved by an anomaly detection tool for head computed tomography (CT), tailored to provide patient-level triage and voxel-based highlighting of pathologies. Methods: Four neuroradiologists with 1–10 years of experience each investigated a set of 80 routinely acquired head CTs containing 40 normal scans and 40 scans with common pathologies. In a random order, scans were investigated with and without AI-predictions. A 4-week wash-out period between runs was included to prevent a reminiscence effect. Performance metrics for identifying pathologies, reporting times, and subjectively assessed diagnostic confidence were determined for both runs. Results: AI-support significantly increased the share of correctly classified scans (normal/pathological) from 309/320 scans to 317/320 scans (p = 0.0045), with a corresponding sensitivity, specificity, negative- and positive- predictive value of 100%, 98.1%, 98.2% and 100%, respectively. Further, reporting was significantly accelerated with AI-support, as evidenced by the 15.7% reduction in reporting times (65.1 ± 8.9 s vs. 54.9 ± 7.1 s; p < 0.0001). Diagnostic confidence was similar in both runs. Conclusion: Our study shows that AI-based triage of CTs can improve the diagnostic accuracy and accelerate reporting for experienced and inexperienced radiologists alike. Through ad hoc identification of normal CTs, anomaly detection promises to guide clinicians towards scans requiring urgent attention.

Deep learning-based algorithm for lung cancer detection on chest radiographs using the segmentation method

We developed and validated a deep learning (DL)-based model using the segmentation method and assessed its ability to detect lung cancer on chest radiographs. Chest radiographs for use as a training dataset and a test dataset were collected separately from January 2006 to June 2018 at our hospital. The training dataset was used to train and validate the DL-based model with five-fold cross-validation. The model sensitivity and mean false positive indications per image (mFPI) were assessed with the independent test dataset. The training dataset included 629 radiographs with 652 nodules/masses and the test dataset included 151 radiographs with 159 nodules/masses. The DL-based model had a sensitivity of 0.73 with 0.13 mFPI in the test dataset. Sensitivity was lower in lung cancers that overlapped with blind spots such as pulmonary apices, pulmonary hila, chest wall, heart, and sub-diaphragmatic space (0.50–0.64) compared with those in non-overlapped locations (0.87). The dice coefficient for the 159 malignant lesions was on average 0.52. The DL-based model was able to detect lung cancers on chest radiographs, with low mFPI.

Weakly supervised pneumonia localization in chest X-rays using generative adversarial networks

PURPOSE

Automatic localization of pneumonia on chest X-rays (CXRs) is highly desirable both as an interpretive aid to the radiologist and for timely diagnosis of the disease. However, pneumonia's amorphous appearance on CXRs and complexity of normal anatomy in the chest present key challenges that hinder accurate localization. Existing studies in this area are either not optimized to preserve spatial information of abnormality or depend on expensive expert-annotated bounding boxes. We present a novel generative adversarial network (GAN)-based machine learning approach for this problem, which is weakly supervised (does not require any location annotations), was trained to retain spatial information, and can produce pixel-wise abnormality maps highlighting regions of abnormality (as opposed to bounding boxes around abnormality).

METHODS

Our method is based on the Wasserstein GAN framework and, to the best of our knowledge, the first application of GANs to this problem. Specifically, from an abnormal CXR as input, we generated the corresponding pseudo normal CXR image as output. The pseudo normal CXR is the "hypothetical" normal, if the same abnormal CXR were not to have any abnormalities. We surmise that the difference between the pseudo normal and the abnormal CXR highlights the pixels suspected to have pneumonia and hence is our output abnormality map. We trained our algorithm on an "unpaired" data set of abnormal and normal CXRs and did not require any location annotations such as bounding boxes/segmentations of abnormal regions. Furthermore, we incorporated additional prior knowledge/constraints into the model and showed that they help improve localization performance. We validated the model on a data set consisting of 14 184 CXRs from the Radiological Society of North America pneumonia detection challenge.

