Artificial intelligence (AI) implementation within the National Health Service (NHS): the South West London AI Working Group experience.

In the rapidly evolving field of artificial intelligence (AI) for radiology, with a plethora of vendor options and use-cases and evidence claims to sift through, the pressing question is how to effectively implement the right tool for enhanced patient care? This article presents a structured approach to AI deployment, drawing from a comprehensive case study in South West London. We underscore the necessity of forming a dedicated AI team with a clear vision and assertive leadership to navigate such complexities. Central to our discussion is the significance of crafting an AI implementation plan, with an overarching aim to augment patient care, promote operational efficiency, and lay down standardized protocols for seamless AI adoption. By presenting a blueprint for AI implementation within the National Health Service (NHS), we intend to demystify the process for radiology departments across the UK, enabling them to make informed decisions and empowering their staff to embrace and leverage AI responsibly ensuring that patient welfare remains at the heart of innovation. Thus, having a framework to follow when implementing an AI solution that addresses a vision for scalable adoption, core team members with diversity of skillset, staff engagement and education, plan for vendor selection, and change management is crucial for success.

Paediatric radiology training in the UK: a national trainee survey by the British Society of Paediatric Radiologists (BSPR).

AIM: To survey current UK radiology trainee experiences and opinions regarding the quality of paediatric radiology training encountered in their core years, and assess their career ambitions with regards to paediatric radiology. MATERIALS AND METHODS: A 22-question online survey, approved by the BSPR committee, was promoted over 12 months (1 February 2022 to 31 January 2023) across current radiology trainees and fellows via regional radiology training programme directors (TPDs), Junior Radiology Forum (JRF) trainee representatives, at BSPR Junior Forum webinar teaching sessions, and via social media/word of mouth. RESULTS: Eighty-three UK survey responses were received from 17/19 (89%) training schemes. Sixty of the 83 (72%) had taken or were due to take a 2-4 month core paediatric radiology placement partly at tertiary centres (66/83, 80%), with 67/83 (81%) receiving dedicated didactic teaching. Only 26/83 (31%) reported fulfilling core curriculum competencies and 32/83 (39%) reported not receiving enough paediatric radiology training. Almost a quarter (18/83, 22%) reported <2 months of paediatric radiology rotation during core training and 12% (10/83) across six training schemes, reported no dedicated paediatric teaching sessions. Respondents who left negative comments around their experience were more likely to reject paediatric radiology as a future career option (p<0.05). CONCLUSIONS: Only 31% of UK radiology trainees reported having enough paediatric radiology training to achieve core competencies. Standardised training, teaching, and increased on-call support could improve confidence in dealing with emergency cases and encourage interest in paediatric radiology beyond junior years.

Forensic post-mortem CT in children.

Post-mortem computed tomography (PMCT) imaging is gaining popularity and acceptance for use alongside forensic autopsies of children, predominantly to aid in the detection of traumatic injuries. Recent research on this topic has provided a breadth of new information regarding the appropriate usage, imaging guidance, and diagnostic accuracy for the identification of different paediatric pathologies. Additionally, advanced CT imaging techniques, such as PMCT angiography or ventilated PMCT, have been trialled, and post-mortem micro-CT is now being used in specialist centres for the assessment of subtle fractures in extracted bone specimens. Various image post-processing methods (e.g., three-dimensional printing from PMCT imaging data) are being used for the illustration of injuries in the medicolegal setting to a lay audience and provide another avenue for the future of forensic radiology research. In this review, the evidence-based principles and benefits of post-mortem imaging for forensic investigation in childhood deaths are presented, with a particular focus on PMCT and current practices. Variations in forensic imaging strategies around the world, published diagnostic accuracy rates, and expected normal post-mortem imaging findings are discussed, as well as potential future applications and research in this area.

Academic clinical fellows in radiology: how can we improve success?

