RSS 1.0Imaging and Pathology
Sub-communities within this community
Recent Submissions
-
Explained deep learning framework for COVID-19 detection in volumetric CT images aligned with the British Society of Thoracic Imaging reporting guidance : a pilot studyIn March 2020, the British Society of Thoracic Imaging (BSTI) introduced a reporting guidance for COVID-19 detection to streamline standardised reporting and enhance agreement between radiologists. However, most current DL methods do not conform to this guidance. This study introduces a multi-class deep learning (DL) model to identify BSTI COVID-19 categories within CT volumes, classified as 'Classic', 'Probable', 'Indeterminate', or 'Non-COVID'. A total of 56 CT pseudoanonymised images were collected from patients with suspected COVID-19 and annotated by an experienced chest subspecialty radiologist following the BSTI guidance. We evaluated the performance of multiple DL-based models, including three-dimensional (3D) ResNet architectures, pre-trained on the Kinetics-700 video dataset. For better interpretability of the results, our approach incorporates a post-hoc visual explainability feature to highlight the areas of the image most indicative of the COVID-19 category. Our four-class classification DL framework achieves an overall accuracy of 75%. However, the model struggled to detect the 'Indeterminate' COVID-19 group, whose removal significantly improved the model's accuracy to 90%. The proposed explainable multi-classification DL model yields accurate detection of 'Classic', 'Probable', and 'Non-COVID' categories with poor detection ability for 'Indeterminate' COVID-19 cases. These findings are consistent with clinical studies that aimed at validating the BSTI reporting manually amongst consultant radiologists.
-
Evaluating explainable Artificial Intelligence (XAI) techniques in chest radiology imaging through a human-centered lensThe field of radiology imaging has experienced a remarkable increase in using of deep learning (DL) algorithms to support diagnostic and treatment decisions. This rise has led to the development of Explainable AI (XAI) system to improve the transparency and trust of complex DL methods. However, XAI systems face challenges in gaining acceptance within the healthcare sector, mainly due to technical hurdles in utilizing these systems in practice and the lack of human-centered evaluation/validation. In this study, we focus on visual XAI systems applied to DL-enabled diagnostic system in chest radiography. In particular, we conduct a user study to evaluate two prominent visual XAI techniques from the human perspective. To this end, we created two clinical scenarios for diagnosing pneumonia and COVID-19 using DL techniques applied to chest X-ray and CT scans. The achieved accuracy rates were 90% for pneumonia and 98% for COVID-19. Subsequently, we employed two well-known XAI methods, Grad-CAM (Gradient-weighted Class Activation Mapping) and LIME (Local Interpretable Model-agnostic Explanations), to generate visual explanations elucidating the AI decision-making process. The visual explainability results were shared through a user study, undergoing evaluation by medical professionals in terms of clinical relevance, coherency, and user trust. In general, participants expressed a positive perception of the use of XAI systems in chest radiography. However, there was a noticeable lack of awareness regarding their value and practical aspects. Regarding preferences, Grad-CAM showed superior performance over LIME in terms of coherency and trust, although concerns were raised about its clinical usability. Our findings highlight key user-driven explainability requirements, emphasizing the importance of multi-modal explainability and the necessity to increase awareness of XAI systems among medical practitioners. Inclusive design was also identified as a crucial need to ensure better alignment of these systems with user needs.
-
PDIA3 inhibits mitochondrial respiratory function in brain endothelial cells and C. elegans through STAT3 signaling and decreases survival after OGD.Background: Protein disulfide isomerase A3 (PDIA3, also named GRP58, ER-60, ERp57) is conserved across species and mediates protein folding in the endoplasmic reticulum. PDIA3 is, reportedly, a chaperone for STAT3. However, the role of PDIA3 in regulating mitochondrial bioenergetics and STAT3 phosphorylation at serine 727 (S727) has not been described. Methods: Mitochondrial respiration was compared in immortalized human cerebral microvascular cells (CMEC) wild type or null for PDIA3 and in whole organism C. Elegans WT or null for pdi-3 (worm homologue). Mitochondrial morphology and cell signaling pathways in PDIA3-/- and WT cells were assessed. PDIA3-/- cells were subjected to oxygen-glucose deprivation (OGD) to determine the effects of PDIA3 on cell survival after injury. Results: We show that PDIA3 gene deletion using CRISPR-Cas9 in cultured CMECs leads to an increase in mitochondrial bioenergetic function. In C. elegans, gene deletion or RNAi knockdown of pdi-3 also increased respiratory rates, confirming a conserved role for this gene in regulating mitochondrial bioenergetics. The PDIA3-/- bioenergetic phenotype was reversed by overexpression of WT PDIA3 in cultured PDIA3-/- CMECs. PDIA3-/- and siRNA knockdown caused an increase in phosphorylation of the S727 residue of STAT3, which is known to promote mitochondrial bioenergetic function. Increased respiration in PDIA3-/- CMECs was reversed by a STAT3 inhibitor. In PDIA3-/- CMECs, mitochondrial membrane potential and reactive oxygen species production, but not mitochondrial mass, was increased, suggesting an increased mitochondrial bioenergetic capacity. Finally, PDIA3-/- CMECs were more resistant to oxygen-glucose deprivation, while STAT3 inhibition reduced the protective effect. Conclusions: We have discovered a novel role for PDIA3 in suppressing mitochondrial bioenergetic function by inhibiting STAT3 S727 phosphorylation.