RESULTS

We evaluated our methods by comparing the generated abnormality maps with radiologist annotated bounding boxes using receiver operating characteristic (ROC) analysis, image similarity metrics such as normalized cross-correlation/mutual information, and abnormality detection rate.We also present visual examples of the abnormality maps, covering various scenarios of abnormality occurrence. Results demonstrate the ability to highlight regions of abnormality with the best method achieving an ROC area under the curve (AUC) of 0.77 and a detection rate of 85%.The GAN tended to perform better as prior knowledge/constraints were incorporated into the model.

CONCLUSIONS

We presented a novel GAN based approach for localizing pneumonia on CXRs that (1) does not require expensive hand annotated location ground truth; and (2) was trained to produce abnormality maps at the pixel level as opposed to bounding boxes. We demonstrated the efficacy of our methods via quantitative and qualitative results.

Diagnostic Errors in Neuroradiology: A Message to Emergency Radiologists and Trainees

Diagnostic errors in neuroradiology are inevitable, yet potentially avoidable. Through extensive literature search, we present an up-to-date review of the psychology of human decision making and how such complex process can lead to radiologic errors. Our focus is on neuroradiology, so we augmented our review with multiple explanatory figures to show how different errors can reflect on real-life clinical practice. We propose a new thematic categorization of perceptual and cognitive biases in this article to simplify message delivery to our target audience: emergency/general radiologists and trainees. Additionally, we highlight individual and organizational remedy strategies to decrease error rate and potential harm.

Optimising breast cancer screening reading: blinding the second reader to the first reader's decisions

Objectives

In breast cancer screening, two readers separately examine each woman’s mammograms for signs of cancer. We examined whether preventing the two readers from seeing each other’s decisions (blinding) affects behaviour and outcomes.

Methods

This cohort study used data from the CO-OPS breast-screening trial (1,119,191 women from 43 screening centres in England) where all discrepant readings were arbitrated. Multilevel models were fitted using Markov chain Monte Carlo to measure whether reader 2 conformed to the decisions of reader 1 when they were not blinded, and the effect of blinding on overall rates of recall for further tests and cancer detection. Differences in positive predictive value (PPV) were assessed using Pearson’s chi-squared test.

Results

When reader 1 recalls, the probability of reader 2 also recalling was higher when not blinded than when blinded, suggesting readers may be influenced by the other’s decision. Overall, women were less likely to be recalled when reader 2 was blinded (OR 0.923; 95% credible interval 0.864, 0.986), with no clear pattern in cancer detection rate (OR 1.029; 95% credible interval 0.970, 1.089; Bayesian p value 0.832). PPV was 22.1% for blinded versus 20.6% for not blinded (p < 0.001).

Conclusions

Our results suggest that when not blinded, reader 2 is influenced by reader 1’s decisions to recall (alliterative bias) which would result in bypassing arbitration and negate some of the benefits of double-reading. We found a relationship between blinding the second reader and slightly higher PPV of breast cancer screening, although this analysis may be confounded by other centre characteristics.

Key Points

• In Europe, it is recommended that breast screening mammograms are analysed by two readers but there is little evidence on the effect of ‘blinding’ the readers so they cannot see each other’s decisions.

• We found evidence that when the second reader is not blinded, they are more likely to agree with a recall decision from the first reader and less likely to make an independent judgement (alliterative error). This may reduce overall accuracy through bypassing arbitration.

• This observational study suggests an association between blinding the second reader and higher positive predictive value of screening, but this may be confounded by centre characteristics.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00330-021-07965-z.

A hybrid deep learning model for breast cancer diagnosis based on transfer learning and pulse-coupled neural networks

Radiology experts often face difficulties in mammography mass lesion labeling, which may lead to conclusive yet unnecessary and expensive breast biopsies. This paper focuses on building an automated diagnosis tool that supports radiologists in identifying and classifying mammography mass lesions. The paper's main contribution is to design a hybrid model based on Pulse-Coupled Neural Networks (PCNN) and Deep Convolutional Neural Networks (CNN). Due to the need for large datasets to train and tune CNNs, which are not available for medical images, Transfer Learning (TL) was exploited in this research. TL can be an effective approach when working with small-sized datasets. The paper's implementation was tested on three public benchmark datasets: DDMS, INbreast, and BCDR datasets for training and testing and MIAS for testing only. The results indicated the enhancement that PCNN provides when combined with CNN compared to other methods for the same public datasets. The hybrid model achieved 98.72% accuracy for DDMS, 97.5% for INbreast, and 96.94% for BCDR. To avoid overfitting, the proposed hybrid model was tested on an unseen MIAS dataset, achieving 98.77% accuracy. Other evaluation metrics are reported in the results section.