AIM: To survey past and current radiology academic clinical fellows (ACFs) for feedback on their experiences, academic achievements, challenges faced in balancing academic and clinical responsibilities, and opinion on how to optimise the fellowship programme. MATERIALS & METHODS: A 26-question online survey approved by the Royal College of Radiologists (RCR) Academic Committee was distributed over a 7-month period (June 2021 to January 2022) to current and past radiology ACFs via the National Institute for Health and Care Research (NIHR) integrated academic training imaging leads, radiology training programme directors, and social media. RESULTS: Thirty-five survey responses were received from past or present ACFs. Of the respondents, 42.8% (15/35) entered ACF training from another research post, and most continued their academic interests after ACF training (59.3%, 16/27 that had completed the post). The majority (22/35, 63%) had or were in the process of obtaining a postgraduate research degree. The most common academic outputs were scientific publications and national/international conference presentations. Most (23/35, 66%) would recommend the ACF post to colleagues, although some found it challenging balancing on-call and examination commitments during training. CONCLUSIONS: Entry into the radiology ACF programme is often after a prior academic post. Many ACFs appear to enjoy their fellowship experience and continue academic interests after training, some achieving higher research degrees. Challenges in balancing clinical workload require some flexibility from local clinical and academic supervisors. Suggestions for alternative structuring of the ACF pathway and how to optimise entry into these competitive posts are also outlined.

Cinematic rendering of paediatric musculoskeletal pathologies: initial experiences with CT.

Cinematic rendering (CR) is a novel post-processing technique similar to volume rendering (VR), which allows for a more photorealistic imaging reconstruction by using a complex light modelling algorithm, incorporating information from multiple light paths and predicted photon scattering patterns. Several recent publications relating to adult imaging have argued that CR gives a better "realism" and "expressiveness" experience over VR techniques. CR has also been shown to improve visualisation of musculoskeletal and vascular anatomy compared with conventional CT viewing, and may help non-radiologists to understand complex patient anatomy. In this review, we provide an overview of how CR could be used in paediatric musculoskeletal imaging, particularly in complex diagnoses, surgical planning, and patient consent processes. We present a direct comparison of VR and CR reconstructions across a range of congenital and acquired musculoskeletal pathologies, highlighting potential advantages and areas for further research.

Value of additional lateral radiographs in paediatric skeletal surveys for suspected physical abuse.

AIM: To determine the added value of lateral limb radiographs in suspected physical abuse (SPA), particularly with regard to fracture detection and reporter confidence. METHODS AND MATERIALS: A 3-year (October 2017 to November 2020) retrospective study was conducted. Two blinded paediatric radiologists independently reviewed the appendicular radiographs for the presence of fractures, first by reviewing just the frontal projections, then both frontal and lateral radiographs. The additional yield of fractures and changes in reporting confidence scores were recorded. RESULTS: One hundred and thirty-eight skeletal surveys (29 live, 109 deceased children) were assessed, consisting of 16 appendicular fractures imaged in two projections (six wrist, five knee, five ankle). In the majority of cases (14/16) the fractures were already visible on the frontal view with only two fractures (one distal radius, one distal tibial) identified only by lateral projection on the blinded review. One fracture (distal tibia) was visible only on the frontal radiograph (not lateral view). The addition of lateral projection did not lead to overcalling of fractures in the remaining normal studies. Radiologist confidence scores showed an improvement with the addition of a lateral projection when a study was thought to be normal (p=0.001-0.003), but not when a fracture was identified (p>0.05). CONCLUSION: The addition of lateral radiographs improves reporting confidence in normal skeletal surveys, but only rarely helped to detect additional fractures given that most were visible on frontal projections. Larger multicentre studies of clinical practice are required to confirm these findings.

Academia in cardiovascular radiology: are we doing enough for the future of the subspecialty?

AIM: To explore barriers to academic activities that cardiovascular radiology-oriented radiology trainees face worldwide. MATERIALS AND METHODS: An international call for participation in an online survey was distributed via social media and radiological societies to radiology trainees. Questions covered barriers and involvement in academic activities during radiology training. Participants interested in cardiovascular radiology were selected for analysis with appropriate statistical methods. RESULTS: Of the 892 respondents, 120 (13.5%) reported an interest in cardiovascular imaging. The majority (63.3%, 76/120) were from Europe and 57.5% (69/120) were men. There were gender discrepancies in academic involvement and in perceived gender-related barriers (perceived gender barrier in academic work between women and men: 15/48 versus 5/69, respectively, p=0.001). The main barriers were lack of time, mentorship, and support. Most did not have protected academic time for research nor for teaching (61.7%, 74/120 and 57.5%, 69/120, respectively). Nonetheless, 40% (48/120) published as first authors, 77.5% (91/120) were involved academically in conferences, 71.7% (85/120) were positive about academic activities, 70.8% plan working in an academic setting, and 78.3% (94/120) would consider research training abroad. CONCLUSION: Although the majority of respondents are positive about academic activities and plan to continue in the future, most do not have protected time. Lack of time, mentorship, and support were the main barriers. Fewer women declare involvement in academic work and output. There are significant perceived gender barriers to academic activities.

