Radiology's Achilles' heel: error and variation in the interpretation of the Röntgen image

Clinical application of radiomics for the prediction of treatment outcome and survival in patients with renal cell carcinoma: a systematic review

PURPOSE

The management of renal cell carcinoma (RCC) relies on clinical and histopathological features for treatment decisions. Recently, radiomics, which involves the extraction and analysis of quantitative imaging features, has shown promise in improving RCC management. This review evaluates the current application and limitations of radiomics for predicting treatment and oncological outcomes in RCC.

METHODS

A systematic search was conducted in Medline, EMBASE, and Web of Science databases or studies that used radiomics to predict response to treatment and survival outcomes in patients with RCC. The study quality was assessed using the Radiomics Quality Score (RQS) tools.

RESULTS

The systematic review identified a total of 27 studies, examining 6,119 patients. The most used imaging modality was contrast-enhanced abdominal CT. The reviewed studies extracted between 19 and 3376 radiomics features, including Histogram, Texture, Filter, or transformation method. Radiomics-based risk stratification models provided valuable insights into treatment response and oncological outcomes. All developed signatures demonstrated at least modest accuracy (AUC range: 0.55-0.99). The studies included in this analysis reported heterogeneous results regarding radiomics methods. The range of Radiomics Quality Score (RQS) was from - 5 to 20, with a mean RQS total of 9.15 ± 7.95.

CONCLUSION

Radiomics has emerged as a promising tool in the management of RCC. It offers the potential for improved risk stratification and response assessment. However, future trials must demonstrate the generalizability of findings to prospective cohorts before progressing towards clinical translation.

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.

Visualization of anatomical structures in the fetlock region of the horse using cone beam computed tomography in comparison with conventional multidetector computed tomography

Introduction

Cone-beam computed tomography (CBCT) is regarded as a convenient and suitable alternative to conventional computed tomography. However, in the horse, the quality of obtained data sets needs to be evaluated. Therefore, the aim of this study was to compare the visibility and accessibility of clinically relevant anatomical structures displayed in CBCT and conventional multidetector computed tomography (MDCT).

Materials and methods

Twenty-nine limbs from horses euthanized for reasons unrelated to this study were used. Native and intraarticular contrast scans of the fetlock (CBCT vs. MDCT) were performed. The visibility and accessibility of selected anatomical structures were blindly scored by three independent experienced observers using a scoring system previously reported and adapted to the fetlock joint.

Results

Only minor differences between CBCT and MDCT were identified concerning the diagnostic quality of images for osseous structures. Soft tissue structures were better evaluated on MDCT images. In CBCT as well as in MDCT articular cartilage could only be visualized after intraarticular injection of contrast medium.

Discussion/conclusion

Cone beam computed tomography of the fetlock is a useful and reliable diagnostic tool when evaluating osseous structures and delineating articular cartilage with contrast medium. However, this modality is limited for assessing soft tissues structures.

Documentation of polysomnographic and home sleep apnea test interpretations: room for improvement?

STUDY OBJECTIVES

Obstructive sleep apnea (OSA), a heterogeneous disorder with many different presentations, is diagnosed with sleep studies. In standard clinical practice, test data are reviewed and scored, and interpretations are documented. Little standardization exists regarding what should be included in interpretations. We aimed to determine how consistently the documented interpretation included references to study quality parameters and accepted disease phenotypes.

METHODS

This study was performed at a single academic center in January 2021. From the literature, we formulated a list of test and titration quality criteria and OSA phenotypes that should be reflected in study interpretations, including total recording time, total sleep time, positionality, and supine rapid eye movement (REM) sleep during titration. We retrospectively reviewed the documentation of sleep studies to determine how often these factors were reflected in interpretation reports or clinical notes.

RESULTS

Of 134 patients in the study, 81 were diagnosed with OSA. A finding of inadequate total recording time during polysomnography or total sleep time on home sleep apnea testing was most often not documented. Positionality of OSA was not documented in 33% of applicable studies. The absence of supine REM sleep during positive airway pressure titration was not mentioned in 15% of interpretations.

CONCLUSIONS

The documentation of quality concerns and clinically important OSA phenotypes in interpretations was inconsistent. Documentation of meaningful test quality information and sleep apnea phenotypes might be improved with report standardization or by developing enhanced data displays.

CITATION

Herberts MB, Morgenthaler TI. Documentation of polysomnographic and home sleep apnea test interpretations: room for improvement? J Clin Sleep Med. 2023;19(6):1043-1049.

Applying Implementation Science Principles to Systematize High-Quality Care for Potentially Significant Imaging Findings

OBJECTIVE

Use principles of implementation science to improve the diagnosis and management of potentially significant imaging findings.

METHODS

Multidisciplinary stakeholders codified the diagnosis and management of potentially significant imaging findings in eight organs and created a finding tracking management system that was embedded in radiologist workflows and IT systems. Radiologists were trained to use this system. An automated finding tracking management system was created to support consistent high-quality care through care pathway visualizations, increased awareness of specific findings in the electronic medical record, templated notifications, and creation of an electronic safety net. Primary outcome was the rate of quality reviews related to eight targeted imaging findings. Secondary outcome was radiologist use of the finding tracking management tool.

RESULTS

In the 4 years after implementation, the tool was used to track findings in 7,843 patients who received 10,015 ultrasound, CT, MRI, x-ray, and nuclear medicine examinations that were interpreted by all 34 radiologists. Use of the tool lead to a decrease in related quality reviews (from 8.0% to 0.0%, P < .007). Use of the system increased from 1.7% of examinations in the early implementation phase to 3.1% (+82%, P < .00001) in the postimplementation phase. Each radiologist used the tool on an average of 294.6 unique examinations (SD 404.8). Overall, radiologists currently use the tool approximately 4,000 times per year.

DISCUSSION

Radiologists frequently used a finding tracking management system to ensure effective communication and raise awareness of the importance of recommended future follow-up studies. Use of this system was associated with a decrease in the rate of quality review requests in this domain.

Bennett's Fracture Management: A Systematic Review of Literature

Bennett's fracture is a relatively common fracture of the base of the thumb with no consensus on optimum management. Determining the optimal treatment method for Bennett's fractures remain a challenge and has been the subject of much debate. This systematic review aims to investigate present and past literature and determine the optimum treatment intervention for Bennett's fracture-dislocation. The primary outcome measure is post-traumatic arthritis, and the secondary outcome measures are reoperation, pain, infection, and nonunion. As per Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, a systematic review of the literature was performed to evaluate patient demographics, clinical profile, management, and treatment outcomes. Two authors independently performed electronic searches of the Embase, Medical Literature Analysis and Retrieval System Online (MEDLINE), and Cochrane databases. Studies conducted between 1963 and 2021 with articles reporting Bennett's fracture management were included. The study was registered with PROSPERO (CRD42021295464). In the initial screening, 58 articles were identified, of which 13 articles met the criteria and were included in the final review, evaluating 558 patients. Out of these, 439 were managed by various surgical procedures, and 119 were treated by manipulation and plaster of Paris (POP) immobilization. Of the 13 studies considered, eight have a clear mention of post-traumatic osteoarthritis, with a total of 50 (9%) patients. Secondary outcomes included pain in 76 (13%) patients, infection in four patients, reoperation in 11 (2%) patients, and no nonunion. This review was conducted with the help of retrospective studies as there is no randomized controlled trial on the management of Bennett's fracture. Our primary outcome measure of post-traumatic arthritis in patients being treated by operative and conservative methods was mentioned in these studies. However, due to the sample size being small and the heterogenicity of these studies, the strength of these findings is low. Due to these shortfalls, this review study cannot recommend any single (or) particular treatment for all patients with Bennett's fracture.

