Towards nationally curated data archives for clinical radiology image analysis at scale: Learnings from national data collection in response to a pandemic.

The prevalence of the coronavirus SARS-CoV-2 disease has resulted in the unprecedented collection of health data to support research. Historically, coordinating the collation of such datasets on a national scale has been challenging to execute for several reasons, including issues with data privacy, the lack of data reporting standards, interoperable technologies, and distribution methods. The coronavirus SARS-CoV-2 disease pandemic has highlighted the importance of collaboration between government bodies, healthcare institutions, academic researchers and commercial companies in overcoming these issues during times of urgency. The National COVID-19 Chest Imaging Database, led by NHSX, British Society of Thoracic Imaging, Royal Surrey NHS Foundation Trust and Faculty, is an example of such a national initiative. Here, we summarise the experiences and challenges of setting up the National COVID-19 Chest Imaging Database, and the implications for future ambitions of national data curation in medical imaging to advance the safe adoption of artificial intelligence in healthcare.

International survey on residency programs in radiology: similarities and differences among 17 countries.

OBJECTIVE: With the initiative of the ACR International Economics Committee, a multinational survey was conducted to evaluate radiology residency programs around the world. METHODS: A 31-question survey was developed. It included: economic issues, program size and length, resident's activities during daytime and call, academic aspects including syllabus and examinations. Data was tabulated using the forementioned thematic framework and was qualitatively analyzed. RESULTS: Responses were received from all 17 countries that were invited to participate (France, Netherlands, Israel, UK, Russia, USA, Japan, India, Germany, Canada, Turkey, Croatia, Serbia, Italy, Ireland, Hungary, and Greece). Residency length varied between 2 and 5 years. The certificate of residency completion is provided by a local hospital [4/17 (23%)], University [6/17 (36%)], National Board [6/17 (36%)], and Ministry of Health [1/17 (6%)]. There was variability among the number of residency programs and residents per program ranging from 15 to 300 programs per nation with a 1-700 residents in each one respectively. Salaries varied significantly and ranged from 8000 to 75,000 USD equivalent. Exams are an integral part of training in all surveyed countries. Length of call varied between 5 and 26 h and the number of monthly calls ranged from 3 to 6. The future of radiology was judged as growing in [12/17 (70%)] countries and stagnant in [5/17 (30%)] countries. DISCUSSION: Radiology residency programs worldwide have many similarities. The differences are in the structure of the residency programs. Stagnation and uncertainties need to be addressed to ensure the continued development of the next generation of radiologists. SUMMARY STATEMENT: There are many similarities in the academic aims and approach to education and training of radiology residency programs worldwide. The differences are in the structure of the residency programs and payments to individual residents.

An International Survey of Quality and Safety Programs in Radiology.

PURPOSE: The aim of this study was to determine the status of radiology quality improvement programs in a variety of selected nations worldwide. METHODS: A survey was developed by select members of the International Economics Committee of the American College of Radiology on quality programs and was distributed to committee members. Members responded on behalf of their country. The 51-question survey asked about 12 different quality initiatives which were grouped into 4 themes: departments, users, equipment, and outcomes. Respondents reported whether a designated type of quality initiative was used in their country and answered subsequent questions further characterizing it. RESULTS: The response rate was 100% and represented Australia, Canada, China, England, France, Germany, India, Israel, Japan, the Netherlands, Russia, and the United States. The most frequently reported quality initiatives were imaging appropriateness (91.7%) and disease registries (91.7%), followed by key performance indicators (83.3%) and morbidity and mortality rounds (83.3%). Peer review, equipment accreditation, radiation dose monitoring, and structured reporting were reported by 75.0% of respondents, followed by 58.3% of respondents for quality audits and critical incident reporting. The least frequently reported initiatives included Lean/Kaizen exercises and physician performance assessments, implemented by 25.0% of respondents. CONCLUSION: There is considerable diversity in the quality programs used throughout the world, despite some influence by national and international organizations, from whom further guidance could increase uniformity and optimize patient care in radiology.

Breast ultrasound region of interest detection and lesion localisation.