Diagnostic accuracy of postmortem ultrasound vs postmortem 1.5-T MRI for non-invasive perinatal autopsy.

OBJECTIVES: To determine the diagnostic accuracy of postmortem magnetic resonance imaging (PM-MRI) and postmortem ultrasound (PM-US) for perinatal autopsy in the same patient cohort, and to determine whether PM-US can provide the same anatomical information as PM-MRI. METHODS: In this prospective, 5-year (July 2014-July 2019) single-center study, we performed 1.5-T PM-MRI and PM-US in an unselected cohort of perinatal deaths. The diagnostic accuracies of both modalities were calculated, using autopsy as the reference standard. As a secondary objective, the concordance rates between the two imaging modalities for the overall main diagnosis and for five anatomical regions (brain, spine, thorax, heart and abdomen) were calculated. RESULTS: During the study period, 136 cases underwent both PM-US and PM-MRI, of which 88 (64.7%) also underwent autopsy. There was no significant difference in the rates of concordance with autopsy between the two modalities for overall diagnosis (PM-US, 86.4% (95% CI, 77.7-92.0%) vs PM-MRI, 88.6% (95% CI, 80.3-93.7%)) or in the sensitivities and specificities for individual anatomical regions. There were more non-diagnostic PM-US than PM-MRI examinations for the brain (22.8% vs 3.7%) and heart (14.7% vs 5.1%). If an 'imaging-only' autopsy had been performed, PM-US would have achieved the same diagnosis as 1.5-T PM-MRI in 86.8% (95% CI, 80.0-91.5%) of cases, with the highest rates of agreement being for spine (99.3% (95% CI, 95.9-99.9%)) and cardiac (97.3% (95% CI, 92.4-99.1%)) findings and the lowest being for brain diagnoses (85.2% (95% CI, 76.9-90.8%)). CONCLUSION: Although there were fewer non-diagnostic cases using PM-MRI than for PM-US, the high concordance rate for overall diagnosis suggests that PM-US could be used for triaging cases when PM-MRI access is limited or unavailable. © 2020 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.

The significance of internal calcifications on perinatal post-mortem radiographs.

AIM: To determine whether the presence of internal calcifications on perinatal post-mortem skeletal surveys (PMSS) are associated with certain diagnoses of fetal loss. METHODS AND MATERIALS: A 6-month retrospective, single-centre, cohort study was conducted on PMSS performed for perinatal death assessment. One reader re-reviewed all PMSS images for the presence and location of internal calcifications, and noted whether these were included within the original radiology report. Findings at autopsy were then reviewed independently by a second researcher and cause of fetal loss or main diagnosis recorded. Chi-squared tests were conducted to identify differences between those with and without internal calcifications at PMSS. RESULTS: Two hundred and thirty perinatal deaths (mean gestational age 18 weeks; average 12-35 weeks) were included in the study, of which 42 (18.3%) demonstrated intra-abdominal calcifications, and 16/42 (38.1%) were mentioned in the radiology reports. Most calcifications were found to be within the lumen of the gastrointestinal tract, and in the left upper quadrant of the abdomen. There was no statistical difference between identifiable causes for fetal loss at autopsy in cases with and without calcification at PMSS (59.5% versus 58.5% respectively, p=0.904). Nevertheless, where calcification and a cause for fetal loss were found, the aetiology was more likely to be due a fetal rather than placental issue. CONCLUSION: The presence of internal calcifications on PMSS was not associated with an increased likelihood of explainable fetal loss or particular diagnosis at autopsy.

Feasibility of INTACT (INcisionless TArgeted Core Tissue) biopsy procedure for perinatal autopsy.