Repeatability of radiographic assessments for feline hip dysplasia suggest consensus scores in radiology are more uncertain than commonly assumed

Variation in the diagnostic interpretation of radiographs is a well-recognised problem in human and veterinary medicine. One common solution is to create a 'consensus' score based on a majority or unanimous decision from multiple observers. While consensus approaches are generally assumed to improve diagnostic repeatability, the extent to which consensus scores are themselves repeatable has rarely been examined. Here we use repeated assessments by three radiologists of 196 hip radiographs from 98 cats within a health-screening programme to examine intra-observer, inter-observer, majority-consensus and unanimous-consensus repeatability scores for feline hip dysplasia. In line with other studies, intra-observer and inter-observer repeatability was moderate (63-71%), and related to the reference assessment and time taken to reach a decision. Consensus scores did show reduced variation between assessments compared to individuals, but consensus repeatability was far from perfect. Only 75% of majority consensus scores were in agreement between assessments, and based on Bayesian multinomial modelling we estimate that unanimous consensus scores can have repeatabilities as low as 83%. These results clearly show that consensus scores in radiology can have large uncertainties, and that future studies in both human and veterinary medicine need to include consensus-uncertainty estimates if we are to properly interpret radiological diagnoses and the extent to which consensus scores improve diagnostic accuracy.

Reliability and accuracy in radiographic measurements of musculoskeletal tumors

Tumor size and growth are important parameters when evaluating bone and soft tissue neoplasms. There are no reports comparing the intra- and interobserver reliability among physicians in their evaluation of musculoskeletal (MSK) tumor imaging. This study investigates the accuracy and precision of measurements made by orthopedic and radiology physicians in different stages of training. Blinded magnetic resonance imaging (MRI) scans from six patients, three soft tissue, and three bone tumors were selected: each case included an "old" and "new" scan that was performed at least 3 months apart. Fourteen participants were selected, representing varying levels of education and experience, including two of each of the following: medical students, orthopedic and radiology residents, oncology and nononcologic orthopedic attendings, and MSK and non-MSK radiology attendings. Participants compared the old and new studies, recording tumor size in the transverse, cranial-caudal, and anterior-posterior dimensions, and determined if the tumor was stable or unstable. The MRI's official report served as the "gold standard." Average intraobserver variability (|Trial 1 - Trial 2|/[(Trial 1 + Trial 2)/2])) in size measurements was 11.08% (0.00%-68.62%). The lowest variability was recorded by the MSK radiologist 1 (6.16%), and the greatest variability by Orthopedic Surgery Resident 1 (16.70%). Participants correctly determined stability 82% of the time (71%-100%). Only MSK radiologists correctly determined stability in over 90% of cases. There is considerable variability and inaccuracy in MRI-based measurements of MSK tumors. These findings motivate opportunities for improving MSK imaging education of radiology and orthopedic residents. Physicians ordering MRI scans should evaluate them themselves, instead of relying on the radiology report alone, to inform clinical decision-making.

Missed cervical spine injuries: aim for the top

PURPOSE

To determine the incidence of missed cervical spine injuries by radiology registrars in a major trauma centre and to identify any common blind spots.

MATERIALS AND METHODS

All patients with an acute traumatic injury who underwent a CT scan of the cervical spine in our unit, which serves a population of approximately 900,000, between September 2016 and December 2017 and whom had a separate radiology trainee report and final neuroradiology consultant report available were included in the study. We recorded the date and time of the scan, the registrar error and the registrar grade. An error was defined as follows: (1) a missed fracture; (2) a missed ligamentous injury; (3) overcall of a fracture (e.g. degenerative calcification or nutrient vessel). Groups were compared with the chi-square test.

RESULTS

Five hundred seventy-three CT scans of the cervical spine fitted the inclusion criteria and were analysed. There were a total of 149 injuries over eight levels in 96 patients. There were 12 registrar errors (2.1% discrepancy rate), of which 11 were missed acute injuries (9 fractures and 2 disco-ligamentous injuries). The grade of the registrar was not significant (p = 0.603). Seventy-three percent (8/11) missed injuries were disproportionately at the cranio-cervical junction, where only 11.6% of traumatic cervical spine injuries occur p < 0.0001. Forty-five percent of the missed injuries included occipital condyle fractures, which occurred in only 12/149 injuries (8%).

CONCLUSIONS

Radiology registrars safely report emergency CT scans of the cervical spine performed following trauma with a low discrepancy rate. Missed cervical spine injuries commonly occur at the cranio-cervical junction, which should become a standard review area.

Agreement in the reporting of General Practitioner requested musculoskeletal radiographs: Reporting radiographers and consultant radiologists compared with an index radiologist

INTRODUCTION

This study assessed the inter-observer agreement of reporting radiographers and consultant radiologists compared with an index radiologist when reporting General Practitioner (GP) requested musculoskeletal radiographs. The potential effect of discordant reports on patient management and outcome was also examined.

METHODS

Three reporting radiographers, three consultant radiologists and an index radiologist reported on a retrospective randomised sample of 219 GP requested musculoskeletal radiographs, in conditions simulating clinical practice. A speciality doctor in radiology compared the observers' reports with the index radiologist report for agreement and assessed whether any discordance between reports was clinically important.

RESULTS

Overall agreement with the index radiologist was 47.0% (95% CI, 40.5-53.6) and 51.6% (95% CI, 45.0-58.1) for the consultant radiologists and reporting radiographers, respectively. The results for the appendicular and axial skeleton were 48.6% (95% CI, 41.3-55.9) and 40.9% (95% CI, 27.7-55.6) for the radiologists, and 52.6% (95% CI, 45.2-59.8) and 47.7% (95% CI, 33.8-62.1) for the radiographers, respectively. The difference in overall observer agreement between the two professional groups with the index radiologist was not statistically significant (p = 0.34). Discordance with the index radiologist's reports was judged to be clinically important in less than 10% of the observer's reports.

CONCLUSION

Reporting radiographers and consultant radiologists demonstrate similar levels of concordance with an index radiologist when reporting GP requested musculoskeletal radiographs.

IMPLICATIONS FOR PRACTICE

These findings contribute to the wider evidence base that selected radiographers with appropriate postgraduate education and training are proficient to report on musculoskeletal radiographs, irrespective of referral source.

