The Efficacy of a Named Entity Recognition AI Model for Identifying Incidental Pulmonary Nodules in CT Reports.

Purpose: This study evaluates the efficacy of a commercial medical Named Entity Recognition (NER) model combined with a post-processing protocol in identifying incidental pulmonary nodules from CT reports. Methods: We analyzed 9165 anonymized CT reports and classified them into 3 categories: no nodules, nodules present, and nodules >6 mm. For each report, a generic medical NER model annotated entities and their relations, which were then filtered through inclusion/exclusion criteria selected to identify pulmonary nodules. Ground truth was established by manual review. To better understand the relationship between model performance and nodule prevalence, a subset of the data was programmatically balanced to equalize the number of reports in each class category. Results: In the unbalanced subset of the data, the model achieved a sensitivity of 97%, specificity of 99%, and accuracy of 99% in detecting pulmonary nodules mentioned in the reports. For nodules >6 mm, sensitivity was 95%, specificity was 100%, and accuracy was 100%. In the balanced subset of the data, sensitivity was 99%, specificity 96%, and accuracy 97% for nodule detection; for larger nodules, sensitivity was 94%, specificity 99%, and accuracy 98%. Conclusions: The NER model demonstrated high sensitivity and specificity in detecting pulmonary nodules reported in CT scans, including those >6 mm which are potentially clinically significant. The results were consistent across both unbalanced and balanced datasets indicating that the model performance is independent of nodule prevalence. Implementing this technology in hospital systems could automate the identification of at-risk patients, ensuring timely follow-up and potentially reducing missed or late-stage cancer diagnoses.

The impact of performance feedback reports on physician ordering behavior in the use of computed tomography pulmonary angiography (CTPA).

PURPOSE: The increased utilization, and potential overutilization, of computed tomography pulmonary angiography (CTPA) is a well-recognized issue within emergency departments (EDs). The objective of this study is to determine the impact of performance feedback reports on CTPA ordering behavior among ED physicians. METHODS: We conducted a prospective study of the impact of individualized performance feedback reports on the ordering behavior of physicians working at two high-volume community EDs in Ontario, Canada. We generated individualized reports (or "Dashboards") for each ED physician containing detailed feedback and peer comparison for each physician's CTPA ordering. Our baseline pre-intervention period was January 1 to December 31, 2018, and our intervention period was January 1, 2019, to December 31, 2021. We tracked individual and group ordering behavior through the study period. Our primary outcomes are impact of feedback on (1) overall group ordering rate and (2) overall diagnostic yield. Secondary analysis was done to determine the impact of the intervention on those physicians with the highest CTPA utilization rate. RESULTS: There was no statistically significant difference in the diagnostic yield of the included physicians in either of the years of the intervention period. There was a statically significant increase in the utilization rate for CTPA from 2018 to 2020 and 2021 from 5.9 to 7.9 and 11.4 CTPAs per 1000 ED visits respectively (p < 0.5). CONCLUSION: Our study found no consistent significant impact of individualized feedback and peer comparison on physician ordering of CTPAs. This points to a potentially greater impact of environmental and institutional factors, as opposed to physician-targeted quality improvement measures, on physician ordering behavior.

Utilization of CT angiography of the head and neck in the era of endovascular therapy for acute ischemic stroke: a retrospective study.

PURPOSE: To describe the impact of a new institutional Code Stroke protocol on ordering volume of head and neck CT angiographies (CTA), and to determine the number and proportion of these studies that resulted in an endovascular or surgical intervention. METHODS: Clinical and administrative data was collected on all head and neck CTAs ordered within the ED at two high-volume community hospitals and an affiliated urgent care centre during the 6-year period between January 1, 2014, and December 31, 2019. Of those patients who underwent CTA, we identified those who were then transferred to a regional stroke centre for consideration of EVT and those who underwent carotid endarterectomy or stenting within 14 days. RESULTS: A total of 4719 CTAs were ordered during the 6-year period. There was nearly a tenfold rise in the yearly number of CTAs ordered per 10,000 ED visits, from 5.3 (in 2014) to 53.1 (in 2019). A total of 164 patients who underwent CTAs (3.5%) were ultimately transferred to a regional tertiary care centre, of whom 43 (0.9%) were transferred to a regional stroke centre for consideration of EVT. A total of 61 (1.3%) patients underwent a carotid intervention within 14 days. CONCLUSION: Little is known of the impacts on healthcare resources that have resulted from the system-wide changes made necessary by the widespread adoption of EVT. Our study shows that at our site, these system changes have resulted in large increases in CTA utilization with very small numbers of patients ultimately undergoing EVT or carotid intervention.

Determination of the Best Parameter for Defining the Hemodynamic Significance of an Iliac Artery Stenosis Detected on Computed Tomography Angiography.