In current breast ultrasound computer aided diagnosis systems, the radiologist preselects a region of interest (ROI) as an input for computerised breast ultrasound image analysis. This task is time consuming and there is inconsistency among human experts. Researchers attempting to automate the process of obtaining the ROIs have been relying on image processing and conventional machine learning methods. We propose the use of a deep learning method for breast ultrasound ROI detection and lesion localisation. We use the most accurate object detection deep learning framework - Faster-RCNN with Inception-ResNet-v2 - as our deep learning network. Due to the lack of datasets, we use transfer learning and propose a new 3-channel artificial RGB method to improve the overall performance. We evaluate and compare the performance of our proposed methods on two datasets (namely, Dataset A and Dataset B), i.e. within individual datasets and composite dataset. We report the lesion detection results with two types of analysis: (1) detected point (centre of the segmented region or the detected bounding box) and (2) Intersection over Union (IoU). Our results demonstrate that the proposed methods achieved comparable results on detected point but with notable improvement on IoU. In addition, our proposed 3-channel artificial RGB method improves the recall of Dataset A. Finally, we outline some future directions for the research.

Primary care referral for knee MRI in the United Kingdom: Association with demography and subsequent surgical intervention.

BACKGROUND: Since 2008 primary care physicians (GPs) in our region have been allowed open access to knee MRI scans. There are questions about whether this changes referral practice and if it is an effective use of resources. PURPOSE: To describe the change in demographics of patients referred for knee MRI following implementation of a new referral pathway. STUDY TYPE: Retrospective observational study. POPULATION: All primary care referrals between 2008 and 2015 for knee MRI from a population of 900,000. FIELD STRENGTH/SEQUENCE: Not applicable. ASSESSMENT: Demographic profile and number of knee MRI referrals and subsequent arthroscopies. STATISTICAL TESTS: Comparisons between urban and rural populations used the t-test. Test for normality used Shapiro-Wilks. Comparison between abnormal MRI proportions used a chi-squared test. RESULTS: There were 23,928 knee MRI referrals (10,695 from GPs) between 2000 and 2015. MRI knee referrals rose from 210 in 2008 to 2379 in 2015. The average age of the patient decreased from 46.8 (SD = 14.9) in 2008 to 41.3 (SD = 14.7) in 2015. Conversion to arthroscopy declined from 15.4% to 10.2%, but there was no significant change in abnormal scan proportion. Conversion rates showed no significant difference between rural (9.6%) and urban populations (10.5%). Referral rates were significantly higher in low socioeconomic status areas (47.3% vs. 34.6%). The median referral rate per 1000 patients was 13.8 (interquartile range = 8.4). Referral rates varied widely between practices. DATA CONCLUSION: Despite a large rise in knee MRI referrals from primary care, there has been no substantial change in the age profile, suggesting that there has been no increase in inappropriate referral of elderly patients in whom MRI is unlikely to influence management. A modest decrease in the conversion rate to arthroscopy may be reasonably offset against a decrease in secondary care referrals. Socioeconomic status of the target population must be considered when planning primary care knee MRI services. LEVEL OF EVIDENCE: 4 Technical Efficacy Stage: 6 J. Magn. Reson. Imaging 2019.

Classification of Microcalcification Clusters in Digital Mammograms Using a Stack Generalization Based Classifier.

This paper presents a machine learning based approach for the discrimination of malignant and benign microcalcification (MC) clusters in digital mammograms. A series of morphological operations was carried out to facilitate the feature extraction from segmented microcalcification. A combination of morphological, texture, and distribution features from individual MC components and MC clusters were extracted and a correlation-based feature selection technique was used. The clinical relevance of the selected features is discussed. The proposed method was evaluated using three different databases: Optimam Mammography Image Database (OMI-DB), Digital Database for Screening Mammography (DDSM), and Mammographic Image Analysis Society (MIAS) database. The best classification accuracy ( 95.00 ± 0.57 %) was achieved for OPTIMAM using a stack generalization classifier with 10-fold cross validation obtaining an A z value equal to 0.97 ± 0.01 .

Breast ultrasound lesions recognition: end-to-end deep learning approaches.