Improving reference standards for validation of AI-based radiography

Objective:

Demonstrate the importance of combining multiple readers' opinions, in a context-aware manner, when establishing the reference standard for validation of artificial intelligence (AI) applications for, e.g. chest radiographs. By comparing individual readers, majority vote of a panel, and panel-based discussion, we identify methods which maximize interobserver agreement and label reproducibility.

Methods:

1100 frontal chest radiographs were evaluated for 6 findings: airspace opacity, cardiomegaly, pulmonary edema, fracture, nodules, and pneumothorax. Each image was reviewed by six radiologists, first individually and then via asynchronous adjudication (web-based discussion) in two panels of three readers to resolve disagreements within each panel. We quantified the reproducibility of each method by measuring interreader agreement.

Results:

Panel-based majority vote improved agreement relative to individual readers for all findings. Most disagreements were resolved with two rounds of adjudication, which further improved reproducibility for some findings, particularly reducing misses. Improvements varied across finding categories, with adjudication improving agreement for cardiomegaly, fractures, and pneumothorax.

Conclusion:

The likelihood of interreader agreement, even within panels of US board-certified radiologists, must be considered before reads can be used as a reference standard for validation of proposed AI tools. Agreement and, by extension, reproducibility can be improved by applying majority vote, maximum sensitivity, or asynchronous adjudication for different findings, which supports the development of higher quality clinical research.

Advances in knowledge:

A panel of three experts is a common technique for establishing reference standards when ground truth is not available for use in AI validation. The manner in which differing opinions are resolved is shown to be important, and has not been previously explored.

The effects of examiner fatigue on the diagnostic accuracy of dental radiographs

OBJECTIVES

The aim of the study was to find out whether and to what extent the performance of dentists regarding diagnostic evaluation of dental radiographs is influenced by symptoms of fatigue.

MATERIALS AND METHODS

Over a period of 40 minutes, 21 dentists evaluated a database of 96 randomly selected, infinitely repeating intraoral dental radiographs for the presence of periapical radiolucencies. Both before and after, participants were asked to assess their subjective fatigue using the Swedish Occupational Fatigue Inventory (SOFI), Visual Analogue Scale (VAS), and Numerical Rating Scale (NRS). Diagnostic accuracy was analyzed using the Receiver Operating Characteristics (ROC) method. Furthermore, the correlation between diagnostic accuracy and radiographic experience, image viewing time, and level of training was also evaluated.

RESULTS

The study showed that despite increasing fatigue, the diagnostic accuracy of the examiners remained consistent with an average AUC value of 0.768 ± 0.091. Within the 40-min reporting period, no statistically significant fluctuations were found. The diagnostic accuracy varied depending on the radiographic experience: with many years of radiographic experience, the diagnostic accuracy increased. At the same time, the older study participants with greater radiographic experience became less tired compared to younger study participants during the examination.

CONCLUSIONS

Although an increase in fatigue was observed during the 40-min examination, the diagnostic accuracy of the doctors remained constant.

CLINICAL RELEVANCE

Due to the high workload which needs to be handled in a limited time, medical and dental professionals have reached a certain level of exposition to stress that can lead to physical fatigue. However, the increasing fatigue should not negatively influence the work of the doctors. The study shows that the radiodiagnostic accuracy remained the same.

The Utility of Adding a Liquid-Nutrient Meal to Aid Interpretation of Small-Bowel Transit Scintigraphy

Small-bowel transit scintigraphy (SBTS) evaluates the accumulation of a radiolabeled meal in the terminal ileal reservoir (TIR) 6 h after ingestion. The location of the TIR may be difficult to determine because anatomic information is limited; for equivocal studies, the patient is asked to return the next day to help determine the TIR location by potential transit into the colon. The purpose of this study was to evaluate whether administration of an additional liquid-nutrient meal (LNM) at 6 h can promote movement of the radiolabeled meal to aid in the interpretation of SBTS and eliminate the need for the patient to return. Methods: This retrospective study reviewed 117 SBTS studies from February 2017 to September 2019. Patients were fed a standardized mixed radiolabeled solid-liquid meal for gastric emptying with SBTS according to Society of Nuclear Medicine and Molecular Imaging practice guidelines. An additional LNM was given at 6 h, and post-LNM images were obtained at least 20 min after the LNM. Two board-certified nuclear medicine physicians independently evaluated all images as equivocal or diagnostic at 6 h. Results: Of the 117 patients (71.8% female; median age, 42.0 y) undergoing SBTS, 37 were equivocal cases at 6 h before the LNM (31.6%; 95% CI, 23.3%-40.9%), compared with 12 equivocal cases after the LNM (10.3%; 95% CI, 5.4%-17.2%). Of the equivocal cases, 25 (69.4%; 95% CI, 51.9%-83.7%) had a definitive result after the LNM, whereas 11 (30.6%; 95% CI, 16.4%-48.1%) remained equivocal and 1 showed rapid transit. Among the 23 patients with gastroparesis, only 13 (57%) responded to the LNM, and none of the 3 patients with irritable bowel syndrome responded. Conclusion: The number of equivocal SBTS cases decreased after administration of an LNM at 6 h, converting to a definitive result. This suggests that with use of an LNM, most patients can complete SBTS in 1 d without the need for repeat imaging at 24 h. Administering an LNM appears to be less effective for patients with gastric disorders. However, the clinical significance remains to be explored, and it is unclear whether such patients have both a gastric and a small-bowel disorder, hence reducing any motility-promoting effect of the LNM.

Deep-Learning-Based Computer-Aided Systems for Breast Cancer Imaging: A Critical Review

This paper provides a critical review of the literature on deep learning applications in breast tumor diagnosis using ultrasound and mammography images. It also summarizes recent advances in computer-aided diagnosis/detection (CAD) systems, which make use of new deep learning methods to automatically recognize breast images and improve the accuracy of diagnoses made by radiologists. This review is based upon published literature in the past decade (January 2010–January 2020), where we obtained around 250 research articles, and after an eligibility process, 59 articles were presented in more detail. The main findings in the classification process revealed that new DL-CAD methods are useful and effective screening tools for breast cancer, thus reducing the need for manual feature extraction. The breast tumor research community can utilize this survey as a basis for their current and future studies.

Correlation of objective image quality and working length measurements in different CBCT machines: An ex vivo study

To investigate potential correlations between objective CBCT image parameters and accuracy in endodontic working length determination ex vivo. Contrast-to-noise ratio (CNR) and spatial resolution (SR) as fundamental objective image parameters were examined using specific phantoms in seven different CBCT machines. Seven experienced observers were instructed and calibrated. The order of the CBCTs was randomized for each observer and observation. To assess intra-operator reproducibility, the procedure was repeated within six weeks with a randomized order of CBCT images. Multivariate analysis (MANOVA) did not reveal any influence of the combined image quality factors CNR and SR on measurement accuracy. Inter-operator reproducibility as assessed between the two observations was poor, with a mean intra-class correlation (ICC) of 0.48 (95%-CI  0.38, 0.59) for observation No. 1. and 0.40 (95%-CI 0.30, 0.51) for observation No. 2. Intra-operator reproducibility pooled over all observers between both observations was only moderate, with a mean ICC of 0.58 (95%-CI 0.52 to 0.64). Within the limitations of the study, objective image quality measures and exposure parameters seem not to have a significant influence on accuracy in determining endodontic root canal lengths in CBCT scans. The main factor of variance is the observer.

