Results for the London investigation into dielectric scanning of lesions study of the MARIA® M6 breast imaging system.

OBJECTIVES: Determine the proportion of breast cancers for which MARIA® findings correspond to the cancer, with stratification by breast density and histological type. Investigate performance in symptomatic lesions. Gain patient feedback on experience with MARIA®. METHODS: MARIA® uses a radio frequency antennae array to measure signal attenuation and back scatter to build up a 3D map of tissue dielectric values. The study was a prospective, single-centre, interventional, post-approval device study. RECRUITMENT: Patients were eligible if they were attending symptomatic breast clinic or had confirmed or suspected breast cancer from any referral source. Recruitment between May 2018 and March 2020. READING: Regions of higher signal compared to background or distinct by shape were considered candidates for lesion correspondence. Up to 4 candidate regions per breast were annotated in likelihood order for representing a true lesion. RESULTS: 389 patients were recruited, and 6 excluded. 114 patients recruited with breast cancers (2 bilateral, 5 multicentric). 57 (47%) malignant lesions showed correspondence between the MARIA® signal and the cancer. Higher correspondence was seen in invasive (50%) than in situ (29%) disease. There was no reduction in correspondence at higher breast density. Reduced signal correspondence in the central scan volume and for small lesions. MARIA® scanning was well tolerated. CONCLUSIONS: We believe that MARIA® signal corresponds to a malignant lesion in 47% of breast cancers examined. ADVANCES IN KNOWLEDGE: Our study, the largest to date for this type of technology, demonstrates successes and limitations of this MARIA® M6 version.

Lessons learned from independent external validation of an AI tool to detect breast cancer using a representative UK data set.

OBJECTIVE: To pilot a process for the independent external validation of an artificial intelligence (AI) tool to detect breast cancer using data from the NHS breast screening programme (NHSBSP). METHODS: A representative data set of mammography images from 26,000 women attending 2 NHS screening centres, and an enriched data set of 2054 positive cases were used from the OPTIMAM image database. The use case of the AI tool was the replacement of the first or second human reader. The performance of the AI tool was compared to that of human readers in the NHSBSP. RESULTS: Recommendations for future external validations of AI tools to detect breast cancer are provided. The tool recalled different breast cancers to the human readers. This study showed the importance of testing AI tools on all types of cases (including non-standard) and the clarity of any warning messages. The acceptable difference in sensitivity and specificity between the AI tool and human readers should be determined. Any information vital for the clinical application should be a required output for the AI tool. It is recommended that the interaction of radiologists with the AI tool, and the effect of the AI tool on arbitration be investigated prior to clinical use. CONCLUSION: This pilot demonstrated several lessons for future independent external validation of AI tools for breast cancer detection. ADVANCES IN KNOWLEDGE: Knowledge has been gained towards best practice procedures for performing independent external validations of AI tools for the detection of breast cancer using data from the NHS Breast Screening Programme.

A longitudinal investigation of dietary diversity during the COVID-19 pandemic in Mandinka households in Kanifing, Brikama, and the West Kiang region in The Gambia.

The Covid pandemic has exposed fissures of inequality through heightened food insecurity and nutritional deficiency for vulnerable social cohorts with limited coping mechanisms. Given the multi-dimensional pathways through which its effects have been felt, several researchers have highlighted the need to analyse the pandemic in specific contexts. Using random and fixed effect regression models, this study analyzed longitudinal survey data collected from 103 Mandinka households in rural and urban Gambia. The study employed convenience and snowball sampling and involved the monthly collection of detailed income, food consumption, expenditure, sourcing, migration, health, and coping mechanism data through mobile phone interviews which yielded 676 observations. Food insecurity was manifest in terms of quality, not quantity, and spread unevenly across food types and households. Dietary outcomes and sourcing strategies were associated with location, improved sanitation, household size, changes in monthly income, Covid policy stringency, and Covid cases but these associations varied by food group. Staples were the most frequently consumed food group, and dark green vegetables were the least. Rural communities were more likely to eat more healthy millets but much less likely to consume dairy products or roots and tubers. Access to own production was also important for Vitamin A-rich foods but higher incomes and markets were key for protein and heme-iron-rich foods. Tighter Covid policy stringency was negatively associated with dietary diversity and, along with fear of market hoarding, was positively associated with reliance on a range of consumption and production coping mechanisms. Resilience was higher in larger households and those with improved water and sanitation. The number of Covid cases was associated with higher consumption of protein-rich foods and greater reliance on own produced iron-rich foods. Very few households received Government aid and those that did already had access to other income sources. Our findings suggest that the nature of food insecurity may have evolved over time during the pandemic. They also reiterate not only the importance of access to markets and employment but also that the capacity to absorb affordability shocks and maintain food choices through switching between sources for specific nutritious food groups varied by household and location.

