Overview of trials on artificial intelligence algorithms in breast cancer screening - A roadmap for international evaluation and implementation.

Accumulating evidence from retrospective studies demonstrate at least non-inferior performance when using AI algorithms with different strategies versus double-reading in mammography screening. In addition, AI algorithms for mammography screening can reduce work load by moving to single human reading. Prospective trials are essential to avoid unintended adverse consequences before incorporation of AI algorithms into UK's National Health Service (NHS) Breast Screening Programme (BSP). A stakeholders' meeting was organized in Newnham College, Cambridge, UK to undertake a review of the current evidence to enable consensus discussion on next steps required before implementation into a screening programme. It was concluded that a multicentre multivendor testing platform study with opt-out consent is preferred. AI thresholds from different vendors should be determined while maintaining non-inferior screening performance results, particularly ensuring recall rates are not increased. Automatic recall of cases using an agreed high sensitivity AI score versus automatic rule out with a low AI score set at a high sensitivity could be used. A human reader should still be involved in decision making with AI-only recalls requiring human arbitration. Standalone AI algorithms used without prompting maintain unbiased screening reading performance, but reading with prompts should be tested prospectively and ideally provided for arbitration.

An optoacoustic imaging feature set to characterise blood vessels surrounding benign and malignant breast lesions.

Combining optoacoustic (OA) imaging with ultrasound (US) enables visualisation of functional blood vasculature in breast lesions by OA to be overlaid with the morphological information of US. Here, we develop a simple OA feature set to differentiate benign and malignant breast lesions. 94 female patients with benign, indeterminate or suspicious lesions were recruited and underwent OA-US. An OA-US imaging feature set was developed using images from the first 38 patients, which contained 14 malignant and 8 benign solid lesions. Two independent radiologists blindly scored the OA-US images of a further 56 patients, which included 31 malignant and 13 benign solid lesions, with a sensitivity of 96.8% and specificity of 84.6%. Our findings indicate that OA-US can reveal vascular patterns of breast lesions that indicate malignancy using a simple feature set based on single wavelength OA data, which is therefore amenable to application in low resource settings for breast cancer management.

Opportunities in cancer imaging: risk-adapted breast imaging in screening.

In the UK, women between 50-70 years are invited for 3-yearly mammography screening irrespective of their likelihood of developing breast cancer. The only risk adaption is for women with >30% lifetime risk who are offered annual magnetic resonance imaging (MRI) and mammography, and annual mammography for some moderate-risk women. Using questionnaires, breast density, and polygenic risk scores, it is possible to stratify the population into the lowest 20% risk, who will develop <4% of cancers and the top 4%, who will develop 18% of cancers. Mammography is a good screening test but has low sensitivity of 60% in the 9% of women with the highest category of breast density (BIRADS D) who have a 2.5- to fourfold breast cancer risk. There is evidence that adding ultrasound to the screening mammogram can increase the cancer detection rate and reduce advanced stage interval and next round cancers. Similarly, alternative tests such as contrast-enhanced mammography (CESM) or abbreviated MRI (ABB-MRI) are much more effective in detecting cancer in women with dense breasts. Scintimammography has been shown to be a viable alternative for dense breasts or for follow-up in those with a personal history of breast cancer and scarring as result of treatment. For supplemental screening to be worthwhile in these women, new technologies need to reduce the number of stage II cancers and be cost effective when tested in large scale trials. This article reviews the evidence for supplemental imaging and examines whether a risk-stratified approach is feasible.

A meta-analysis comparing the diagnostic performance of abbreviated MRI and a full diagnostic protocol in breast cancer.