-
Rapid implementation of triaging system for assessment of breast referrals from primary care centres during the COVID-19 pandemic.Objective: The aim of this study was to establish a triaging system for assessment of breast referrals from primary care to ensure safe and effective breast services without compromising breast cancer management. Background: COVID-19 was officially declared a global pandemic on 11 March 2020, and with no effective treatment available, preventing spread has been paramount. Previously, all referrals from primary care were seen in the rapid-access breast clinic (RABC). Clinic appointments exposed patients and healthcare professionals to risk. Method: Initial triage during the lockdown was in line with national governing body guidance, rejected low risk referrals and streamed remaining patients through a telephone consultation to RABC or discharge. A modified triage pathway streamed all patients through virtual triage to RABC, telephone clinic or discharge with advice and guidance categories. Demographics, reasons for referral and outcomes data were collected and presented as median with range and frequency with percentages. Results: Initial triage (23 March-23 April 2020) found fewer referrals with a higher percentage of breast cancer diagnoses. Modified triage (22 June-17 July 2020) resulted in a 35.1% (99/282) reduction in RABC attendance. Overall cancer detection rate remained similar at 4.2% of all referrals pre-COVID (18/429) and 4.3% (12/282) during modified triage. After six months follow-up of patients not seen in RABC during the modified triage pathway, 18 patients were re-referred to RABC and none were diagnosed with cancer. Conclusion: A modified triage pathway has the potential to improve triage efficiency and prevent unnecessary visits during the COVID-19 pandemic. Further refinement of pathway is feasible in collaboration with primary care.
-
Safety of haem arginate in overdose: report of an uneventful fourfold accidental overdose.Safety of haem arginate in overdose: report of an uneventful fourfold accidental overdose
-
Treatment of chronic or relapsing COVID-19 in immunodeficiency.Background: Patients with some types of immunodeficiency can experience chronic or relapsing infection with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). This leads to morbidity and mortality, infection control challenges, and the risk of evolution of novel viral variants. The optimal treatment for chronic coronavirus disease 2019 (COVID-19) is unknown. Objective: Our aim was to characterize a cohort of patients with chronic or relapsing COVID-19 disease and record treatment response. Methods: We conducted a UK physician survey to collect data on underlying diagnosis and demographics, clinical features, and treatment response of immunodeficient patients with chronic (lasting ≥21 days) or relapsing (≥2 episodes) of COVID-19. Results: We identified 31 patients (median age 49 years). Their underlying immunodeficiency was most commonly characterized by antibody deficiency with absent or profoundly reduced peripheral B-cell levels; prior anti-CD20 therapy, and X-linked agammaglobulinemia. Their clinical features of COVID-19 were similar to those of the general population, but their median duration of symptomatic disease was 64 days (maximum 300 days) and individual patients experienced up to 5 episodes of illness. Remdesivir monotherapy (including when given for prolonged courses of ≤20 days) was associated with sustained viral clearance in 7 of 23 clinical episodes (30.4%), whereas the combination of remdesivir with convalescent plasma or anti-SARS-CoV-2 mAbs resulted in viral clearance in 13 of 14 episodes (92.8%). Patients receiving no therapy did not clear SARS-CoV-2. Conclusions: COVID-19 can present as a chronic or relapsing disease in patients with antibody deficiency. Remdesivir monotherapy is frequently associated with treatment failure, but the combination of remdesivir with antibody-based therapeutics holds promise.