Recombinant Human BMP6 Applied Within Autologous Blood Coagulum Accelerates Bone Healing: Randomized Controlled Trial in High Tibial Osteotomy Patients

Bone morphogenetic proteins (BMPs) are potent osteogenic proteins that induce new bone formation in vivo. However, their effect on bone healing in the trabecular bone surfaces remains challenging. We evaluated the safety and efficacy of recombinant human BMP6 (rhBMP6) applied within an autologous blood coagulum (ABC) in a surgically created wedge defect of the proximal tibia in patients undergoing high tibial osteotomy (HTO) for varus deformity and medial osteoarthritis of the knee. We enrolled 20 HTO patients in a randomized, placebo-controlled, double-blinded phase I/II clinical trial. RhBMP6/ABC (1.0 mg/10 mL ABC prepared from peripheral blood) or placebo (10 mL ABC containing excipients) was administered into the tibial wedge defects. Patients were followed for 0 to 24 months by clinical examination (safety) and computed tomography (CT) and serial radiographic analyses (efficacy). The results show that there were no detectable anti-rhBMP6 antibodies in the blood of any of the 20 patients at 14 weeks after implantation. During the 24 months of follow-up, there were no serious adverse reactions recorded. The CT scans from defects of patients treated with rhBMP6/ABC showed an accelerated bone healing compared with placebo at 9 weeks (47.8 ± 24.1 versus 22.2 ± 12.3 mg/cm³ ; p = 0.008) and at 14 weeks (89.7 ± 29.1 versus 53.6 ± 21.9 mg/cm³ ; p = 0.006) follow-up. Radiographic analyses at weeks 6 and 24 and months 12 and 24 suggested the advanced bone formation and remodeling in rhBMP6/ABC-treated patients. In conclusion, we show that rhBMP6/ABC at a dose of 100 μg/mL accelerated bone healing in patients undergoing HTO without serious adverse events and with a good tolerability compared with placebo alone. Overall, for the first time, a BMP-based osteogenic implant was examined against a placebo for bone healing efficacy in the trabecular bone surface, using an objective bone mineral density measurement system. © 2020 The Authors. Journal of Bone and Mineral Research published by American Society for Bone and Mineral Research.

Inter- and intraobserver agreement in standard and ultra-fast single-photon emission computed tomography/computed tomography for the assessment of bone metastases

OBJECTIVE

To investigate inter- and intraobserver agreement using standard (std)-single-photon emission computed tomography/computed tomography (SPECT/CT) or ultra-fast acquisition (UF)-SPECT/CT as an add-on to planar whole-body bone scintigraphy (WB-BS) for the detection of bone metastases in patients with known cancer.

METHODS

Images from patients (n = 104) participating in a prospective trial comparing SPECT/CT as an add-on to WB-BS with two different acquisition methods (std-SPECT/CT; 20 s per view, 32 views and UF-SPECT/CT; 16 views of 10 s per view) were included. The combinations of WB-BS with the two different SPECT/CT acquisition methods were independently evaluated by two experienced observers using a three-point scale (M0: no metastases, Me: equivocal for bone metastases and M1: bone metastases). The observers re-evaluated the scans 6 months later. Inter- and intraobserver agreement was calculated using linear weighted kappa.

RESULTS

The linear weighted kappa for the interobserver agreement was 0.78 [95% confidence interval (CI) 0.61-0.94] for WB-BS with std-SPECT/CT and 0.84 (95% CI 0.67-1.00) for WB-BS with UF-SPECT/CT. Similar values were observed for the intraobserver agreement, in which the linear weighted kappa was 0.76 (95% CI 0.60-0.92) and 0.73 (95% CI 0.57-0.90) for std-SPECT/CT and UF-SPECT/CT, respectively.

CONCLUSION

Satisfactory inter- and intraobserver agreement was seen for both acquisition methods, emphasizing that the use of SPECT/CT as an add-on to WB-BS is a robust method for the detection of bone metastases even when conducted with a very short acquisition time.

Errors and artifacts on radiographs

Introduction: The process of recording a patient includes a procedure with several separate segments during work that together provide the imaging to be obtained for adequate radiological analysis. Throughout the process, it is possible to experience errors that create artifacts on X-rays which ultimately results in an inadequate recording that is not for valid analysis. Aim: Determine the total number of radiological films that are not for valid analysis. Sort out and analyze errors in radiographs according to the work process. Provide recommendations for improving the quality in the process of recording the patient. Material and methods: A prospective study was conducted at the Radiology Clinic of the Clinical Hospital Center Pristina-Gracanica, for two calendar years. All films that are not for valid analysis were considered. The radiological procedure of patient imaging was broken down into logical segments so that possible errors could be observed. We have summarized the causes of the artifacts in five appropriate groups (errors made by the recording technique, during the acquisition of the image, caused by the object of recording, during the processing of films in an automated machine and improper handling of films). Results: The total amount of used X-ray films is 32600 pieces, of which 242 (0.74%) were errors and artifacts. The most common format of a film with an error or artifact was 30x40 cm. A frequency of errors according to the cause of the occurrence is classified into appropriate groups. The largest number was in a group 1 - 155 (64.04%), in a group 2 - 3 (1.24%), in a group 3 - 13 (5.37%), in a group 4 - 67 (27.69%), and in a group 5 - 4 (1.66%). Conclusion: In the proper systematization of all observed errors and artifacts of X-ray film, it allows us to realise the place of error during the whole process of recording and processing of the film. We hereby wish to propose their elimination and improve the quality of the radiology department.

Observer Agreement and Accuracy of 18F-Sodium Fluoride PET/CT in the Diagnosis of Bone Metastases in Prostate Cancer

Our aim was to evaluate the interobserver agreement in ¹⁸F-sodium fluoride (NaF) PET/CT for the detection of bone metastases in patients with prostate cancer (PCa). Methods: ¹⁸F-NaF PET/CT scans were retrieved from all patients who participated in 4 recent prospective trials. Two experienced observers independently evaluated the ¹⁸F-NaF PET/CT scans on a patient level using a 3-category scale (no bone metastases [M0], equivocal for bone metastases, and bone metastases present [M1]) and on a dichotomous scale (M0/M1). In patients with no more than 10 lesions, the location and number of lesions were recorded. On a patient level, the diagnostic performance was calculated using a sensitivity analysis, in which equivocal lesions were handled as M0 as well as M1. Results: ¹⁸F-NaF PET/CT scans from 219 patients with PCa were included, of whom 129 patients were scanned for primary staging, 67 for biochemical recurrence, and 23 for metastatic castration-resistant PCa. Agreement between the observers was almost perfect on a patient level (3-category unweighted κ = 0.83 ± 0.05, linear weighted κ = 0.90 ± 0.06, and dichotomous κ = 0.91 ± 0.07). On a lesion level (dichotomous scale), the observers agreed on the number and location of bone metastases in 205 (93.6%) patients. In the remaining 14 patients, the readers disagreed on the number of lesions in 13 patients and the location of bone metastases in 1 patient. A final diagnosis of bone metastases was made for 211 of 219 patients. The sensitivity ranged from 0.86 to 0.92, specificity from 0.83 to 0.97, positive predictive value from 0.70 to 0.93, and negative predictive value from 0.94 to 0.96. Conclusion: The interobserver agreement on ¹⁸F-NaF PET/CT for the detection of bone metastases in patients with PCa was very high among trained observers, both on a patient level and on a lesion level. Moreover, the diagnostic performance of ¹⁸F-NaF PET/CT was satisfactory, rendering ¹⁸F-NaF PET/CT a robust tool in the diagnostic armamentarium.