Emergency Computed Tomography: How Misinterpretations Vary According to the Periods of the Nightshift?

OBJECTIVE

To evaluate the accuracy of initial computed tomography (CT) interpretations made by radiology residents during nightshifts in the emergency department.

METHODS

Preliminary CT reports performed by radiology residents during 120 consecutive nightshifts (08:30 pm to 08:30 am) were reviewed, attendings' final interpretation being the reference standard. Nightshifts were divided into four consecutive periods of 3 hours. Major misinterpretations were related to potentially life-threatening conditions if not treated immediately after CT. The rate of misinterpretations was calculated for all CT examinations, separately for nightshift's periods and for residents' training years.

RESULTS

Misinterpretations were recorded in 155 (7.4%) of 2102 CT examinations, 0.6% (13/2102) were major. There were 2.2% (4/186) major misinterpretations that occurred during the last period of the nightshift versus 0.4% (9/1916) during the first periods of the night (P < 0.05). Of all misinterpretations, 8.5% (130/1526) were made by third- and fourth-year residents and 4.3% (25/576) by fifth-year residents (P < 0.005).

CONCLUSIONS

Major misinterpretations occur at the end of the nightshift, which may be explained by the fatigue effect. The rate of misinterpretations is lower among fifth-year residents, which may be related to their prior experience in reading emergency cases.

What do radiologists look for? Advances and limitations of perceptual learning in radiologic search

Supported by guidance from training during residency programs, radiologists learn clinically relevant visual features by viewing thousands of medical images. Yet the precise visual features that expert radiologists use in their clinical practice remain unknown. Identifying such features would allow the development of perceptual learning training methods targeted to the optimization of radiology training and the reduction of medical error. Here we review attempts to bridge current gaps in understanding with a focus on computational saliency models that characterize and predict gaze behavior in radiologists. There have been great strides toward the accurate prediction of relevant medical information within images, thereby facilitating the development of novel computer-aided detection and diagnostic tools. In some cases, computational models have achieved equivalent sensitivity to that of radiologists, suggesting that we may be close to identifying the underlying visual representations that radiologists use. However, because the relevant bottom-up features vary across task context and imaging modalities, it will also be necessary to identify relevant top-down factors before perceptual expertise in radiology can be fully understood. Progress along these dimensions will improve the tools available for educating new generations of radiologists, and aid in the detection of medically relevant information, ultimately improving patient health.

Analysis of Perceptual Expertise in Radiology - Current Knowledge and a New Perspective

Radiologists rely principally on visual inspection to detect, describe, and classify findings in medical images. As most interpretive errors in radiology are perceptual in nature, understanding the path to radiologic expertise during image analysis is essential to educate future generations of radiologists. We review the perceptual tasks and challenges in radiologic diagnosis, discuss models of radiologic image perception, consider the application of perceptual learning methods in medical training, and suggest a new approach to understanding perceptional expertise. Specific principled enhancements to educational practices in radiology promise to deepen perceptual expertise among radiologists with the goal of improving training and reducing medical error.

Intraosteal Behavior of Porous Scaffolds: The mCT Raw-Data Analysis as a Tool for Better Understanding

The aim of the study is to determine the existing correlation between high-resolution 3D imaging technique obtained through Micro Computed Tomography (mCT) and histological-histomorphometric images to determine in vivo bone osteogenic behavior of bioceramic scaffolds. A Ca-Si-P scaffold ceramic doped and non-doped (control) with a natural demineralized bone matrix (DBM) were implanted in rabbit tibias for 1, 3, and 5 months. A progressive disorganization and disintegration of scaffolds and bone neoformation occurs, from the periphery to the center of the implants, without any differences between histomorphometric and radiological analysis. However, significant differences (p < 0.05) between DMB-doped and non-doped materials where only detected through mathematical analysis of mCT. In this way, average attenuation coefficient for DMB-doped decreased from 0.99 ± 0.23 Hounsfield Unit (HU) (3 months) to 0.86 ± 0.32 HU (5 months). Average values for non-doped decreased from 0.86 ± 0.25 HU (3 months) to 0.66 ± 0.33 HU. Combination of radiological analysis and mathematical mCT seems to provide an adequate in vivo analysis of bone-implanted biomaterials after surgery, obtaining similar results to the one provided by histomorphometric analysis. Mathematical analysis of Computed Tomography (CT) would allow the conducting of long-term duration in vivo studies, without the need for animal sacrifice, and the subsequent reduction in variability.