-
A UK multicentre audit of the management of patients with primary hypercholesterolaemia or mixed dyslipidaemia with bempedoic acid against published lipid-lowering treatment targetsBackground: Bempedoic acid, an adenosine triphosphate citrate lyase inhibitor, was introduced to UK practice via a pre-reimbursement access scheme for adults with primary hypercholesterolaemia or mixed dyslipidaemia who are at high risk of cardiovascular disease, in whom statins are either not tolerated or contraindicated, who have not achieved target cholesterol, despite being on ezetimibe therapy, and do not qualify for PCSK9 inhibitor treatment. This retrospective multicentre audit aimed to evaluate the achievement of lipid-lowering targets with bempedoic acid in UK patients based on recommendations in the Joint British Societies (JBS) guidelines for the prevention of cardiovascular disease. Methods: Pseudo-anonymized medical record data for 221 adults treated with bempedoic acid as part of the UK scheme were entered into a bespoke data collection tool at four UK hospitals. Patient demographics, clinical characteristics, treatment pathways and lipid assessment results (against JBS lipid-lowering targets) were collected against pre-specified criteria. Results: Overall, 54% (99/184) of patients achieved the JBS2 audit standard (total cholesterol (TC) <5 mmol/L and low-density lipoprotein cholesterol (LDL-C) <3 mmol/L or ≥25% reduction in TC and ≥30% reduction in LDL-C) at 12 weeks post-initiation. At week 12, the mean absolute change in LDL-C was -1.0 mmol/L; the mean percentage reduction from baseline was 22.0%. Additionally, 52% (96/185) of patients had an LDL-C of <3 mmol/L and 10% (18/185) an LDL-C of <1.8 mmol/L at 12 weeks (as per JBS3). Conclusion: This audit highlights the role of bempedoic acid as part of combination therapy for a population with previously limited treatment options.
-
Take a break : should breaks be enforced during digital breast tomosynthesis reading sessions?Objectives: Digital breast tomosynthesis (DBT) can improve diagnostic accuracy compared to 2D mammography, but DBT reporting is time-consuming and potentially more fatiguing. Changes in diagnostic accuracy and subjective and objective fatigue were evaluated over a DBT reporting session, and the impact of taking a reporting break was assessed. Materials and methods: Forty-five National Health Service (NHS) mammography readers from 6 hospitals read a cancer-enriched set of 40 DBT cases whilst eye tracked in this prospective cohort study, from December 2020 to April 2022. Eye-blink metrics were assessed as objective fatigue measures. Twenty-one readers had a reporting break, 24 did not. Subjective fatigue questionnaires were completed before and after the session. Diagnostic accuracy and subjective and objective fatigue measures were compared between the cohorts using parametric and non-parametric significance testing. Results: Readers had on average 10 years post-training breast screening experience and took just under 2 h (105.8 min) to report all cases. Readers without a break reported greater levels of subjective fatigue (44% vs. 33%, p = 0.04), which related to greater objective fatigue: an increased average blink duration (296 ms vs. 286 ms, p < 0.001) and a reduced eye-opening velocity (76 mm/s vs. 82 mm/s, p < 0.001). Objective fatigue increased as the trial progressed for the no break cohort only (ps < 0.001). No difference was identified in diagnostic accuracy between the groups (accuracy: 87% vs. 87%, p = 0.92). Conclusions: Implementing a break during a 2-h DBT reporting session resulted in lower levels of subjective and objective fatigue. Breaks did not impact diagnostic accuracy, which may be related to the extensive experience of the readers. Clinical relevance statement: DBT is being incorporated into many mammography screening programmes. Recognising that reporting breaks are required when reading large volumes of DBT studies ensures this can be factored in when setting up reading sessions. Trial registration: Clinical trials registration number: NCT03733106 KEY POINTS: • Use of digital breast tomosynthesis (DBT) in breast screening programmes can cause significant reader fatigue. • The effectiveness of incorporating reading breaks into DBT reporting sessions, to reduce mammography reader fatigue, was investigated using eye tracking. • Integrating breaks into DBT reporting sessions reduced reader fatigue; however, diagnostic accuracy was unaffected.