OBJECTIVES: To determine the feasibility and tissue yield of a perinatal incisionless ultrasound-guided biopsy procedure, the INcisionless Targeted Core Tissue (INTACT) technique, in the context of minimally invasive autopsy. METHODS: Cases of perinatal death in which the parents consented for minimally invasive autopsy underwent postmortem magnetic resonance imaging and an INTACT biopsy procedure, defined as needle biopsy of organs via the umbilical cord, performed under ultrasound guidance. In each case, three cores of tissue were obtained from seven target organs (both lungs, both kidneys, heart, spleen and liver). Biopsy success was predefined as an adequate volume of the intended target organ for pathological analysis, as judged by a pathologist blinded to the case and biopsy procedure. RESULTS: Thirty fetuses underwent organ sampling. Mean gestational age was 30 weeks (range, 18-40 weeks) and mean delivery-to-biopsy interval was 12 days (range, 6-22 days). The overall biopsy success rate was 153/201 (76.1%) samples, with the success rates in individual organs being highest for the heart and lungs (93% and 91%, respectively) and lowest for the spleen (11%). Excluding splenic samples, the biopsy success rate was 150/173 (86.7%). Histological abnormalities were found in 4/201 (2%) samples, all of which occurred in the lungs and kidneys of a fetus with pulmonary hypoplasia and multicystic kidney disease. CONCLUSIONS: Incisionless ultrasound-guided organ biopsy using the INTACT procedure is feasible, with an overall biopsy success rate of over 75%. This novel technique offers the ideal combination of an imaging-led autopsy with organ sampling for parents who decline the conventional invasive approach. © 2019 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of the International Society of Ultrasound in Obstetrics and Gynecology.

Finite element modelling of the developing infant femur using paired CT and MRI scans.

Bone finite element (FE) studies based on infant post-mortem computed tomography (CT) examinations are being developed to provide quantitative information to assist the differentiation between accidental and inflicted injury, and unsuspected underlying disease. As the growing skeleton contains non-ossified cartilaginous regions at the epiphyses, which are not well characterised on CT examinations, it is difficult to evaluate the mechanical behaviour of the developing whole bone. This study made use of paired paediatric post mortem femoral CT and magnetic resonance imaging (MRI) examinations at two different stages of development (4 and 7 months) to provide anatomical and constitutive information for both hard and soft tissues. The work aimed to evaluate the effect of epiphyseal ossification on the propensity to shaft fractures in infants. The outcomes suggest that the failure load of the femoral diaphysis in the models incorporating the non-ossified epiphysis is within the range of bone-only FE models. There may however be an effect on the metaphysis. Confirmation of these findings is required in a larger cohort of children.

Automated data extraction and report analysis in computer-aided radiology audit: practice implications from post-mortem paediatric imaging.

AIM: To determine local departmental adherence to the paediatric post-mortem magnetic resonance imaging (MRI) protocols, using a customised automated computational approach. MATERIALS AND METHODS: A retrospective review of 460 whole-body post-mortem MRI examinations performed at Great Ormond Street Hospital for Children over a 5.5-year period was assessed for adherence to a full or abbreviated imaging sequence protocol. A simple computer program was developed to batch process DICOM (digital imaging and communications in medicine) files, extracting imaging sequence details, followed by natural language processing (NLP) of authorised reports to automate information extraction of diagnostic image quality. RESULTS: The program was able to extract study parameters from the entire dataset (approximately 80 GB of data) in a few hours, and retrieve information on diagnostic image quality using NLP with an overall diagnostic accuracy for data extraction of 96.7% (445/460, 95% confidence interval [CI]: 94.7-98%). The full imaging protocol was adhered to in 305/460 (66.3%) cases, and an abbreviated protocol in 140/460 (30.4%) cases. Overall, 423/460 (91.9%) of studies were of diagnostic quality. These included 298/305 (97.7%) of the full protocol, 111/140 (79.3%) of the abbreviated protocol. In only five cases were the examinations non-diagnostic for all body systems, all of whom weighed <100 g (24.7-72 g) and imaged using the abbreviated protocol. CONCLUSION: The present study demonstrated a successful application of an automated approach for data collection for audit and quality assessment purposes using paediatric post-mortem imaging as a specific example. Re-audit of these data following change implementation will be straightforward now that the automated workflow is clearly established.

Minimally invasive perinatal and pediatric autopsy with laparoscopically assisted tissue sampling: feasibility and experience of the MinImAL procedure.