“Learning from my experience”: Acute abdomen - Perforated Meckel's diverticulitis

•

Acute surgical abdomen has a limited differential in pediatric population.

•

Meckel’s diverticulum can mimic the clinical findings of acute appendicitis.

•

Careful radiographic review to identify the appendix in the pediatric population with acute abdomen.

•

Surgical pathology of a normal appendix and a perforated Meckel’s Diverticulitis, with a retrospective view of the appendix.

This is a case report documenting the risk in imaging misinterpretation of a pediatric patient that presented with an acute abdomen. Computed Tomography (CT) demonstrated an inflamed blind ending loop of bowel in the pelvis without an obvious cecal connection. The patient was taken to the operative theater, a normal appendix and perforated Meckel’s diverticulitis were resected. Meckel’s diverticulum is the most common small bowel abnormality and can have complications. We will emphasize that imaging studies in a pediatric patient with Meckel’s diverticulum are easily subject to errors radiologists make. This is such a cautionary and learning tale.

DXA Errors Are Common and Reduced by Use of a Reporting Template

OBJECTIVE

High quality dual energy X-ray absorptiometry (DXA) acquisition, analysis, and reporting demands technical and interpretive excellence. We hypothesized that DXA errors are common and of such magnitude that incorrect clinical decisions might result. In this 2-phase study, we evaluated DXA technical and interpretation error rates in a clinical population and subsequently assessed if implementing an interpretation template reduced errors.

METHODS

In phase 1, DXA scans of 345 osteoporosis clinic referrals were reviewed by International Society for Clinical Densitometry-certified technologists (n = 3) and physicians (n = 3). Technologists applied International Society for Clinical Densitometry performance standards to assess technical quality. Physicians assessed reporting compliance with published guidance, relevance of technical errors and determined overall and major error prevalence. Major errors were defined as "provision of inaccurate information that could potentially lead to incorrect patient care decisions." In phase 2, a DXA reporting template was implemented at 2 clinical DXA sites after which the 3 physicians reviewed 200 images and reports as above. The error prevalence was compared with the 298 patients in phase 1 from these sites.

RESULTS

In phase 1, technical errors were identified in 90% of patients and affected interpretation in 13%. Interpretation errors were present in 80% of patients; 42% were major. The most common major errors were reporting incorrect information on bone mineral density change (70%) and incorrect diagnosis (22%). In phase 2, at these 2 clinical sites, major errors were present in 37% before and 17% after template implementation. Template usage reduced the odds of major error by 66% (odds ratio 0.34, 95% confidence interval 0.21, 0.53, and p < 0.0001).

CONCLUSION

DXA technical and interpretation errors are extremely common and likely adversely affect patient care. Implementing a DXA reporting template reduces major errors and should become common practice. Additional interventions, such as requiring initial and ongoing training and/or certification for technologists and interpreters, are suggested.

Image interpretation by radiographers in brain, spine and knee MRI examinations: Findings from an accredited postgraduate module

INTRODUCTION

The aim of the study was to evaluate the performance of radiographers in image interpretation of magnetic resonance imaging (MRI) brain, spine and knee examinations following a nine-month work based postgraduate MRI module.

METHODS

Twenty-seven participants each submitted 60 image commentaries taken from prospective clinical workloads. The image interpretations (n = 1620) comprised brain, spine, and knee MRI examinations. Prevalence of abnormal examinations approximated 53% (brain), 74% (spine), and 73% (knee), and included acute and chronic pathology, normal variants and incidental findings. Each image interpretation was graded against reference standard consultant radiologist definitive report.