Recent Updates on Molecular Imaging Reporting and Data Systems (MI-RADS) for Theranostic Radiotracers-Navigating Pitfalls of SSTR- and PSMA-Targeted PET/CT

The theranostic concept represents a paradigmatic example of personalized treatment. It is based on the use of radiolabeled compounds which can be applied for both diagnostic molecular imaging and subsequent treatment, using different radionuclides for labelling. Clinically relevant examples include somatostatin receptor (SSTR)-targeted imaging and therapy for the treatment of neuroendocrine tumors (NET), as well as prostate-specific membrane antigen (PSMA)-targeted imaging and therapy for the treatment of prostate cancer (PC). As such, both classes of radiotracers can be used to triage patients for theranostic endoradiotherapy using positron emission tomography (PET). While interpreting PSMA- or SSTR-targeted PET/computed tomography scans, the reader has to navigate certain pitfalls, including (I.) varying normal biodistribution between different PSMA- and SSTR-targeting PET radiotracers, (II.) varying radiotracer uptake in numerous kinds of both benign and malignant lesions, and (III.) resulting false-positive and false-negative findings. Thus, two novel reporting and data system (RADS) classifications for PSMA- and SSTR-targeted PET imaging (PSMA- and SSTR-RADS) have been recently introduced under the umbrella term molecular imaging reporting and data systems (MI-RADS). Notably, PSMA- and SSTR-RADS are structured in a reciprocal fashion, i.e., if the reader is familiar with one system, the other system can readily be applied. Learning objectives of the present case-based review are as follows: (I.) the theranostic concept for the treatment of NET and PC will be briefly introduced, (II.) the most common pitfalls on PSMA- and SSTR-targeted PET/CT will be identified, (III.) the novel framework system for theranostic radiotracers (MI-RADS) will be explained, applied to complex clinical cases and recent studies in the field will be highlighted. Finally, current treatment strategies based on MI-RADS will be proposed, which will demonstrate how such a generalizable framework system truly paves the way for clinically meaningful molecular imaging-guided treatment of either PC or NET. Thus, beyond an introduction of MI-RADS, the present review aims to provide an update of recently published studies which have further validated the concept of structured reporting systems in the field of theranostics.

Computer-Aided Detection for Breast Cancer Screening in Clinical Settings: Scoping Review

Background

With the growth of machine learning applications, the practice of medicine is evolving. Computer-aided detection (CAD) is a software technology that has become widespread in radiology practices, particularly in breast cancer screening for improving detection rates at earlier stages. Many studies have investigated the diagnostic accuracy of CAD, but its implementation in clinical settings has been largely overlooked.

Objective

The aim of this scoping review was to summarize recent literature on the adoption and implementation of CAD during breast cancer screening by radiologists and to describe barriers and facilitators for CAD use.

Methods

The MEDLINE database was searched for English, peer-reviewed articles that described CAD implementation, including barriers or facilitators, in breast cancer screening and were published between January 2010 and March 2018. Articles describing the diagnostic accuracy of CAD for breast cancer detection were excluded. The search returned 526 citations, which were reviewed in duplicate through abstract and full-text screening. Reference lists and cited references in the included studies were reviewed.

Results

A total of nine articles met the inclusion criteria. The included articles showed that there is a tradeoff between the facilitators and barriers for CAD use. Facilitators for CAD use were improved breast cancer detection rates, increased profitability of breast imaging, and time saved by replacing double reading. Identified barriers were less favorable perceptions of CAD compared to double reading by radiologists, an increase in recall rates of patients for further testing, increased costs, and unclear effect on patient outcomes.

Conclusions

There is a gap in the literature between CAD’s well-established diagnostic accuracy and its implementation and use by radiologists. Generally, the perceptions of radiologists have not been considered and details of implementation approaches for adoption of CAD have not been reported. The cost-effectiveness of CAD has not been well established for breast cancer screening in various populations. Further research is needed on how to best facilitate CAD in radiology practices in order to optimize patient outcomes, and the views of radiologists need to be better considered when advancing CAD use.

Reproducibility of CBCT image analysis: a clinical study on intrapersonal and interpersonal errors in bone structure determination

OBJECTIVE

For correct implant planning based on cone-beam computed tomography (CBCT), the bone contour must be accurately determined. Identification of the contour is difficult in bones with incomplete mineralization. In this clinical study, we investigated the intrapersonal and interpersonal reproducibilities of manual bone contour determination on CBCT images using a semi-automated computerized process.

METHODS

The bone surface level in the area of the socket in 20 patients who had undergone tooth extraction from the upper jaw at 10 ± 1 weeks previously was determined on CBCT images. Two investigators with different levels of experience determined the bone structure initially (T₀) and repeated the procedure after 3 months (T₁). The bone structure marked on CBCT images was converted into a surface data set. The resulting data sets were superimposed on one another. In the analyses, the shortest distances between the datasets were identified and measured. The average deviations were statistically evaluated.

RESULTS

The intrapersonal evaluation resulted in an average deviation of 0.18 mm across both investigators. The interpersonal analysis comparing the two investigators resulted in average deviations of 0.15 mm at T₀ and 0.26 mm at T₁. Significant differences were not found.

CONCLUSIONS

The low intrapersonal deviation indicates that the procedure has satisfactory reproducibility. All deviations were within the range of the selected resolution of the CBCT device. Application of a semi-automated procedure to detect the bone border in areas with incomplete mineralization is a predictable process.

TRIAL REGISTRATION

The study was registered in the German Clinical Trials Register and the International Clinical Trials Registry Platform of the WHO: DRKS00004769, date of registration: 28 February 2013; and DRKS00005978, date of registration: 09 November 2015.

Error in radiology-where are we now?

Error is inherent in radiological practice. Our awareness of the extent of this and the reasons behind it has increased in recent times. Our next step must be the development of a shared understanding with our patients of the limitations as well as the huge benefits of medical imaging.