-
Automation of the Whole-Blood Thiopurine -Methyltransferase (TPMT) Phenotyping Assay Using the Biomek NX and Biomek i5 Liquid-Handling Workstations.Assessment of thiopurine S-methyltransferase (TPMT) status is required before commencing thiopurine treatment to reduce the potential for adverse drug reactions. Our laboratory has provided a national TPMT phenotyping service since 2003. Our assay uses 6-thioguanine as substrate and detection of 6-methylthioguanine via high-performance liquid chromatography (HPLC) fluorescence. Here, we report the automation of this complex, labor-intensive, manual assay using the Biomek NXP and Biomek i5 liquid-handling workstations. We optimized assay performance and validated for precision, linearity, and lower limit of quantitation. We also compared results from the manual and automated methods. Primary sample mixing and aliquoting were performed on the Biomek NXP. On-board inversions (n = 10) replaced offline mixing. No carryover was observed. Eleven percent of primary sample tubes were incompatible with the Biomek NXP, and these were assayed manually. The Biomek i5 was used to automate the enzyme assay. Optimum vortex mixing was achieved at 2500 rpm for 60 s, and the temperature was set to 37.0 °C for the 60 min enzyme incubation. Intra- and inter-assay precision were excellent, with coefficients of variation (CVs) of ≤2.3% and ≤7.4%, respectively, for patient samples. Linearity was demonstrated up to 199 mIU/L (R2 = 0.992), with a lower limit of quantitation of 3.9 mIU/L. Correlation between the manual and automated methods was good (R2 = 0.979, n = 405), with results being interchangeable. We have successfully developed and validated a novel automated method for the TPMT phenotyping enzyme assay. The two methods are cost-neutral. Automation of other complex enzyme assays may be possible using this approach.
-
Apparent diffusion coefficient (ADC): A potential in vivo biological surrogate of the incidentally discovered bone lesions at 3T MRI.The mean ADC value (mean±SD) of all malignant tumors (including cartilaginous neoplasms) was [0.92 ± 0.40] × 10-3 mm2/s. This significantly differed from those of both primary benign tumors [1.14 ± 0.24] × 10-3 mm2/s, (p = 0.011), and all non-malignant lesions collectively [1.29 ± 0.44] × 10-3 mm2/s, (p < 0.001). Using mADC value of ≤ 1.1 × 10-3 mm2/s resulted in 86.1% sensitivity and 62.5% specificity for characterizing a lesion as a malignant. The inter-rater reliability was almost perfect (95% CI = 0.954-0.985).
-
A real-world evaluation of the diagnostic accuracy of radiologists using positive predictive values verified from deep learning and natural language processing chest algorithms deployed retrospectivelyObjectives: This diagnostic study assessed the accuracy of radiologists retrospectively, using the deep learning and natural language processing chest algorithms implemented in Clinical Review version 3.2 for: pneumothorax, rib fractures in digital chest X-ray radiographs (CXR); aortic aneurysm, pulmonary nodules, emphysema, and pulmonary embolism in CT images. Methods: The study design was double-blind (artificial intelligence [AI] algorithms and humans), retrospective, non-interventional, and at a single NHS Trust. Adult patients (≥18 years old) scheduled for CXR and CT were invited to enroll as participants through an opt-out process. Reports and images were de-identified, processed retrospectively, and AI-flagged discrepant findings were assigned to two lead radiologists, each blinded to patient identifiers and original radiologist. The radiologist's findings for each clinical condition were tallied as a verified discrepancy (true positive) or not (false positive). Results: The missed findings were: 0.02% rib fractures, 0.51% aortic aneurysm, 0.32% pulmonary nodules, 0.92% emphysema, and 0.28% pulmonary embolism. The positive predictive values (PPVs) were: pneumothorax (0%), rib fractures (5.6%), aortic dilatation (43.2%), pulmonary emphysema (46.0%), pulmonary embolus (11.5%), and pulmonary nodules (9.2%). The PPV for pneumothorax was nil owing to lack of available studies that were analysed for outpatient activity. Conclusions: The number of missed findings was far less than generally predicted. The chest algorithms deployed retrospectively were a useful quality tool and AI augmented the radiologists' workflow. Advances in knowledge: The diagnostic accuracy of our radiologists generated missed findings of 0.02% for rib fractures CXR, 0.51% for aortic dilatation, 0.32% for pulmonary nodule, 0.92% for pulmonary emphysema, and 0.28% for pulmonary embolism for CT studies, all retrospectively evaluated with AI used as a quality tool to flag potential missed findings. It is important to account for prevalence of these chest conditions in clinical context and use appropriate clinical thresholds for decision-making, not relying solely on AI.