RESULTS

The radiographer's performance on brain image interpretations demonstrated mean accuracy at 86.7% (95% CI 83.4-89.3) with sensitivity and specificity of 84% (95% CI 80.9-86.4) and 89.7% (95% CI 86.2-92.6) respectively. For spinal interpretations the mean accuracy was 86.4% (95% CI 83.4-89.0), sensitivity was 90.2% (95% CI 88.2-92), mean specificity was 75.3% (95% CI 69.4-80.4). The mean results for knee interpretation accuracy were 80.9% (95% CI 77.3-84.1), sensitivity was 83.3% (95% CI 80.8-85.5), with 74.3% specificity (95% CI 67.4-80.4).

CONCLUSIONS

The radiographer's demonstrated skills in brain, spine and knee MRI examination image interpretation. These skills are not to replace radiologist reporting but to meet regulating body standards of proficiency, and to assist decision making in communicating unexpected serious findings, and/or extend scan range and sequences. Further research is required to investigate the impact of these skills on adjusting scan protocols or flagging urgent findings in clinical practice.

256 Shades of gray: uncertainty and diagnostic error in radiology

Radiologists practice in an environment of extraordinarily high uncertainty, which results partly from the high variability of the physical and technical aspects of imaging, partly from the inherent limitations in the diagnostic power of the various imaging modalities, and partly from the complex visual-perceptual and cognitive processes involved in image interpretation. This paper reviews the high level of uncertainty inherent to the process of radiological imaging and image interpretation vis-à-vis the issue of radiological interpretive error, in order to highlight the considerable degree of overlap that exists between these. The scope of radiological error, its many potential causes and various error-reduction strategies in radiology are also reviewed.

Deep Convolutional Neural Networks for breast cancer screening

BACKGROUND AND OBJECTIVE

Radiologists often have a hard time classifying mammography mass lesions which leads to unnecessary breast biopsies to remove suspicions and this ends up adding exorbitant expenses to an already burdened patient and health care system.

METHODS

In this paper we developed a Computer-aided Diagnosis (CAD) system based on deep Convolutional Neural Networks (CNN) that aims to help the radiologist classify mammography mass lesions. Deep learning usually requires large datasets to train networks of a certain depth from scratch. Transfer learning is an effective method to deal with relatively small datasets as in the case of medical images, although it can be tricky as we can easily start overfitting.

RESULTS

In this work, we explore the importance of transfer learning and we experimentally determine the best fine-tuning strategy to adopt when training a CNN model. We were able to successfully fine-tune some of the recent, most powerful CNNs and achieved better results compared to other state-of-the-art methods which classified the same public datasets. For instance we achieved 97.35% accuracy and 0.98 AUC on the DDSM database, 95.50% accuracy and 0.97 AUC on the INbreast database and 96.67% accuracy and 0.96 AUC on the BCDR database. Furthermore, after pre-processing and normalizing all the extracted Regions of Interest (ROIs) from the full mammograms, we merged all the datasets to build one large set of images and used it to fine-tune our CNNs. The CNN model which achieved the best results, a 98.94% accuracy, was used as a baseline to build the Breast Cancer Screening Framework. To evaluate the proposed CAD system and its efficiency to classify new images, we tested it on an independent database (MIAS) and got 98.23% accuracy and 0.99 AUC.

CONCLUSION

The results obtained demonstrate that the proposed framework is performant and can indeed be used to predict if the mass lesions are benign or malignant.

Errors in Mammography Cannot be Solved Through Technology Alone

Mammography has been the frontline screening tool for breast cancer for decades. However, high error rates in the form of false negatives (FNs) and false positives (FPs) have persisted despite technological improvements. Radiologists still miss between 10% and 30% of cancers while 80% of woman recalled for additional views have normal outcomes, with 40% of biopsied lesions being benign. Research show that the majority of cancers missed is actually visible and looked at, but either go unnoticed or are deemed to be benign. Causal agents for these errors include human related characteristics resulting in contributory search, perception and decision-making behaviours. Technical, patient and lesion factors are also important relating to positioning, compression, patient size, breast density and presence of breast implants as well as the nature and subtype of the cancer itself, where features such as architectural distortion and triple-negative cancers remain challenging to detect on screening. A better understanding of these causal agents as well as the adoption of technological and educational interventions, which audits reader performance and provide immediate perceptual feedback, should help. This paper reviews the current status of our knowledge around error rates in mammography and explores the factors impacting it. It also presents potential solutions for maximizing diagnostic efficacy thus benefiting the millions of women who undergo this procedure each year.