AIM: To undertake a meta-analysis of the diagnostic performance of abbreviated (ABB) magnetic resonance imaging (MRI) and full diagnostic protocol MRI (FDP-MRI) in breast cancer. MATERIALS AND METHODS: This meta-analysis was performed using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Diagnostic Test Accuracy (PRISMA-DTA) guidelines. The PubMed and EMBASE databases were searched through August 2019 for studies comparing the diagnostic performance of ABB-MRI and FDP-MRI in the breast. Studies were reviewed by two authors independently according to eligibility and exclusion criteria and split into two subgroups (screening population studies and studies using cohorts enriched with known cancers) to avoid bias. Quality assessment and bias for diagnostic accuracy was determined with Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2). The diagnostic accuracy for each subgroup was pooled using a bivariate random effects model and summary receiver operating characteristic (sROC) curves produced. Sensitivities and specificities were compared using a paired t-test. RESULTS: Five screening (62/2,588 cancers/patients) and eight enriched cohort (540/1,432 cancers/patients) studies were included in the meta-analysis. QUADAS-2 assessment showed a low risk of bias in most studies. The pooled sensitivity/specificity/area under the receiver operating characteristic curve (AUC) for screening studies was 0.90/0.92/0.94 for ABB-MRI and 0.92/0.95/0.97 for FDP-MRI. The pooled sensitivity/specificity/AUC for enriched cohort studies was 0.93/0.83/0.94 for ABB-MRI and 0.93/0.84/0.95 for FDP-MRI. There was no significant difference in sensitivity or specificity using ABB-MRI or FDP-MRI (p=0.18 and 0.27, p=0.18 and 0.93, respectively). CONCLUSION: The diagnostic performances of the ABB-MRI and FDP-MRI protocols used in either screening or enriched cohorts were comparable. There was a large variation in patient population, study methodology, and abbreviated protocols reported.

Ten lessons of leadership: reflections of a female academic.

The defining qualities and characteristics of a good leader - visionary, decisive, clear communicator, honesty and integrity, commitment and passion, being inspirational, accountable, well organised with the ability to delegate and empower-have been well documented. There are many articles with advice on how to lead, and there are excellent courses on leadership. In this article, I would like to explore how the general advice pertains to academia (and an academic radiologist), and whether this advice is relevant and whether the principles of leadership are the same. I have reflected on my own career and I have come up with my own 10 lessons of leadership that I have learnt over my working life. There are elements of leadership that are common across different sectors of business and healthcare, and examining how these are relevant to academia can be illuminating.

Artificial intelligence in clinical imaging: a health system approach.

The development and application of artificial intelligence (AI) to radiology requires an approach that encompasses a health system. The UK government and National Health Service (NHS) are creating an ecosystem to facilitate academic/industrial partnerships aimed at accelerating the creation of relevant and robust AI tools, which will improve the development and delivery of healthcare imaging. A series of recent initiatives are described, which will drive the development and adoption of AI in clinical imaging.

Artificial intelligence in breast imaging.

This article reviews current limitations and future opportunities for the application of computer-aided detection (CAD) systems and artificial intelligence in breast imaging. Traditional CAD systems in mammography screening have followed a rules-based approach, incorporating domain knowledge into hand-crafted features before using classical machine learning techniques as a classifier. The first commercial CAD system, ImageChecker M1000, relies on computer vision techniques for pattern recognition. Unfortunately, CAD systems have been shown to adversely affect some radiologists' performance and increase recall rates. The Digital Mammography DREAM Challenge was a multidisciplinary collaboration that provided 640,000 mammography images for teams to help decrease false-positive rates in breast cancer screening. Winning solutions leveraged deep learning's (DL) automatic hierarchical feature learning capabilities and used convolutional neural networks. Start-ups Therapixel and Kheiron Medical Technologies are using DL for breast cancer screening. With increasing use of digital breast tomosynthesis, specific artificial intelligence (AI)-CAD systems are emerging to include iCAD's PowerLook Tomo Detection and ScreenPoint Medical's Transpara. Other AI-CAD systems are focusing on breast diagnostic techniques such as ultrasound and magnetic resonance imaging (MRI). There is a gap in the market for contrast-enhanced spectral mammography AI-CAD tools. Clinical implementation of AI-CAD tools requires testing in scenarios mimicking real life to prove its usefulness in the clinical environment. This requires a large and representative dataset for testing and assessment of the reader's interaction with the tools. A cost-effectiveness assessment should be undertaken, with a large feasibility study carried out to ensure there are no unintended consequences. AI-CAD systems should incorporate explainable AI in accordance with the European Union General Data Protection Regulation (GDPR).