Is an indistinct picture "exactly what we need"? Objectivity, accuracy, and harm in imaging for cancer

Assumptions about the epistemic ideal of objectivity, closely related to ontological assumptions about the nature of disease as pathophysiological abnormality, lead us into oversimplified ways of thinking about medical imaging. This is illustrated by current controversies in the early detection of cancer. Improvements in the technical quality of imaging failed to address the problem of overdiagnosis in breast cancer screening and exacerbate the problem in thyroid cancer diagnosis. Drawing on Douglas and on Daston and Galison, I distinguish 3 dimensions of objectivity (accuracy, reliability, and precision) and demonstrate ways they may be at odds, as illustrated in the early detection of cancer. Guidelines for evaluating the efficacy of diagnostic imaging are insufficiently sensitive to this complexity. Improving imaging quality may raise epistemic issues, place disease definitions in question, and lead to overall harm or to changes in the distribution of harms and benefits among population subgroups. With a nod to Wittgenstein, I argue that we cannot take for granted that "an indistinct picture" is not "exactly what we need."

Does the Use of a Checklist Help Medical Students in the Detection of Abnormalities on a Chest Radiograph?

The interpretation of chest radiographs is a complex task that is prone to diagnostic error, especially for medical students. The aim of this study is to investigate the extent to which medical students benefit from the use of a checklist regarding the detection of abnormalities on a chest radiograph. We developed a checklist based on literature and interviews with experienced thorax radiologists. Forty medical students in the clinical phase assessed 18 chest radiographs during a computer test, either with (n = 20) or without (n = 20) the checklist. We measured performance and asked participants for feedback using a survey. Participants that used a checklist detected more abnormalities on images with multiple abnormalities (M = 50.1%) than participants that could not use a checklist (M = 41.9%), p = 0.04. The post-experimental survey shows that on average, participants considered the checklist helpful (M = 3.25 on a five-point scale), but also time consuming (M = 3.30 on a five-point scale). In conclusion, a checklist can help medical students to detect abnormalities in chest radiographs. Moreover, students tend to appreciate the use of a checklist as a helpful tool during the interpretation of a chest radiograph. Therefore, a checklist is a potentially important tool to improve radiology education in the medical curriculum.

Interobserver Agreement of Interim and End-of-Treatment 18F-FDG PET/CT in Diffuse Large B-Cell Lymphoma: Impact on Clinical Practice and Trials

We aimed to assess the interobserver agreement of interim PET (I-PET) and end-of-treatment PET (EoT-PET) using the Deauville score (DS) in first-line diffuse large B-cell lymphoma (DLBCL) patients. Methods: I-PET and EoT-PET scans of DLBCL patients were performed in the HOVON84 study (2007-2012), an international multicenter randomized controlled trial. Patients received R-CHOP14 and were randomized to receive rituximab intensification in the first 4 cycles or not. I-PET was performed after 4 cycles (for observational purposes), and EoT-PET after 6 or 8 cycles. Two independent central reviewers retrospectively scored all scans according to the DS system, masked to clinical outcomes. Results were dichotomized as negative (DS of 1-3) or positive (DS of 4-5). Besides percentage overall agreement (OA), we calculated agreement for positive and negative scores, expressed as positive agreement (PA) and negative agreement (NA), respectively. Results: 465 I-PET and 457 EoT-PET scans were centrally reviewed; baseline ¹⁸F-FDG PET or PET/CT was available in 75%-77%, and CT in the remaining cases. Percentage OA for I-PET and EoT-PET were 87.7% and 91.7% (P = 0.049), with NA of 92.0% and 95.0% (P = 0.091), and PA of 73.7% and 76.3% (P = 0.656), respectively. Conclusion: Interobserver agreement using DS in DLBCL patients in I-PET and EoT-PET yields high OA and NA. The lower PA suggests that EoT-PET/CT treatment evaluation in daily practice and I-PET-adapted trials may benefit from dual reads and central review, respectively.

Medical Image Data and Datasets in the Era of Machine Learning-Whitepaper from the 2016 C-MIMI Meeting Dataset Session

At the first annual Conference on Machine Intelligence in Medical Imaging (C-MIMI), held in September 2016, a conference session on medical image data and datasets for machine learning identified multiple issues. The common theme from attendees was that everyone participating in medical image evaluation with machine learning is data starved. There is an urgent need to find better ways to collect, annotate, and reuse medical imaging data. Unique domain issues with medical image datasets require further study, development, and dissemination of best practices and standards, and a coordinated effort among medical imaging domain experts, medical imaging informaticists, government and industry data scientists, and interested commercial, academic, and government entities. High-level attributes of reusable medical image datasets suitable to train, test, validate, verify, and regulate ML products should be better described. NIH and other government agencies should promote and, where applicable, enforce, access to medical image datasets. We should improve communication among medical imaging domain experts, medical imaging informaticists, academic clinical and basic science researchers, government and industry data scientists, and interested commercial entities.

Radiology reporting-from Hemingway to HAL?

The job of the diagnostic radiologist is two-fold: identifying and interpreting the information available from diagnostic imaging studies and communicating that interpretation meaningfully to the referring clinician. However skilled our interpretive abilities, our patients are not well served if we fail to convey our conclusions effectively. Despite the central importance of communication skills to the work of radiologists, trainees rarely receive significant formal training in reporting skills, and much of the training given simply reflects the trainer's personal preferences. Studies have shown a preference among referrers for reports in a structured form, with findings given in a standard manner, followed by a conclusion. The technical competence to incorporate structured report templates into PACS/RIS systems is growing, "...and radiology societies (including the European Society of Radiology (ESR)) are active in producing and validating templates for a wide range of modalities and clinical circumstances. While some radiologists may prefer prose format reports, and much literature has been produced addressing "dos and don'ts" for such prose reports, it seems likely that structured reporting will become the norm in the near future. Benefits will include homogenisation and standardisation of reports, certainty that significant information has not been omitted, and capacity for data-mining of structured reports for research and teaching purposes.

TEACHING POINTS

• The radiologist's job includes interpretation of imaging studies AND communication. • Traditionally, communication has taken the form of a prose report. • Referrers have been shown to prefer reports in a structured format. • Structured reports have many advantages over traditional prose reports. • It is likely that structured reports represent the future standard.

Detection accuracy of maxillary sinus floor septa in panoramic radiographs using CBCT as gold standard: a multi-observer receiver operating characteristic (ROC) study

OBJECTIVES

To investigate diagnostic accuracy of panoramic radiography in detecting maxillary sinus floor septa by means of a multi-observer receiver operating characteristic (ROC) analysis and a standardized protocol for reporting (STARD protocol; Clin Chem 49(1):1-6, 2003).

MATERIAL AND METHODS

From our database, 25 cone beam computed tomographies (CBCTs) were selected with one maxillary sinus floor septum (height ≥ 2.5 mm). For the same patient, a recent panoramic radiograph (PAN) had to be available in the database. As controls, 28 CBCTs plus corresponding PANs without evidence of a sinus septum were selected. Using the CBCTs as ground truth, 17 observers from our dental school on a five-point confidence scale rated both sinuses in all 53 PANs with respect to presence/absence of a sinus septum. Areas beneath ROC curves (Az-values), sensitivity/specificity (SNT/SPF), positive/negative predictive values (PPV, NPV), and positive/negative likelihood ratios (LR+, LR-) were computed for each observer and pooled over all observers. Inter-rater reproducibility was assessed by means of the intraclass coefficient (ICC) using a two-way random effects model.