OBJECTIVE: Less invasive autopsy techniques in cases of fetal or infant death have good acceptability among parents, but the published sampling adequacy in needle biopsy studies is generally poor. Minimally Invasive Autopsy with Laparoscopically assisted sampling (MinImAL) has the potential to increase the diagnostic yield of less invasive autopsy by improving the quality and quantity of tissue samples obtained, whilst permitting visualization, extraction and examination of internal organs through a small incision. The aim of this study was to present the findings of our experience with the MinImAL procedure in cases of fetal, neonatal and pediatric death. METHODS: This was a retrospective analysis of 103 prospectively recruited unselected cases of fetal, neonatal or pediatric death that underwent the MinImAL procedure at a tertiary referral center over a 5-year period. Following preprocedure 1.5-T whole-body postmortem magnetic resonance imaging, MinImAL autopsy was performed. Procedure duration, sampling adequacy and cause of death were assessed. Chi-square analysis was used to compare the 'unexplained' rate of intrauterine deaths in the cohort with that in a previously published cohort of > 1000 cases of intrauterine death examined by standard autopsy. RESULTS: MinImAL autopsy was performed successfully in 97.8% (91/93) of the cases undergoing a complete procedure. There was a satisfactory rate of adequate histological sampling in most major organs; heart (100%, 91 cases), lung (100%, 91 cases), kidney (100%, 91 cases), liver (96.7%, 88 cases), spleen (94.5%, 86 cases), adrenal glands (89.0%, 81 cases), pancreas (82.4%, 75 cases) and thymus (56.0%, 51 cases). Procedure duration was similar to that of standard autopsy in a previously published cohort of intrauterine deaths. The unexplained rate in stillbirths and intrauterine fetal deaths that underwent MinImAL autopsy was not significantly different from that following standard autopsy. CONCLUSIONS: The MinImAL procedure provides good histological yield from major organs with minimal cosmetic damage and can be learned by an autopsy practitioner. The MinImAL procedure is an appropriate minimally invasive alternative for the investigation of perinatal and pediatric deaths in which consent to full autopsy is withheld, and may have applications in both high- and low/middle-income settings. Copyright © 2019 ISUOG. Published by John Wiley & Sons Ltd.

Postmortem examination of human fetuses: comparison of two-dimensional ultrasound with invasive autopsy.

OBJECTIVE: To assess the diagnostic accuracy of postmortem ultrasound performed by operators blinded to prenatal findings and to invasive autopsy results in fetuses at different gestational ages and to investigate the effect of various parameters on its diagnostic success. METHODS: We performed postmortem two-dimensional ultrasound examination, blinded to clinical details, on 163 fetuses at 13-42 weeks' gestation. Logistic regression analysis was used to investigate the effect of: (i) gestational age at postmortem ultrasound, (ii) presence of maceration and (iii) mode of death, on whether the exam succeeded or failed to reach a diagnosis. In 123 cases in which invasive autopsy was available, the diagnostic accuracy of ultrasound in detecting major organ abnormalities was evaluated, using invasive autopsy as the gold standard. RESULTS: For the fetal brain, postmortem ultrasound exam was non-diagnostic in significantly more fetuses with maceration (39.5%; 17/43) vs those without maceration (20.0%; 24/120) (P = 0.013). For the fetal thorax, the exam was non-diagnostic in 34.1% (15/44) of fetuses < 20 weeks of gestation and in 10.9% (13/119) of fetuses ≥ 20 weeks (P < 0.001). For the heart and abdominal organs, there was no association between non-diagnostic postmortem ultrasound and the variables tested. For fetuses < 20 weeks, specificity of postmortem ultrasound examination was 83.3% for detection of anomalies of the brain, 68.6% for the thorax and 77.4% for the heart. For fetuses ≥ 20 weeks, sensitivity and specificity were, respectively, 61.9% and 74.2% for detection of anomalies of the brain, 29.5% and 87.0% for the thorax and 65.0% and 83.1% for the heart. For the fetal abdominal organs, sensitivity was 60.7% and specificity 75.8%, and postmortem ultrasound was particularly useful for detection of abnormalities of the kidneys, irrespective of gestational age. CONCLUSION: Although maceration may lead to failure of postmortem ultrasound examination in some cases, this technique achieves diagnostically acceptable levels of accuracy for fetal brain and abdominal organs, compared with conventional autopsy. It may therefore play a role as a first-line examination before other virtual autopsy techniques are indicated. Copyright © 2017 ISUOG. Published by John Wiley & Sons Ltd.