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.

Incidental findings in medical imaging.

The success of medical imaging as a diagnostic tool has resulted in a continuing increase in its use. Technological advances mean that images are now acquired at higher resolution and in greater volumes than ever before. This has led to an increase in the detection of findings which do not appear to be related to the primary purpose of the examination and have been termed "incidental". Many of these will be harmless but some will carry significant implications for the patient's health. Determining which of these findings are significant and which may be safely disregarded is an increasing problem in radiology practice. Radiologists should familiarise themselves with the more common incidental findings in order to make the best possible estimation of their importance in each case and to allow them to make appropriate recommendations for further investigation where this is indicated. The decision to advise further investigation carries implications for the patient and the service as a whole and requires careful consideration.

Diagnostics: a major priority for the NHS.

Diagnostic capacity in the NHS in England was much lower than that in many other developed countries before the COVID-19 pandemic. The relative lack of diagnostic equipment and workforce is now hampering recovery from the pandemic. In response to this, a major programme of work is now underway to improve access to a wide range of diagnostic tests. Establishment of community diagnostic centres is a key component of this programme.

The angiographic anatomy of the small arteries and their collaterals in colorectal resections: some insights into anastomotic perfusion.

OBJECTIVE: The aim of this study was to examine by screening angiography the anatomy of the small arteries and their collaterals in colorectal resections in order to identify factors that might be implicated in anastomotic leak. SUMMARY BACKGROUND DATA: Anastomotic leak is more frequent following low anterior resection. Vascular compromise is frequently implicated but poorly understood as a mechanism. METHODS: High definition screening angiography was performed on 17 colorectal resection specimens. RESULTS: (1) The small arteries of the colon (the vasa recta that arise from the marginal artery) show variability in their spacing and in their collaterals based on their anatomical positions. At the splenic flexure and the proximal and mid descending colons, the vasa recta are spaced 2-cm apart and have few collaterals. At the right, transverse, distal descending and sigmoid colons, the vasa recta are spaced <1 cm apart and have more extensive collaterals. (2) The small arteries of the rectum are spaced <1 cm apart and also show variability in their collaterals based on their anatomical level. In the mid-to-upper rectum there are good collaterals between the small arteries within the mesorectum based upon the bifurcation of the superior rectal artery and its main branches. In the lower rectum, however, there are only a few and very variable intramural collaterals between the small arteries. CONCLUSIONS: Based on these findings, unrecognized disruption of small artery collaterals during colorectal resection might be implicated in anastomotic leak and in particular might explain the higher leak rate in low anterior resection.

Error in radiology-where are we now?

Error is inherent in radiological practice. Our awareness of the extent of this and the reasons behind it has increased in recent times. Our next step must be the development of a shared understanding with our patients of the limitations as well as the huge benefits of medical imaging.

Metastatic prostate cancer masquerading clinically and radiologically as a primary caecal carcinoma.

BACKGROUND: Prostatic carcinoma is the second most common cause of cancer-related deaths in males in the West. Approximately 20% of patients present with metastatic disease. We describe the case of a patient with metastatic prostate cancer to the bowel presenting clinically and radiologically as a primary caecal cancer. CASE PRESENTATION: A 72 year-old man presented with abdominal discomfort and a clinically palpable caecal mass and a firm nodule on his thigh, the latter behaving clinically and radiologically as a lipoma. Computed tomographic (CT) scan showed a luminally protuberant caecal mass with regional nodal involvement. The patient was being treated (Zoladex) for prostatic cancer diagnosed 6 years previously and was known to have bony metastases. On admission his PSA was 245.4 nmol/ml. The patient underwent a right hemicolectomy. Histology showed a poorly differentiated adenocarcinoma which was PSA positive, confirming metastatic prostatic adenocarcinoma to the caecum. The patient underwent adjuvant chemotherapy and is free from recurrence a year later. CONCLUSION: Metastasis of prostatic carcinoma to the bowel is a very rare occurrence and presents a challenging diagnosis. The diagnosis is supported by immunohistochemistry for PSA. The treatment for metastatic prostate cancer is mainly palliative.

Stomal varices: a rare cause of stomal hemorrhage. A report of three cases.

Stomal varices secondary to portal hypertension are a rare but potentially fatal cause of hemorrhage. Management, determined by the site of the bleeding, centers on preventing additional bleeds and may include providing local pressure, applying silver nitrate, injection sclerotherapy, suture ligation of the bleeding point, and/or the placement of transjugular intrahepatic portosystemic shunts and refashioning the stoma. Two patients (60- and 69-year-old women) had panproctocolectomy for inflammatory bowel disease and presented at the authors' hospital with bleeding from the ileostomy 1 and 19 years, respectively, following the creation of their stomas. A third patient (a 72-year-old man) bled from an end colostomy following an abdominoperineal resection for Duke's C rectal adenocarcinoma performed 3 years previous. All three patients had recurrent admissions for stomal bleeding and stomal varices secondary to portal hypertension and were initially treated with local measures (pressure, silver nitrate, and suture ligation). Two had undergone revision of their stomas prior to current treatment. One patient responded to local treatment but later died due to liver failure, one stopped bleeding after transjugular portosystemic shunt placement, and one died from metastatic cancer. Clinicians should maintain a high index of suspicion of stomal varices in patients with underlying liver disease who present with recurrent stomal bleeds and provide appropriate treatment to stop active bleeding and reduce portal venous pressure.

Impact of detecting potentially serious incidental findings during multi-modal imaging.

Background: There are limited data on the impact of feedback of incidental findings (IFs) from research imaging.  We evaluated the impact of UK Biobank's protocol for handling potentially serious IFs in a multi-modal imaging study of 100,000 participants (radiographer 'flagging' with radiologist confirmation of potentially serious IFs) compared with systematic radiologist review of all images. Methods: Brain, cardiac and body magnetic resonance, and dual-energy x-ray absorptiometry scans from the first 1000 imaged UK Biobank participants were independently assessed for potentially serious IFs using both protocols. We surveyed participants with potentially serious IFs and their GPs up to six months after imaging to determine subsequent clinical assessments, final diagnoses, emotional, financial and work or activity impacts. Results: Compared to systematic radiologist review, radiographer flagging resulted in substantially fewer participants with potentially serious IFs (179/1000 [17.9%] versus 18/1000 [1.8%]) and a higher proportion with serious final diagnoses (21/179 [11.7%] versus 5/18 [27.8%]). Radiographer flagging missed 16/21 serious final diagnoses (i.e., false negatives), while systematic radiologist review generated large numbers of non-serious final diagnoses (158/179) (i.e., false positives). Almost all (90%) participants had further clinical assessment (including invasive procedures in similar numbers with serious and non-serious final diagnoses [11 and 12 respectively]), with additional impact on emotional wellbeing (16.9%), finances (8.9%), and work or activities (5.6%). Conclusions: Compared with systematic radiologist review, radiographer flagging missed some serious diagnoses, but avoided adverse impacts for many participants with non-serious diagnoses. While systematic radiologist review may benefit some participants, UK Biobank's responsibility to avoid both unnecessary harm to larger numbers of participants and burdening of publicly-funded health services suggests that radiographer flagging is a justifiable approach in the UK Biobank imaging study. The potential scale of non-serious final diagnoses raises questions relating to handling IFs in other settings, such as commercial and public health screening.