Fair shares: building and benefiting from healthcare AI with mutually beneficial structures and development partnerships.

Artificial intelligence (AI) algorithms are used in an increasing range of aspects of our lives. In particular, medical applications of AI are being developed and deployed, including many in image analysis. Deep learning methods, which have recently proved successful in image classification, rely on large volumes of clinical data generated by healthcare institutions. Such data is collected from their served populations. In this opinion article, using digital mammographic screening as an example, we briefly consider the background to AI development and some issues around its deployment. We highlight the importance of high quality clinical data as fundamental to these technologies, and question how the ownership of resultant tools should be defined. Though many of the ethical issues concerning the development and use of medical AI technologies continue to be discussed, the value of the data on which they rely remains a subject that is seldom considered. This potentially controversial issue can and should be addressed in a way which is beneficial to all parties, particularly the population in general and the patients we serve.

International evaluation of an AI system for breast cancer screening.

Screening mammography aims to identify breast cancer at earlier stages of the disease, when treatment can be more successful1. Despite the existence of screening programmes worldwide, the interpretation of mammograms is affected by high rates of false positives and false negatives2. Here we present an artificial intelligence (AI) system that is capable of surpassing human experts in breast cancer prediction. To assess its performance in the clinical setting, we curated a large representative dataset from the UK and a large enriched dataset from the USA. We show an absolute reduction of 5.7% and 1.2% (USA and UK) in false positives and 9.4% and 2.7% in false negatives. We provide evidence of the ability of the system to generalize from the UK to the USA. In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%. We ran a simulation in which the AI system participated in the double-reading process that is used in the UK, and found that the AI system maintained non-inferior performance and reduced the workload of the second reader by 88%. This robust assessment of the AI system paves the way for clinical trials to improve the accuracy and efficiency of breast cancer screening.

MARIA® M5: A multicentre clinical study to evaluate the ability of the Micrima radio-wave radar breast imaging system (MARIA®) to detect lesions in the symptomatic breast.

The MARIA® breast imaging system is a clinical diagnostic tool that uses a hemispherical array of radiowave antennas to generate three-dimensional images of the internal breast. The system utilises the variance of dielectric contrast within the breast volume in order to identify areas of interest for further diagnostic investigation. This multicentre study of 225 patients was conducted at three trial sites and recruited women with both malignant and benign lesions. The MARIA® images from the study were read by both clinicians who had access to the patient's clinical information, as well as by 'blind' reviewers who did not. Results from the study show an overall sensitivity of 76% for the system, which was similar across benign and malignant findings, and in denser breasts. The results from this study are outlined here and discussions on ongoing and future work with MARIA® are deliberated.

Co-existing Paget's disease and ankylosing spondylitis resulting in panthoracic pagetic vertebral ankylosis.

Pagetic vertebral ankylosis is an uncommon presentation and occurs when Paget's disease is associated with diffuse idiopathic skeletal hyperostosis and ankylosing spondylitis. In these cases, the pagetic trait extends from one vertebra to another by invasion of the intervertebral disc space. Such acquired vertebral ankylosis is extremely uncommon but possible when bony bridging syndesmophytes or osteophytes are present. We describe one such case, where a delayed diagnosis resulted in the most extensive pagetic vertebral ankylosis described in the literature and severe patient morbidity.

Microbial keratitis in keratoglobus-associated vernal keratoconjunctivitis and atopic dermatitis.

PURPOSE: The authors describe a patient with keratoglobus and a history of vernal keratoconjunctivitis and atopic dermatitis who acutely developed corneal hydrops and severe microbial keratitis. The infectious keratitis responded poorly to medical management and resulted in enucleation of the eye. METHODS: A 25-year-old man presented with an acutely painful, red left eye. He had an ocular history of keratoglobus in association with vernal keratoconjunctivitis and atopic dermatitis. His visual acuity was light perception in the left eye and 20/40 in the right eye. Ocular examination showed a grossly edematous cornea with breaks in Descemet's membrane, a central infiltrative ulcer, and hypopyon. RESULTS: Corneal scrapings showed gram-positive beta-hemolytic streptococci, for which topical treatment of ceftazidime and benzylpenicillin every 30 minutes by day and night was commenced with an oral course of ciprofloxacin 750 mg twice a day. Despite aggressive therapy, the microbial keratitis progressed, and the hypopyon increased in size. The patient subsequently underwent enucleation of his left eye. CONCLUSIONS: This is the first reported case of microbial keratitis with corneal hydrops in a patient with keratoglobus, vernal keratoconjunctivitis, and atopic dermatitis. His ocular and dermatologic comorbidity may have impaired corneal integrity and allowed penetration of organisms, resulting in a severe case of keratitis that responded poorly to medical management.