Multistage processing of automated breast ultrasound lesions recognition is dependent on the performance of prior stages. To improve the current state of the art, we propose the use of end-to-end deep learning approaches using fully convolutional networks (FCNs), namely FCN-AlexNet, FCN-32s, FCN-16s, and FCN-8s for semantic segmentation of breast lesions. We use pretrained models based on ImageNet and transfer learning to overcome the issue of data deficiency. We evaluate our results on two datasets, which consist of a total of 113 malignant and 356 benign lesions. To assess the performance, we conduct fivefold cross validation using the following split: 70% for training data, 10% for validation data, and 20% testing data. The results showed that our proposed method performed better on benign lesions, with a top "mean Dice" score of 0.7626 with FCN-16s, when compared with the malignant lesions with a top mean Dice score of 0.5484 with FCN-8s. When considering the number of images with Dice score > 0.5 , 89.6% of the benign lesions were successfully segmented and correctly recognised, whereas 60.6% of the malignant lesions were successfully segmented and correctly recognized. We conclude the paper by addressing the future challenges of the work.

Deep learning in mammography and breast histology, an overview and future trends.

Recent improvements in biomedical image analysis using deep learning based neural networks could be exploited to enhance the performance of Computer Aided Diagnosis (CAD) systems. Considering the importance of breast cancer worldwide and the promising results reported by deep learning based methods in breast imaging, an overview of the recent state-of-the-art deep learning based CAD systems developed for mammography and breast histopathology images is presented. In this study, the relationship between mammography and histopathology phenotypes is described, which takes biological aspects into account. We propose a computer based breast cancer modelling approach: the Mammography-Histology-Phenotype-Linking-Model, which develops a mapping of features/phenotypes between mammographic abnormalities and their histopathological representation. Challenges are discussed along with the potential contribution of such a system to clinical decision making and treatment management.

Classification of micro-calcification in mammograms using scalable linear Fisher discriminant analysis.

Breast cancer is one of the major causes of death in women. Computer Aided Diagnosis (CAD) systems are being developed to assist radiologists in early diagnosis. Micro-calcifications can be an early symptom of breast cancer. Besides detection, classification of micro-calcification as benign or malignant is essential in a complete CAD system. We have developed a novel method for the classification of benign and malignant micro-calcification using an improved Fisher Linear Discriminant Analysis (LDA) approach for the linear transformation of segmented micro-calcification data in combination with a Support Vector Machine (SVM) variant to classify between the two classes. The results indicate an average accuracy equal to 96% which is comparable to state-of-the art methods in the literature. Graphical Abstract Classification of Micro-calcification in Mammograms using Scalable Linear Fisher Discriminant Analysis.

Provision of out-of-hours services for acute upper gastrointestinal bleeding in England: results of the 2014-2015 BSG/NHS England national survey.

BACKGROUND: There has been a drive to raise the standard of management of acute upper gastrointestinal bleeding (AUGIB) in the UK, including three previous audits, sponsored by the British Society of Gastroenterology (BSG). OBJECTIVE: To review the results of the latest BSG/National Health Service (NHS) England national survey of endoscopy services in England between 2014 and 2015. METHOD: All NHS hospitals accepting acute admissions in England (168) were invited to respond to the survey. RESULTS: Overall, 142 hospitals (84%) returned data. 85% of hospitals used a validated risk assessment score at the time of patient's admission. While 80% of hospitals provided a 24/7 endoscopy service for unstable patients, and another 10% were in network to provide an acute service, only 60% performed an endoscopy within 24 hours for stable acute admissions or inpatients with AUGIB. 11% of hospitals operated an out-of-hours ad hoc rota. 43% felt that pressure from routine work affected their ability to offer a next-day oesophagogastroduodenoscopy service, while 20% of hospitals struggled to recruit endoscopists. 28% of units reported that the previous national audit performed in 2013 had a positive influence on service development. CONCLUSIONS: This survey has revealed significant deficiencies in provision of services for patients with AUGIB in England, without a significant increase in number of hospitals providing an emergency AUGIB service since the last national audit in 2013.

Impact of errors in recorded compressed breast thickness measurements on volumetric density classification using volpara v1.5.0 software.

