Evaluation of the impact of artificial intelligence-assisted image interpretation on the diagnostic performance of clinicians in identifying pneumothoraces on plain chest X-ray: a multi-case multi-reader study.

BACKGROUND: Artificial intelligence (AI)-assisted image interpretation is a fast-developing area of clinical innovation. Most research to date has focused on the performance of AI-assisted algorithms in comparison with that of radiologists rather than evaluating the algorithms' impact on the clinicians who often undertake initial image interpretation in routine clinical practice. This study assessed the impact of AI-assisted image interpretation on the diagnostic performance of frontline acute care clinicians for the detection of pneumothoraces (PTX). METHODS: A multicentre blinded multi-case multi-reader study was conducted between October 2021 and January 2022. The online study recruited 18 clinician readers from six different clinical specialties, with differing levels of seniority, across four English hospitals. The study included 395 plain CXR images, 189 positive for PTX and 206 negative. The reference standard was the consensus opinion of two thoracic radiologists with a third acting as arbitrator. General Electric Healthcare Critical Care Suite (GEHC CCS) PTX algorithm was applied to the final dataset. Readers individually interpreted the dataset without AI assistance, recording the presence or absence of a PTX and a confidence rating. Following a 'washout' period, this process was repeated including the AI output. RESULTS: Analysis of the performance of the algorithm for detecting or ruling out a PTX revealed an overall AUROC of 0.939. Overall reader sensitivity increased by 11.4% (95% CI 4.8, 18.0, p=0.002) from 66.8% (95% CI 57.3, 76.2) unaided to 78.1% aided (95% CI 72.2, 84.0, p=0.002), specificity 93.9% (95% CI 90.9, 97.0) without AI to 95.8% (95% CI 93.7, 97.9, p=0.247). The junior reader subgroup showed the largest improvement at 21.7% (95% CI 10.9, 32.6), increasing from 56.0% (95% CI 37.7, 74.3) to 77.7% (95% CI 65.8, 89.7, p<0.01). CONCLUSION: The study indicates that AI-assisted image interpretation significantly enhances the diagnostic accuracy of clinicians in detecting PTX, particularly benefiting less experienced practitioners. While overall interpretation time remained unchanged, the use of AI improved diagnostic confidence and sensitivity, especially among junior clinicians. These findings underscore the potential of AI to support less skilled clinicians in acute care settings.

Staging by Thoracoscopy in potentially radically treatable Lung Cancer associated with Minimal Pleural Effusion (STRATIFY): protocol of a prospective, multicentre, observational study.

INTRODUCTION: Recurrence rate following radical therapy for lung cancer remains high, potentially reflecting occult metastatic disease, and better staging tools are required. Minimal pleural effusion (mini-PE) is associated with particularly high recurrence risk and is defined as an ipsilateral pleural collection (<1/3 hemithorax on chest radiograph), which is either too small to safely aspirate fluid for cytology using a needle, or from which fluid cytology is negative. Thoracoscopy (local anaesthetic thoracoscopy (LAT) or video-assisted thoracoscopic surgery (VATS)) is the gold-standard diagnostic test for pleural malignancy in patients with larger symptomatic effusions. Staging by Thoracoscopy in potentially radically treatable Lung Cancer associated with Minimal Pleural Effusion (STRATIFY) will prospectively evaluate thoracoscopic staging in lung cancer associated-mini-PE for the first time. METHODS AND ANALYSIS: STRATIFY is a prospective multicentre observational study. Recruitment opened in January 2020. The primary objective is to determine the prevalence of detectable occult pleural metastases (OPM). Secondary objectives include assessment of technical feasibility and safety, and the impact of thoracoscopy results on treatment plans, overall survival and recurrence free survival. Inclusion criteria are (1) suspected/confirmed stages I-III lung cancer, (2) mini-PE, (3) Performance Status 0-2 (4), radical treatment feasible if OPM excluded, (5) ≥16 years old and (6) informed consent. Exclusion criteria are any metastatic disease or contraindication to the chosen thoracoscopy method (LAT/VATS). All patients have LAT or VATS within 7 (±5) days of registration, with results returned to lung cancer teams for treatment planning. Following an interim analysis, the sample size was reduced from 96 to 50, based on a lower-than-expected OPM rate. An MRI substudy was removed in November 2022 due to pandemic-related site setup/recruitment delays. These also necessitated a no-cost recruitment extension until October 2023. ETHICS AND DISSEMINATION: Protocol approved by the West of Scotland Research Ethics Committee (Ref: 19/WS/0093). Results will be published in peer-reviewed journals and presented at international meetings. TRIAL REGISTRATION NUMBER: ISRCTN13584097.