-
How long does it take to read a mammogram? Investigating the reading time of digital breast tomosynthesis and digital mammographyPurpose: To analyse digital breast tomosynthesis (DBT) reading times in the screening setting, compared to 2D full-field digital mammography (FFDM), and investigate the impact of reader experience and professional group on interpretation times. Method: Reading time data were recorded in the PROSPECTS Trial, a prospective randomised trial comparing DBT plus FFDM or synthetic 2D mammography (S2D) to FFDM alone, in the National Health Service (NHS) breast screening programme, from January 2019-February 2023. Time to read DBT+FFDM or DBT+S2D and FFDM alone was calculated per case and reading times were compared between modalities using dependent T-tests. Reading times were compared between readers from different professional groups (radiologists and radiographer readers) and experience levels using independent T-tests. The learning curve effect of using DBT in screening on reading time was investigated using a Kruskal-Wallis test. Results: Forty-eight readers interpreted 1,242 FFDM batches (34,210 FFDM cases) and 973 DBT batches (13,983 DBT cases). DBT reading time was doubled compared to FFDM (2.09 ± 0.64 min vs. 0.98 ± 0.30 min; p < 0.001), and DBT+S2D reading was longer than DBT + FFDM (2.24 ± 0.62 min vs. 2.04 ± 0.46 min; p = 0.006). No difference was identified in reading time between radiologists and radiographers (2.06 ± 0.71 min vs. 2.14 ± 0.46 min, respectively; p = 0.71). Readers with five or more years of experience reading DBT were quicker than those with less experience (1.86 ± 0.56 min vs. 2.37 ± 0.65 min; p = 0.008), and DBT reading time decreased after less than 9 months accrued screening experience (p = 0.01). Conclusions: DBT reading times were double those of FFDM in the screening setting, but there was a short learning curve effect with readers showing significant improvements in reading times within the first nine months of DBT experience.
-
British Nuclear Medicine Society SeHCAT guidelinesNo abstract available
-
Using VARSKIN+v1.2 to estimate dose from direct skin contamination with radionuclides 223Ra, 212Pb and 225Ac; considerations for Nuclear Medicine staff and associated Personal Protective Equipment (PPE)Objective: To calculate depth-weighted doses for 223Ra, 212Pb and 225Ac for the skin sites of trunk, arms/legs, face, wrist, back of hand, fingertip, back and side of fingers using VARSKIN+v1.2. Methods: Published depth distribution histograms of the basal cells were used with dose averaging in VARSKIN+v1.2. A density correction factor was applied for the 1 g/cc within VARSKIN. Results were compared to the regulatory 70 µm depth and to average depth values for the skin sites. Results: 223Ra has no alpha component at the regulatory 70 µm. This dose is exceeded by the depth-weighted dose rates for all sites (except the fingertip) with factors ×74 (back of finger) to x3600 (trunk). 212Pb and 225Ac have alpha contributions at 70 µm. . For 212Pb, this dose value is greater by over ×2 than the depth-weighted dose rate for the wrist, back of hand, and finger sites, and underestimates dose rates for the other sites. For 225Ac, the 70µm dose rate is exceeded by the depth-weighted dose rates for the trunk, face, arms/legs by factors of ×4-10. Using fixed depth values, the depth-weighted dose rates are larger for all sites except the fingertip. The skin dose is also calculated for biological half-lives of 1, 3 and 6 h. Using the depth-weighted dose rates and a 3 h biological half-life, the activity for 500 mSv is in the range 9-177 Bq for the trunk, face, arms/legs, wrist and hand for all three radionuclides. Conclusion: For alpha-emitting radionuclides a depth-weighted calculation gives more representative dose values. The very low activity values for 500 mSv skin dose to be exceeded have implications for appropriate staff PPE and training.
-
A prospective surveillance study to determine the prevalence of 16S rRNA methyltransferase-producing Gram-negative bacteria in the UK.Objectives: To determine the prevalence of 16S rRNA methyltransferase- (16S RMTase-) producing Gram-negative bacteria in patients in the UK and to identify potential risk factors for their acquisition. Methods: A 6 month prospective surveillance study was conducted from 1 May to 31 October 2016, wherein 14 hospital laboratories submitted Acinetobacter baumannii, Enterobacterales and Pseudomonas aeruginosa isolates that displayed high-level amikacin resistance according to their testing methods, e.g. no zone of inhibition with amikacin discs. Isolates were linked to patient travel history, medical care abroad, and previous antibiotic exposure using a surveillance questionnaire. In the reference laboratory, isolates confirmed to grow on Mueller-Hinton agar supplemented with 256 mg/L amikacin were screened by PCR for 16S RMTase genes armA, rmtA-rmtH and npmA, and carbapenemase genes (blaKPC, blaNDM, blaOXA-48-like and blaVIM). STs and total antibiotic resistance gene complement were determined via WGS. Prevalence was determined using denominators for each bacterial species provided by participating hospital laboratories. Results: Eighty-four isolates (44.7%), among 188 submitted isolates, exhibited high-level amikacin resistance (MIC >256 mg/L), and 79 (94.0%) of these harboured 16S RMTase genes. armA (54.4%, 43/79) was the most common, followed by rmtB (17.7%, 14/79), rmtF (13.9%, 11/79), rmtC (12.7%, 10/79) and armA + rmtF (1.3%, 1/79). The overall period prevalence of 16S RMTase-producing Gram-negative bacteria was 0.1% (79/71 063). Potential risk factors identified through multivariate statistical analysis included being male and polymyxin use. Conclusions: The UK prevalence of 16S RMTase-producing Gram-negative bacteria is low, but continued surveillance is needed to monitor their spread and inform intervention strategies.