PURPOSE: Mammographic density has been demonstrated to predict breast cancer risk. It has been proposed that it could be used for stratifying screening pathways and recommending additional imaging. Volumetric density tools use the recorded compressed breast thickness (CBT) of the breast measured at the x-ray unit in their calculation; however, the accuracy of the recorded thickness can vary. The aim of this study was to investigate whether inaccuracies in recorded CBT impact upon volumetric density classification and to examine whether the current quality control (QC) standard is sufficient for assessing mammographic density. METHODS: Raw data from 52 digital screening mammograms were included in the study. For each image, the clinically recorded CBT was artificially increased and decreased in increments of 1 mm to simulate measurement error, until ±15% from the recorded CBT was reached. New images were created for each 1 mm step in thickness resulting in a total of 974 images which then had volpara density grade (VDG) and volumetric density percentage assigned. RESULTS: A change in VDG was observed in 38.5% (n = 20) of mammograms when applying ±15% error to the recorded CBT and 11.5% (n = 6) was within the QC standard prescribed error of ±5 mm. CONCLUSIONS: The current QC standard of ±5 mm error in recorded CBT creates the potential for error in mammographic density measurement. This may lead to inaccurate classification of mammographic density. The current QC standard for assessing mammographic density should be reconsidered.

Breast image pre-processing for mammographic tissue segmentation.

During mammographic image acquisition, a compression paddle is used to even the breast thickness in order to obtain optimal image quality. Clinical observation has indicated that some mammograms may exhibit abrupt intensity change and low visibility of tissue structures in the breast peripheral areas. Such appearance discrepancies can affect image interpretation and may not be desirable for computer aided mammography, leading to incorrect diagnosis and/or detection which can have a negative impact on sensitivity and specificity of screening mammography. This paper describes a novel mammographic image pre-processing method to improve image quality for analysis. An image selection process is incorporated to better target problematic images. The processed images show improved mammographic appearances not only in the breast periphery but also across the mammograms. Mammographic segmentation and risk/density classification were performed to facilitate a quantitative and qualitative evaluation. When using the processed images, the results indicated more anatomically correct segmentation in tissue specific areas, and subsequently better classification accuracies were achieved. Visual assessments were conducted in a clinical environment to determine the quality of the processed images and the resultant segmentation. The developed method has shown promising results. It is expected to be useful in early breast cancer detection, risk-stratified screening, and aiding radiologists in the process of decision making prior to surgery and/or treatment.

Breast cancer screening and diagnosis in the 21st century within the U.K.

In the UK, breast cancer represents the most commonly diagnosed cancer in females and carries with it a significant morbidity and mortality. An established national screening programme is in place to identify those with the disease at an early stage in order to optimise treatment and prognosis. This article provides an overview for the non-breast specialist clinician whose practice regularly deals with women who may be the subject of the screening programme. It outlines the diagnostic methods employed in the screening programme, the controversies that have surrounded it and the evolving technologies that will improve detection and diagnostic accuracy.

A Review on Automatic Mammographic Density and Parenchymal Segmentation.

Breast cancer is the most frequently diagnosed cancer in women. However, the exact cause(s) of breast cancer still remains unknown. Early detection, precise identification of women at risk, and application of appropriate disease prevention measures are by far the most effective way to tackle breast cancer. There are more than 70 common genetic susceptibility factors included in the current non-image-based risk prediction models (e.g., the Gail and the Tyrer-Cuzick models). Image-based risk factors, such as mammographic densities and parenchymal patterns, have been established as biomarkers but have not been fully incorporated in the risk prediction models used for risk stratification in screening and/or measuring responsiveness to preventive approaches. Within computer aided mammography, automatic mammographic tissue segmentation methods have been developed for estimation of breast tissue composition to facilitate mammographic risk assessment. This paper presents a comprehensive review of automatic mammographic tissue segmentation methodologies developed over the past two decades and the evidence for risk assessment/density classification using segmentation. The aim of this review is to analyse how engineering advances have progressed and the impact automatic mammographic tissue segmentation has in a clinical environment, as well as to understand the current research gaps with respect to the incorporation of image-based risk factors in non-image-based risk prediction models.