Fully automated volumetric measurement of malignant pleural mesothelioma by deep learning AI: validation and comparison with modified RECIST response criteria.

BACKGROUND: In malignant pleural mesothelioma (MPM), complex tumour morphology results in inconsistent radiological response assessment. Promising volumetric methods require automation to be practical. We developed a fully automated Convolutional Neural Network (CNN) for this purpose, performed blinded validation and compared CNN and human response classification and survival prediction in patients treated with chemotherapy. METHODS: In a multicentre retrospective cohort study; 183 CT datasets were split into training and internal validation (123 datasets (80 fully annotated); 108 patients; 1 centre) and external validation (60 datasets (all fully annotated); 30 patients; 3 centres). Detailed manual annotations were used to train the CNN, which used two-dimensional U-Net architecture. CNN performance was evaluated using correlation, Bland-Altman and Dice agreement. Volumetric response/progression were defined as ≤30%/≥20% change and compared with modified Response Evaluation Criteria In Solid Tumours (mRECIST) by Cohen's kappa. Survival was assessed using Kaplan-Meier methodology. RESULTS: Human and artificial intelligence (AI) volumes were strongly correlated (validation set r=0.851, p<0.0001). Agreement was strong (validation set mean bias +31 cm3 (p=0.182), 95% limits 345 to +407 cm3). Infrequent AI segmentation errors (4/60 validation cases) were associated with fissural tumour, contralateral pleural thickening and adjacent atelectasis. Human and AI volumetric responses agreed in 20/30 (67%) validation cases κ=0.439 (0.178 to 0.700). AI and mRECIST agreed in 16/30 (55%) validation cases κ=0.284 (0.026 to 0.543). Higher baseline tumour volume was associated with shorter survival. CONCLUSION: We have developed and validated the first fully automated CNN for volumetric MPM segmentation. CNN performance may be further improved by enriching future training sets with morphologically challenging features. Volumetric response thresholds require further calibration in future studies.

A comparison between MRI and CT in the assessment of primary tumour volume in mesothelioma.

INTRODUCTION: Primary tumour staging in Malignant Pleural Mesothelioma (MPM) using Computed Tomography (CT) imaging is confounded by perception errors reflecting low spatial resolution between tumour and adjacent structures. Augmentation using perfusion CT is constrained by radiation dosage. In this study, we evaluated an alternative tumour staging method using perfusion-tuned Magnetic Resonance Imaging (MRI). METHODS: Consecutive patients with suspected MPM were recruited to a prospective observational study. All had MRI (T1-weighted, isotropic, contrast-enhanced 3-Tesla perfusion imaging) and CT (contrast-enhanced) pre-biopsy. Patients diagnosed with MPM underwent MRI and CT volumetry, with readers blinded to clinical data. MRI volumetry was semi-automated, using signal intensity limits from perfusion studies to grow tumour regions within a pleural volume. A similar CT method was not possible, therefore all visible tumour was manually segmented. MRI and CT volumes were compared (agreement, correlation, analysis time, reproducibility) and associations with survival examined using Cox regression. RESULTS: 58 patients were recruited and had MRI before biopsy. 31/58 were diagnosed with MPM and these scans were used for volumetry. Mean (SD) MRI and CT volumes were 370 cm3 and 302 cm3, respectively. MRI volumes were larger (average bias 61.9 cm3 (SD 116), 95 % limits (-165.5 - 289 cm3), moderately correlated with CT (r = 0.56, p = 0.002) and independently associated with survival (HR 4.03 (95 % CI 1.5-11.55), p = 0.006). CT volumes were not associated with survival, took longer to compute than MRI volumes (mean (SD) 151 (19) v 14 (2) minutes, p=<0.0001) and were less reproducible (inter-observer ICC 0.72 for CT, 0.96 for MRI). CONCLUSIONS: MRI and CT generate different tumour volumes in MPM. In this study, MRI volumes were larger and were independently associated with survival. MRI volumetry was quicker and more reproducible than CT.

Early Contrast Enhancement: A novel magnetic resonance imaging biomarker of pleural malignancy.