-
British Nuclear Medicine Society clinical guidelines for gastric emptyNo abstract available.
-
Audit of adequacy of the large joints magnetic resonance imagingNo abstract available
-
Whole-genome sequencing enhances existing pathogen and antimicrobial-resistance surveillance schemes within a neonatal unit.In some neonatal units, the screening of isolates for antimicrobial-resistant organisms is a matter of routine, with theoretical benefits including the prevention or early detection of outbreaks. This study sought to use whole-genome sequencing (WGS) retrospectively to characterize the genomic epidemiology of Gram-negative organisms obtained from a screening programme in a 32-bed unit providing intensive, high-dependency and special care at City Hospital, Birmingham, UK, identifying occult transmission events and clinically important antimicrobial-resistance (AMR) genes. WGS was performed for 155 isolates collected from rectal and umbilical screening swabs over a 2 month period from 44 individual neonates. Genomic epidemiological analysis showed possible transmission events involving Escherichia coli, Enterobacter cloacae, Klebsiella oxytoca and Klebsiella pneumoniae not detected by routine screening, with eight putative clusters involving different individuals identified. Within phylogenetic clusters, the relatedness of organisms - as determined by the abundance of SNPs - varied widely, indicating that a variety of transmission routes may be implicated. While clinically important AMR genes were not present in the putative transmission clusters, our observation of suspected interspecies horizontal transfer of blaCTX-M-15 within individuals highlights the potential for their spread between organisms as well as individuals in this environment, with implications for surveillance. Our data show that WGS may reveal occult Gram-negative transmission events, demonstrating the potential of sequencing-based surveillance systems for nosocomial pathogens. Challenges remain in understanding how to utilize WGS surveillance to maximum effect in real-world settings.
-
Optimising cylinder model dimensions for VARSKIN to simulate a droplet of radionuclide skin contamination using Geant4 Monte Carlo codeAim: VARSKIN provides a convenient way of calculating skin dose from predefined geometries but the models are limited to concentric shapes such as discs, cylinders and point sources. The aim of this article is to use the Geant4 Monte Carlo code to independently compare the cylindrical geometries available in VARSKIN to more realistic droplet models obtained from photography. It may then be possible to recommend an appropriate cylinder model that can be used to represent a droplet within acceptable accuracy. Method: Geant4 Monte Carlo code was used to model various droplets of radioactive liquid on the skin based on photographs. The dose rates were then calculated to the sensitive basal layer 70 µm beneath the surface for three droplet volumes (10, 30 and 50 µl) and 26 radionuclides. The dose rates from the cylinder models were then compared against the dose rates from the 'true' droplet models. Results: The table gives the optimum cylinder dimensions that best approximate a true droplet shape for each volume. The mean bias and 95% confidence interval (CI) from the true droplet model are also quoted. Conclusion: The evidence from the Monte Carlo data suggests that different droplet volumes require different cylinder aspect ratios to approximate the true droplet shape. Using the cylinder dimensions in the table in software packages such as VARSKIN, dose rates from radioactive skin contamination are expected to be within ± 7.4% of a 'true' droplet model at 95% CI.
-
Breast cancer metastasis to thymusDespite advancements in breast cancer management, metastatic disease remains a challenge. Thymic metastasis is an infrequent site of involvement. We intend to report a rare case of a woman in her 40s who presented with bilateral nipple discharge for 2 months. A triple assessment confirmed left breast carcinoma. A staging CT scan and positron emission tomography scan revealed an anterior mediastinal mass of thymic origin, and histology confirmed metastatic breast cancer. The patient made a good recovery following surgical and oncological treatment.