Radiology education: a radiology curriculum for all medical students?

Diagnostic errors in radiology are frequent and can cause severe patient harm. Despite large performance differences between radiologists and non-radiology physicians, the latter often interpret medical images because electronic health records make images available throughout the hospital. Some people argue that non-radiologists should not diagnose medical images at all, and that medical school should focus on teaching ordering skills instead of image interpretation skills. We agree that teaching ordering skills is crucial as most physicians will need to order medical images in their professional life. However, we argue that the availability of medical images is so ubiquitous that it is important that non-radiologists are also trained in the basics of medical image interpretation and, additionally in recognizing when radiological consultancy should be sought. In acute situations, basic image interpretations skills can be life-saving. We plead for a radiology curriculum for all medical students. This should include the interpretation of common abnormalities on chest and skeletal radiographs and a basic distinction of normal from abnormal images. Furthermore, substantial attention should be given to the correct ordering of radiological images. Finally, it is critical that students are trained in deciding when to consult a radiologist.

Diagnostic errors in abdominopelvic CT interpretation: characterization based on report addenda

PURPOSE

The purpose of the article is to characterize the diagnostic errors in abdominopelvic CT interpretation through review of radiology report addenda.

METHODS

We searched abdominopelvic CT reports for the word "addendum" over a nearly seven-year period. Addenda were reviewed to identify those reporting a diagnostic error. Cases were characterized by a spectrum of features.

RESULTS

709 addenda describing 785 diagnostic errors were identified, representing approximately 0.5% of searched reports. 84.1% were a new finding, 5.1% an upgrade in severity of an originally reported finding, 3.9% a downgrade in severity, and 6.9% other modification. The most common anatomic sites, as well as the most common missed abnormality per site, were vasculature (9.8%, atherosclerosis/thrombus), abdominal wall (8.3%, ventral hernia), bone [7.4%, osseous lesion (not clearly benign)], kidney [6.9%, renal lesion (not clearly benign)], liver (6.1%, steatosis), and ureter (5.1%, calculus). Of 209 addenda providing a reason for the change, 30.6% related to comparison with prior imaging, 22.5% additional surgical history, 13.4% referrer request for re-review, 8.6% additional signs, symptoms, or lab abnormality, 8.6% additional known diagnosis, 5.7% attention to patient gender, 5.3% multi-planar reconstructions, and 5.3% consultation with other radiologist.

CONCLUSION

Missed findings rather than misinterpretations of detected abnormalities were the most common reason for abdominopelvic CT report addenda. Awareness of the most common misses by anatomic location may help guide quality assurance initiatives. A wide variety of contributing factors were identified. Informatics and workflow optimization may be warranted to facilitate radiologists' access to all available patient-related data, as well as communication with other physicians, and thereby help reduce diagnostic errors.

An automated confirmatory system for analysis of mammograms

This paper presents an integrated system for the automatic analysis of mammograms to assist radiologists in confirming their diagnosis in mammography screening. The proposed automated confirmatory system (ACS) can process a digitalized mammogram online, and generates a high quality filtered segmentation of an image for biological interpretation and a texture-feature based diagnosis. We use a serial of image pre-processing and segmentation techniques, including 2D median filtering, seeded region growing (SRG) algorithm, image contrast enhancement, to remove noise, delete radiopaque artifacts and eliminate the projection of the pectoral muscle from a digitalized mammogram. We also develop an entire-image texture-feature based classification method, by combining a Rough-set approach to extract five fundamental texture features from images, and then an Artificial Neural Network technique to classify a mammogram as: normal; indicating the presence of a benign lump; or representing a malignant tumor. Here, 222 random images from the Mammographic Image Analysis Society (MIAS) database are used for the offline ACS training. Once the system is tuned and trained, it is ready for the automated use for the analysis and diagnosis of new mammograms. To test the trained system, a separate set of 100 random images from the MIAS and another set of 100 random images from the independent BancoWeb database are selected. The proposed ACS is shown to be successful in confirming diagnosis of mammograms from the two independent databases.