Topological modeling and classification of mammographic microcalcification clusters.

GOAL: The presence of microcalcification clusters is a primary sign of breast cancer; however, it is difficult and time consuming for radiologists to classify microcalcifications as malignant or benign. In this paper, a novel method for the classification of microcalcification clusters in mammograms is proposed. METHODS: The topology/connectivity of individual microcalcifications is analyzed within a cluster using multiscale morphology. This is distinct from existing approaches that tend to concentrate on the morphology of individual microcalcifications and/or global (statistical) cluster features. A set of microcalcification graphs are generated to represent the topological structure of microcalcification clusters at different scales. Subsequently, graph theoretical features are extracted, which constitute the topological feature space for modeling and classifying microcalcification clusters. k-nearest-neighbors-based classifiers are employed for classifying microcalcification clusters. RESULTS: The validity of the proposed method is evaluated using two well-known digitized datasets (MIAS and DDSM) and a full-field digital dataset. High classification accuracies (up to 96%) and good ROC results (area under the ROC curve up to 0.96) are achieved. A full comparison with related publications is provided, which includes a direct comparison. CONCLUSION: The results indicate that the proposed approach is able to outperform the current state-of-the-art methods. Significance: This study shows that topology modeling is an important tool for microcalcification analysis not only because of the improved classification accuracy but also because the topological measures can be linked to clinical understanding.

A reporter-targeted approach to reducing the use of the term "superficial femoral vein" in ultrasound reports for deep vein thrombosis.

BACKGROUND: The term "superficial femoral vein" is felt to be potentially misleading and dangerous in the context of deep venous thrombosis (DVT) of the lower limb, with the preferred term now being "femoral vein." The purpose of this work was to identify a means of reducing its use in reports of lower limb sonographic scans for DVT. METHODS: A retrospective search of the reports for all lower limb sonographic scans for the detection of DVT in 2011 was performed using the Radiology Information System to identify the baseline rate of use of the term. Nontargeted and subsequently targeted e-mails were then sent to reporters of such sonographic scans over a 6-month follow-up period, and the results were reevaluated. RESULTS: The baseline rate of use of the term was 17% before any intervention. It decreased to 10% following nontargeted intervention, and to 6% following targeted e-mail intervention. There were no cases of patients coming to harm as a result of the use of the term during the study period. CONCLUSIONS: A simple and inexpensive system of e-mails to reporters resulted in a decline in the rate of use of the misleading term "superficial femoral vein."

The clinical and functional outcomes of ultrasound-guided vs landmark-guided injections for adults with shoulder pathology--a systematic review and meta-analysis.

OBJECTIVE: To compare the clinical and functional outcomes of US-guided (USG) vs landmark-guided (LMG) injection for the treatment of adults with shoulder pathology. METHOD: MEDLINE, AMED and Embase in addition to unpublished literature databases were searched from 1950 to August 2011. Studies were included if they were randomized or non-randomized controlled trials comparing USG vs LSG injections for the treatment of adults with shoulder pathology. Two reviewers independently performed data extraction and appraisal of the studies. Meta-analyses were performed where possible and when inappropriate a narrative review of the data was presented. RESULTS: Six papers including 307 patients were reviewed; 142 received LMG injections and 165 received USG injections. There was a statistically significant difference in favour of USG for pain at 6 weeks (standardized mean difference 1.03; 95% CI 0.12, 1.93; P = 0.03). There was no statistically significant difference between the injection methods with respect to shoulder function (standardized mean difference 0.33; 95% CI -0.59, 1.25; P = 0.48). There was a significant difference between interventions for shoulder abduction at 6 weeks in favour of the USG method (mean difference 2.81; 95% CI 0.67, 4.95; P = 0.01). No other movements showed a statistically significant difference. CONCLUSION: There is a statistically significant difference in pain and abduction between LMG and USG steroid injections for adults with shoulder pathology. However, these differences are small and may not represent clinically useful differences. The current evidence base is limited by a number of important methodological weaknesses, which should be considered when interpreting these findings. The cost-effectiveness of the intervention should be considered in the design of future studies.