Systematic review and meta-analysis of the reliability and discriminative validity of cartilage compositional MRI in knee osteoarthritis.

OBJECTIVE: To assess reliability and discriminative validity of cartilage compositional magnetic resonance imaging (MRI) in knee osteoarthritis (OA). DESIGN: The study was carried out per PRISMA recommendations. We searched MEDLINE and EMBASE (1974 - present) for eligible studies. We performed qualitative synthesis of reliability data. Where data from at least two discrimination studies were available, we estimated pooled standardized mean difference (SMD) between subjects with and without OA. Discrimination analyses compared controls and subjects with mild OA (Kellgren-Lawrence (KL) grade 1-2), severe OA (KL grade 3-4) and OA not otherwise specified (NOS) where not possible to stratify. We assessed quality of the evidence using Quality Appraisal of Diagnostic Reliability (QAREL) and Quality Assessment of Diagnostic Accuracy (QUADAS-2) tools. RESULTS: Fifty-eight studies were included in the reliability analysis and 26 studies were included in the discrimination analysis, with data from a total of 2,007 knees. Intra-observer, inter-observer and test-retest reliability of compositional techniques were excellent with most intraclass correlation coefficients >0.8 and coefficients of variation <10%. T1rho and T2 relaxometry were significant discriminators between subjects with mild OA and controls, and between subjects with OA (NOS) and controls (P < 0.001). T1rho showed best discrimination for mild OA (SMD [95% CI] = 0.73 [0.40 to 1.06], P < 0.001) and OA (NOS) (0.60 [0.41 to 0.80], P < 0.001). Quality of evidence was moderate for both parts of the review. CONCLUSIONS: Cartilage compositional MRI techniques are reliable and, in the case of T1rho and T2 relaxometry, can discriminate between subjects with OA and controls.

Personalised screening: is this the way forward?

Screening with mammography has been implemented in many countries across the world with most offering 2-yearly examinations between the ages of 50-69 years. Robust modelling tools that include breast density and single nucleotide polymorphisms (SNPs) have been developed to predict which women are most likely to develop breast cancer. Mammographic sensitivity is poor in women with the densest category of breast tissue, and even women with heterogeneously dense tissue may benefit from additional supplemental imaging. Digital breast tomosynthesis (DBT), automated breast ultrasound (ABUS), contrast-enhanced mammography (CESM) or abbreviated (ABB) magnetic resonance imaging (MRI) all offer the opportunity to increase cancer detection, especially in women with dense breasts at increased risk of cancer. DBT increases cancer detection by around 15% with a corresponding reduction in recall rates; ABUS has been shown to increase cancer detection by between 2-4/1,000 depending on the cohort being examined and results in increased recalls, which tend to fall in subsequent screening rounds; CESM has very high sensitivity almost matching MRI with slightly improved specificity; ABB-MRI has been shown to be virtually equivalent to standard protocol MRI examinations, making this a technique that could be considered as a screening tool in high-risk women. This article reviews the literature to establish the current status of these techniques. The cost-effectiveness of these techniques requires further investigation and screening trials should report the nature of any additional tumours that are found.

Accuracy and cost-effectiveness of dynamic contrast-enhanced CT in the characterisation of solitary pulmonary nodules-the SPUtNIk study.