RESULTS

A pooled Az-value of 0.839 was observed (SNT 84.6%, SPF 73.5%). PPV ranged between 0.492 and 0.824 (median 0.627) and NPV between 0.838 and 0.976 (median 0.917). A median LR+ of 3.567 was computed (LR- median 0.193). Inter-rater reliability revealed an ICC of 0.55 (95% confidence interval 0.48 to 0.62).

CONCLUSIONS

Our results indicate that PAN is a moderately accurate method for sinus elevation planning for the purpose of septum detection. Ruling out a septum by PAN seems to work more accurately than ruling in.

CLINICAL RELEVANCE

For the purpose of maxillary sinus floor septa detection, panoramic radiography can be relatively safely advocated, particularly for judgment of a septum-free sinus.

Traumatic fractures in adults: missed diagnosis on plain radiographs in the Emergency Department

Radiography remains the imaging standard for fracture detection after trauma. The radiographic diagnosis of most fractures and dislocations poses little difficulty to radiologists: however, occasionally these injuries are quite subtle or even impossible to detect on radiographs. Missed diagnoses of fracture potentially have important consequences for patients, clinicians, and radiologists. Radiologists play a pivot role in the diagnostic assessment of the trauma patients: emergency radiologists who are more practiced at seeking out and discerning traumatic fractures can provide an invaluable service to their clinical colleagues by ensuring that patients do not endure delayed diagnoses. This is a narrative review article aims to highlight the spectrum of fractures in adults potentially missed on plain radiographs, the causes of error in diagnosis of fractures in the emergency setting and the key elements to reduce misdiagnosis of fractures.

Radiologic discrepancies in children with special healthcare needs in a pediatric emergency department

BACKGROUND

After-hours radiologic interpretation by nonradiology attendings or resident radiologists introduces the risk of discrepancies. Clinical outcomes following radiologic discrepancies among pediatric emergency department (ED) patients are poorly described. In particular, children with special healthcare needs (CSHCN), have more opportunities for discrepancies and potential consequences than non- CSHCN. Our objective was to determine the rates and types of radiologic discrepancies, and to compare CSHCN to non-CSHCN.

METHODS

From July 2014 to February 2015, all children who underwent a diagnostic imaging study at a free-standing children's ED were included. Data collected included radiologic studies - type and location - and clinical details - chief complaint and CSHCN type. Differences between preliminary reads and final pediatric radiology attending reads were defined as discrepancies, and categorized by clinical significance. Descriptive statistics, z-tests, and chi-square were used.

RESULTS

Over 8months, 8310 visits (7462 unique patients) had radiologic studies (2620 CSHCN, 5690 non-CSHCN). A total of 198 (2.4%) radiologic discrepancies [56 (28.3%) CSHCN, 142 (71.7%) non-CSHCN] were found. Chief complaints for CSCHN were more often within the cardiac, pulmonary and neurologic systems (p<0.001 for each), whereas non-CSHCN presented with more trauma (p<0.001). The rates of discrepancies (CSHCN 2.1%, non- CSHCN 2.5%, p=0.3) and severity of clinical consequences (p=0.6) were not significantly different between CSHCN and non-CSHCN.

CONCLUSION

Though the frequency and type of radiologic studies performed between CSHCN and non-CSHCN were different, we found no significant difference in the rate of radiologic discrepancies or the rate of clinically significant radiologic discrepancies.

Prospective evaluation of computer-assisted analysis of skeletal lesions for the staging of prostate cancer

Background

The purpose of this study was to compare the agreement of the bone scan index (BSI) using EXINI Bone^BSI versus experts’ readings in the initial staging for bone metastasis in prostate cancer. In addition, the diagnostic outcome was assessed in a large subset of patients where a true reference for metastases could be determined based on clinical and biochemical follow-up and/or supplementary imaging.

Methods

A total of 342 patients had a bone scintigraphy as part of routine staging for prostate cancer. Supplementary imaging was obtained at the discretion of the referring urologist. After full recruitment, the BSI and the number of malignant lesions were calculated using EXINI Bone^BSI, and three imaging experts independently classified bone status by a dichotomous outcome (M1 for bone metastasis, M0 for no bone metastasis). A true reference was available in a subset of the patients based on post-operative prostate-specific antigen responses after radical prostatectomy and/or supplementary imaging.

Results

Software analysis with a BSI > 0 as the cut-off for metastasis showed excellent agreement with expert classification for M1 disease (96% of the patients) but modest agreement for M0 disease (38%). With a BSI > 1, the agreement was 58% for M1 and 98% for M0. Software analyses based on individual European Association of Urology risk classification did not improve the diagnostic performance. Among patients with a true reference, the software showed metastasis in 64% of the M0 patients but correctly classified metastases in all M1 patients. The sensitivity was 100%, the specificity was 36%, the positive predictive value was 12.6% and the negative predictive value was 100% with a BSI >0 compared with 66.7%, 97.8%, 72.7%, and 97.0% with a BSI > 1.

Conclusion

The diagnostic value of using EXINI Bone for the BSI in the staging of newly diagnosed prostate cancer is limited.

Error and discrepancy in radiology: inevitable or avoidable?

Errors and discrepancies in radiology practice are uncomfortably common, with an estimated day-to-day rate of 3-5% of studies reported, and much higher rates reported in many targeted studies. Nonetheless, the meaning of the terms "error" and "discrepancy" and the relationship to medical negligence are frequently misunderstood. This review outlines the incidence of such events, the ways they can be categorized to aid understanding, and potential contributing factors, both human- and system-based. Possible strategies to minimise error are considered, along with the means of dealing with perceived underperformance when it is identified. The inevitability of imperfection is explained, while the importance of striving to minimise such imperfection is emphasised.

TEACHING POINTS

• Discrepancies between radiology reports and subsequent patient outcomes are not inevitably errors. • Radiologist reporting performance cannot be perfect, and some errors are inevitable. • Error or discrepancy in radiology reporting does not equate negligence. • Radiologist errors occur for many reasons, both human- and system-derived. • Strategies exist to minimise error causes and to learn from errors made.

Is the Training of Imaging Informatics Personnel in New Zealand Adequate?