INTRODUCTION: Pleural Malignancy (PM) is often occult on subjective radiological assessment. We sought to define a novel, semi-objective Magnetic Resonance Imaging (MRI) biomarker of PM, targeted to increased tumour microvessel density (MVD) and applicable to minimal pleural thickening. MATERIALS AND METHODS: 60 consecutive patients with suspected PM underwent contrast-enhanced 3-T MRI then pleural biopsy. In 58/60, parietal pleura signal intensity (SI) was measured in multiple regions of interest (ROI) at multiple time-points, generating ROI SI/time curves and Mean SI gradient (MSIG: SI increment/time). The diagnostic performance of Early Contrast Enhancement (ECE; which was defined as a SI peak in at least one ROI at or before 4.5 min) was compared with subjective MRI and Computed Tomography (CT) morphology results. MSIG was correlated against tumour MVD (based on Factor VIII immunostain) in 31 patients with Mesothelioma. RESULTS: 71% (41/58) patients had PM. Pleural thickening was <10 mm in 49/58 (84%). ECE sensitivity was 83% (95% CI 61-94%), specificity 83% (95% CI 68-91%), positive predictive value 68% (95% CI 47-84%), negative predictive value 92% (78-97%). ECE performance was similar or superior to subjective CT and MRI. MSIG correlated with MVD (r = 0.4258, p = .02). DISCUSSION: ECE is a semi-objective, perfusion-based biomarker of PM, measurable in minimal pleural thickening. Further studies are warranted.

Changing trends in a decade of vascular radiology-the impact of technical developments of non-invasive techniques on vascular imaging.

OBJECTIVES: This review aims to establish the impact on conventional angiography and endovascular intervention of contrast-enhanced magnetic resonance angiography (CE-MRA) and contrast-enhanced computed tomography angiography (CE-CTA) on a background of evolving technology, changing clinical requirements and resulting clinical repercussions. METHODS: The angiographic and interventional caseload was prospectively recorded between 1997 and 2010, along with the CE-MRA and CE-CTA caseload. Waiting times and the marginal cost analyses for 2001 and 2009 were also prospectively established. RESULTS: Conventional diagnostic angiographies declined from a peak of 847 to 121 per year while endovascular interventions continue in similar numbers. CE-MRA increased from effectively none initially to 620 per year while CE-CTA has currently risen to 396 per year. Total diagnostic study numbers have increased but at reduced cost. Various influences are clear, including on-site modality availability, capability and accuracy along with impact of new therapies, research studies and adverse events. CONCLUSIONS: Vascular imaging has undergone a metamorphosis in little over a decade because of CE-MRA and CE-CTA. With waiting times significantly reduced since the start of the study and the cost-effectiveness of both CE-MRA and CE-CTA as primary diagnostic investigations established, further development of these services is inevitable. MAIN MESSAGES: • The availability of CE-MRA and CE-CTA has reduced the need for conventional angiography. • Both waiting times and the marginal cost analyses for CE-MRA and CE-CTA have reduced. • The impact of new therapies, research studies (e.g. ASTRAL) and adverse events is illustrated.

Transscaphoid perilunate dislocation--a tale of two carpals?

The authors describe two cases of perilunate dislocation with transscaphoid fractures both presenting following hyperextension injuries within 4 days of each other at a district general hospital. After initial management in accident and emergency, both patients underwent open reduction and internal fixation under general anaesthetic, with plaster cast immobilisation and careful clinical follow-up. Given the relative rarity of this injury, the significance of the radiological findings, the degree of associated bony and ligamentous damage and the need for prompt intervention may be overlooked. Thus, a summary of presentation and management of the condition is given to aid junior emergency staff in diagnosis and initial management.

Acute portal vein thrombosis: an important mimic of malignancy on FDG PET/CT.

The authors report a case of portal vein thrombosis, with no underlying malignant cause identified, which was initially detected by fludeoxyglucose positron emission tomography/CT (FDG PET/CT) and subsequently confirmed by both contrast enhanced CT and MRI. The multimodality imaging findings are outlined, the potential clinical implications discussed and note made of the possible FDG PET/CT mimics of this disorder.

Thrombosis of popliteal artery pseudoaneurysm by deployment of angioplasty balloon after total knee arthroplasty.

Vascular complications after total knee arthroplasty are rare, with incidence ranging between 0.03% and 0.2%. We describe a case of a popliteal artery false aneurysm after total knee arthroplasty successfully treated by deployment of an angioplasty balloon across the aneurysm neck resulting in thrombosis.