INTRODUCTION: Solitary pulmonary nodules (SPNs) are common on CT. The most cost-effective investigation algorithm is still to be determined. Dynamic contrast-enhanced CT (DCE-CT) is an established diagnostic test not widely available in the UK currently. METHODS AND ANALYSIS: The SPUtNIk study will assess the diagnostic accuracy, clinical utility and cost-effectiveness of DCE-CT, alongside the current CT and 18-flurodeoxyglucose-positron emission tomography) (18FDG-PET)-CT nodule characterisation strategies in the National Health Service (NHS). Image acquisition and data analysis for 18FDG-PET-CT and DCE-CT will follow a standardised protocol with central review of 10% to ensure quality assurance. Decision analytic modelling will assess the likely costs and health outcomes resulting from incorporation of DCE-CT into management strategies for patients with SPNs. ETHICS AND DISSEMINATION: Approval has been granted by the South West Research Ethics Committee. Ethics reference number 12/SW/0206. The results of the trial will be presented at national and international meetings and published in an Health Technology Assessment (HTA) Monograph and in peer-reviewed journals. TRIAL REGISTRATION NUMBER: ISRCTN30784948; Pre-results.

Imaging tumour hypoxia with positron emission tomography.

Hypoxia, a hallmark of most solid tumours, is a negative prognostic factor due to its association with an aggressive tumour phenotype and therapeutic resistance. Given its prominent role in oncology, accurate detection of hypoxia is important, as it impacts on prognosis and could influence treatment planning. A variety of approaches have been explored over the years for detecting and monitoring changes in hypoxia in tumours, including biological markers and noninvasive imaging techniques. Positron emission tomography (PET) is the preferred method for imaging tumour hypoxia due to its high specificity and sensitivity to probe physiological processes in vivo, as well as the ability to provide information about intracellular oxygenation levels. This review provides an overview of imaging hypoxia with PET, with an emphasis on the advantages and limitations of the currently available hypoxia radiotracers.

Tumour expression of leptin is associated with chemotherapy resistance and therapy-independent prognosis in gastro-oesophageal adenocarcinomas.

BACKGROUND: Cytotoxic chemotherapy remains the main systemic therapy for gastro-oesophageal adenocarcinoma, but resistance to chemotherapy is common, resulting in ineffective and often toxic treatment for patients. Predictive biomarkers for chemotherapy response would increase the probability of successful therapy, but none are currently recommended for clinical use. We used global gene expression profiling of tumour biopsies to identify novel predictive biomarkers for cytotoxic chemotherapy. METHODS: Tumour biopsies from patients (n=14) with TNM stage IB-IV gastro-oesophageal adenocarcinomas receiving platinum-based combination chemotherapy were used as a discovery cohort and profiled with Affymetrix ST1.0 Exon Genechips. An independent cohort of patients (n=154) treated with surgery with or without neoadjuvant platinum combination chemotherapy and gastric adenocarcinoma cell lines (n=22) were used for qualification of gene expression profiling results by immunohistochemistry. A cisplatin-resistant gastric cancer cell line, AGS Cis5, and the oesophageal adenocarcinoma cell line, OE33, were used for in vitro validation investigations. RESULTS: We identified 520 genes with differential expression (Mann-Whitney U, P<0.020) between radiological responding and nonresponding patients. Gene enrichment analysis (DAVID v6.7) was used on this list of 520 genes to identify pathways associated with response and identified the adipocytokine signalling pathway, with higher leptin mRNA associated with lack of radiological response (P=0.011). Similarly, in the independent cohort (n=154), higher leptin protein expression by immunohistochemistry in the tumour cells was associated with lack of histopathological response (P=0.007). Higher leptin protein expression by immunohistochemistry was also associated with improved survival in the absence of neoadjuvant chemotherapy, and patients with low leptin protein-expressing tumours had improved survival when treated by neoadjuvant chemotherapy (P for interaction=0.038). In the gastric adenocarcinoma cell lines, higher leptin protein expression was associated with resistance to cisplatin (P=0.008), but not to oxaliplatin (P=0.988) or 5fluorouracil (P=0.636). The leptin receptor antagonist SHLA increased the sensitivity of AGS Cis5 and OE33 cell lines to cisplatin. CONCLUSIONS: In gastro-oesophageal adenocarcinomas, tumour leptin expression is associated with chemoresistance but a better therapy-independent prognosis. Tumour leptin expression determined by immunohistochemistry has potential utility as a predictive marker of resistance to cytotoxic chemotherapy, and a prognostic marker independent of therapy in gastro-oesophageal adenocarcinoma. Leptin antagonists have been developed for clinical use and leptin and its associated pathways may also provide much needed novel therapeutic targets for gastro-oesophageal adenocarcinoma.