The purpose of this study of Imaging Informatics Professionals (IIPs) in New Zealand was to assess their experience, background, educational qualifications and needs for support and continuing education. The IIP role includes administration of DICOM modalities, picture archiving and communication systems (PACS), radiology information systems (RIS) and many additional software and hardware systems, including the interface to New Zealand's nationwide individual electronic medical records (EMR) system. Despite the complexity of current systems, training programmes for IIPs are almost non-existent in Australasia. This cross-sectional qualitative case study used triangulated data sources, via online questionnaire, interview and critical incident analysis. Demographic data was also obtained from the questionnaire. Participants included about one third of the IIPs in New Zealand. Quantitative results were summarised with descriptive statistics or frequency data. Qualitative data was assessed by iterative multi-staged thematic analysis. This study found that the IIP role is undertaken by personnel from diverse backgrounds. Most of the IIPs learned what they know from vendors and on the job. Many feel that their biggest issue is in not knowing what they do not know and therefore not having sufficient understanding of the imaging informatics field. Only one IIP had any formal certification in PACS administration. Most respondents indicated their desire for some form of additional training. The number of IIPs in New Zealand healthcare is very small, so neither a formal training programme nor regulatory body is viable or justified. However, IIPs believe there is a need for education, regulation and recognition that their role is a critical component in healthcare.

Opportunities for Targeted Education: Critical Neuroradiologic Findings Missed or Misinterpreted by Residents and Fellows

OBJECTIVE

We reviewed neuroradiology cases in which a resident or fellow missed a significant finding, to identify potential areas of deficiency that could be strengthened through targeted education.

MATERIALS AND METHODS

Included in the study were all neuroradiology reports from 2011 through 2013 that were marked with an electronic flag to indicate a significant modification between the preliminary and final versions. The reports were examined to determine whether a critical finding (CF) or a non-CF was missed, with the use of a hospital-approved list of 17 neuroradiology CFs. Results were analyzed for all trainees.

RESULTS

A total of 978 modified reports were found among reports from 225,628 neuroradiology examinations. Of these modified reports, 891 (91.1%) contained an addendum that identified the discrepancy: 658 (73.8%) contained a CF,192 (21.7%) contained a non-CF, and 41 (4.6%) were changed from containing a CF to not containing a CF. A total of 725 missed CFs were found in the 658 modified reports. The CF miss rate for all trainees was 6.0% (95% CI, 5.6-6.4%), whereas that for residents was 8.6% and that for fellows was 4.8%. Residents missed hydrocephalus, intracranial pressure or edema, new hemorrhage, and new infarction more frequently than did fellows. The five most frequently missed CFs were congenital variation, infection, misplaced hardware, a new or enlarging mass, and vascular abnormality.

CONCLUSION

Our trainees' overall CF miss rate was 6.0%. Five CFs had miss rates of approximately 10% or more, and residents missed four of the CFs more frequently than did fellows. With the use of these data, our curriculum could potentially be strengthened and our trainee error rates decreased, leading to improved patient care.

Computer-assisted interpretation of planar whole-body bone scintigraphy in patients with newly diagnosed prostate cancer

PURPOSE

The aim of this study was to compare the diagnostic properties of EXINI Bone(BSI) in newly diagnosed prostate cancer in comparison with expert reading.

MATERIALS AND METHODS

Bone scintigraphy was performed in consecutive patients referred for staging at three clinics (342 patients with DICOM file format, 272 with Interfile format). Images were reported by three independent readers on a four-point scale (class 1-4) and by using a dichotomous outcome (M1 or M0). The software analyzed data in balanced mode, as well as using 'patient-specific' settings (based on tumor characteristics), and classified outcome as normal (N), probably normal (pN), probably abnormal (pA), and abnormal (A).

RESULTS

Classification of bone metastasis using the software (pA+A) versus experts (class 3+4) showed a sensitivity of 93.3%, specificity of 89.3%, positive predictive value of 57.5%, and negative predictive value of 98.9% with DICOM files. The diagnostic properties of the software were notably different with Interfile format. For example, expert M1 versus software A showed a sensitivity of 90.0%, specificity of 98.9%, positive predictive value of 88.2%, and negative predictive value of 98.3% with DICOM files, versus 69.2, 88.2, 38.3, and 96.4% with Interfile format, respectively. Generally, patient-specific settings did not influence the diagnostic characteristics of the software versus balanced setting with expert reading as reference.

CONCLUSION

EXINI Bone(BSI) showed high sensitivity and specificity for bone metastasis in patients with newly diagnosed prostate cancer. The software ruled out metastasis with confidence, whereas the positive predictive value was modest. The diagnostic properties were different for DICOM and Interfile file formats.

Errors in imaging patients in the emergency setting

Emergency and trauma care produces a "perfect storm" for radiological errors: uncooperative patients, inadequate histories, time-critical decisions, concurrent tasks and often junior personnel working after hours in busy emergency departments. The main cause of diagnostic errors in the emergency department is the failure to correctly interpret radiographs, and the majority of diagnoses missed on radiographs are fractures. Missed diagnoses potentially have important consequences for patients, clinicians and radiologists. Radiologists play a pivotal role in the diagnostic assessment of polytrauma patients and of patients with non-traumatic craniothoracoabdominal emergencies, and key elements to reduce errors in the emergency setting are knowledge, experience and the correct application of imaging protocols. This article aims to highlight the definition and classification of errors in radiology, the causes of errors in emergency radiology and the spectrum of diagnostic errors in radiography, ultrasonography and CT in the emergency setting.

Observer agreement and accuracy in the evaluation of bone scans in newly diagnosed prostate cancer

OBJECTIVE

The aim of the study was to assess the agreement in the interpretation of bone scintigraphy (BS) in a newly diagnosed prostate cancer.

MATERIALS AND METHODS

A total of 635 consecutive patients had their planar whole-body BS independently reviewed by three nuclear medicine physicians and classified by two grading systems: (a) a four-category scale (1: benign; 2: equivocal; 3: most likely malignant; and 4: multiple metastases) and (b) a dichotomous scale (bone metastasis present or absent).

RESULTS

Agreement in the same category, or with one or two categories of differences, was found in 66, 34, and 1.3% of the readings, respectively. Average κ-values were 0.59, 0.72, and 0.83 for unweighted, linear, and quadratic weighted variants, respectively. Very high agreement was observed (96% of the readings) with the dichotomous scale (average κ=0.87); a comparison with a final imaging diagnosis with additional CT or MRI showed a sensitivity of 83% and a specificity of 98%. BS categories 1, 3, and 4 were consistent with the final imaging diagnosis in 96-99% of cases. The prevalence of metastasis was 10% in category 2. To optimize the diagnostic characteristics, category 2 should be regarded as a separate option.

CONCLUSION

Close agreement was found among trained observers for the evaluation of BS in prostate cancer. The high level of agreement with a dichotomous scale was hampered by diagnostic misclassification. A scale with equivocal findings on planar BS is considered important to allow for additional imaging and correct staging at the bone level of BS in a population with newly diagnosed prostate cancer.

Breast Imaging in the Era of Big Data: Structured Reporting and Data Mining

OBJECTIVE

The purpose of this article is to describe structured reporting and the development of large databases for use in data mining in breast imaging.

CONCLUSION

The results of millions of breast imaging examinations are reported with structured tools based on the BI-RADS lexicon. Much of these data are stored in accessible media. Robust computing power creates great opportunity for data scientists and breast imagers to collaborate to improve breast cancer detection and optimize screening algorithms. Data mining can create knowledge, but the questions asked and their complexity require extremely powerful and agile databases. New data technologies can facilitate outcomes research and precision medicine.