The effect of waiting times from general practitioner referral to MRI or orthopaedic consultation for the knee on patient-based outcomes.

OBJECTIVES: The purpose of this study was to test for the effect of waiting time from general practitioner (GP) referral to MRI or to orthopaedic consultation on outcomes of patients with knee problems, and to test whether any characteristics of trial participants predicted waiting time to MRI or orthopaedics. METHODS: We undertook secondary analyses of data on 553 participants from a randomised trial who were recruited from 163 general practices during November 2002 to October 2004. RESULTS: Of the patients allocated to MRI, 263 (94%) had an MRI, and of those referred to orthopaedics, 236 (86%) had an orthopaedic consultation. The median (interquartile range) waiting time in days from randomisation to MRI was 41.0 (21.0-71.0) and to orthopaedic appointment was 78.5 (54.5-167.5). Waiting time was found to have no significant effect on patient outcome for both the Short Form 36-item (SF-36) physical functioning score (p=0.570) and the Knee Quality of Life 26-item (KQoL-26) physical functioning score (p=0.268). There was weak evidence that males waited less time for their MRI (p=0.049) and older patients waited longer for their orthopaedic referral (p=0.049). For patients who resided in the catchment areas of some centres there were significantly longer waiting times for both MRI and orthopaedic appointment. CONCLUSION: Where patients reside is a strong predictor of waiting time for access to services such as MRI or orthopaedics. There is no evidence to suggest, however, that this has a significant effect on physical well-being in the short term for patients with knee problems.

Training the next generation of radiology researchers. Report on a joint meeting of the Royal College of Radiologists and the Wellcome Trust and an overview of College strategies in developing radiology research.

As the use of imaging continues to increase in medical research, the Royal College of Radiologists (RCR) Research Committee and the Wellcome Trust held a joint meeting in 2010 to explore the current issues critical to training the next generation of radiology researchers. It was stated that despite an increase in consultant radiologist numbers, the number of academic posts has declined. This has resulted in radiology lagging behind some of the other specialties in terms of research activity and output. Some of the obstacles arising from the meeting have since been addressed by the RCR. These and other possible solutions that emanated from the interactive sessions include establishing a research culture in radiology departments, establishing a network of research radiologist mentors, improving trainee opportunities, and improving consultant opportunities. The meeting demonstrated the demand and need for extended support for research activity from the RCR. The RCR Academic Committee is formulating recommendations based on these and other sources of evidence. However, it is acknowledged that this comes at a time when national financial resources for research support are severely limited. Research in radiology remains an important part of the future development of the specialty. Supporting the next generation of potential researchers is essential, and this meeting was a means to understand how this might best be achieved.

The biological correlates of macroscopic breast tumour structure measured using fractal analysis in patients undergoing neoadjuvant chemotherapy.

Breast cancers are evolving, multi-scale systems that are characterized by varied complex spatial structures. In this study, we measured the structural characteristics of 33 breast tumours in patients who were to receive neoadjuvant chemotherapy using dynamic contrast enhanced MRI and fractal geometry. The results showed a significant association between fractal measurements and tumour characteristics. The fractal dimension was associated with receptor status (ER and PR) and the fractal fit was associated with response to chemotherapy, measured using a validated pathological response scale, tumour grade and size. This study describes structure measures that may be a consequence of known prognostic factors during the initial and/or maturation phase of tumour growth. These results suggest that measuring tumour structure in this way can predict an individual's response to neoadjuvant therapy and may identify those who will benefit least from neoadjuvant chemotherapy, allowing alternative treatment options to be selected in those patients.