PURPOSE: The purpose of this study was to determine the best parameter, derived from computed tomography angiography (CTA) for accurate prediction of a hemodynamically significant stenosis of the common or external iliac artery. METHODS: A retrospective keyword search was performed on the Radiology Information System at our tertiary academic medical centre. Reports from January 2008 to September 2013 were searched using the keywords iliac, stenosis, and pressure. Patients who had both and CTA and a pelvic angiogram with pressure measurements obtained across a potential stenosis were selected. Using 3D postprocessing software (TeraRecon, Foster City, CA), the CTAs were analysed for the following parameters of each lesion: minimum diameter of stenosis, minimum cross-sectional area of stenosis, percent narrowing of vessel diameter, and percent reduction in vessel area. The percent stenosis was calculated in reference to the outer diameter at the point of maximal narrowing and also in reference to a normal segment of vessel more distal to the stenosis. These parameters were then compared with the measured pressure gradient using receiver-operating characteristic analysis and the Mann-Whitney U test to determine which best predicted a significant stenosis, defined as a greater than 10% drop in systolic pressure across a lesion. RESULTS: One hundred and two stenoses in 83 patients (26 women, 57 men; 47-88 years old) were identified. Mean diameter of the stenosis was 2.8 mm for significant stenosis compared to 3.8 mm in nonsignificant stenoses (P = .005). Mean minimum area for significant stenoses was 11.8 mm(2) compared to 17.22 mm(2) for nonsignificant stenoses (P = .032) No other variables showed a significant difference between significant and nonsignificant stenoses. A minimum diameter of ≤4.0 mm at the level of a stenosis is 92% sensitive and 48% specific for predicting a hemodynamically significant iliac artery stenosis, with a positive predictive value of 88%. CONCLUSIONS: A simple measurement of the minimum diameter of an iliac artery at the level of stenosis is the best predictor of the hemodynamic significance of a stenosis in the common or external iliac artery.

The use of dynamic volumetric CT angiography (DV-CTA) for the characterization of endoleaks following fenestrated endovascular aortic aneurysm repair (f-EVAR).

Accurate endoleak classification is essential following fenestrated endovascular aneurysm repair (f-EVAR). Both endoleak type and exact source of endoleak have implications upon the urgency and complexity of future management strategies. Herein we report on a patient with a documented endoleak post-f-EVAR, in which the source of blood flow into the aneurysm sac could not be determined using conventional computed tomographic angiography. Consequently, dynamic volumetric computed tomographic angiography (DV-CTA) was employed, which clearly illustrated the site of origin of the endoleak. DV-CTA enables accurate endoleak characterization following f-EVAR, with excellent conspicuity of the source of blood flow into the aneurysm sac.

Effectiveness of stent-graft placement for salvage of dysfunctional arteriovenous hemodialysis fistulas.

PURPOSE: To determine the clinical effectiveness of stent-graft placement in dysfunctional autogenous hemodialysis fistulas. MATERIALS AND METHODS: Between September 2006 and June 2008, 24 Fluency Plus stent-grafts were inserted in 17 patients with failing autogenous hemodialysis fistulas (two radiocephalic, 12 brachiocephalic, and three brachiobasilic). Six fistulas were thrombosed at presentation. Indications for stent-graft insertion included 10 residual stenoses, six pseudoaneurysms, and one fistuloplasty-induced rupture. Fistula function before and after intervention was assessed with ultrasound dilution technique. RESULTS: The technical success rate was 100%, as were the anatomic and clinical success rates. Via Kaplan-Meier method, the primary access patency rates were 94.1% at 3 months (95% CI, 0.461-0.995) and 88.2% at 6 and 12 months (95% CI, 0.468-0.980). The primary lesion patency rate was 94.1% (95% CI, 0.650-0.992) at 3, 6, and 12 months. The secondary access patency rate had a point estimate of 100% at 3 months, and was 93.8% (95% CI, 0.632-0.991) at 6 and 12 months. Univariate analysis suggested that male sex and left-sided accesses were associated with a longer primary access patency rate (P = .02). The binary restenosis rate at 6 months was 5.9%. Mean transonic flow rates were 667.5 mL/min before intervention and 2,076.0 mL/min after intervention (P = .03, Wilcoxon signed-rank test). No patients were lost to follow-up. CONCLUSIONS: Stent-graft placement in dysfunctional autogenous hemodialysis fistulas is technically feasible and effective in preserving function and preventing access abandonment, with patency rates that exceed historical patency rates with angioplasty and/or uncovered stents.

Acute ischemic cholecystitis after transarterial chemoembolization of hepatocellular carcinoma: incidence and clinical outcome.

PURPOSE: To determine the incidence and clinical outcome of acute ischemic cholecystitis after transarterial chemoembolization (TACE) of hepatocellular carcinoma. MATERIALS AND METHODS: In this ethics board-approved study, a retrospective review of 355 TACE procedures performed in 246 patients during a 5-year period was performed. The review of postintervention computed tomography (CT) reports for findings indicative of acute cholecystitis identified 12 patients (4.9%). In these patients, all CT scans, laboratory results (white blood cell count, alkaline phosphatase level, total bilirubin level), and clinical reports were analyzed to assess imaging findings and outcomes at the following time points: before TACE, within the first week after the procedure, as well as 1 and 6 months post-TACE. RESULTS: In 11 of 12 cases, the dominant finding on CT was new gallbladder wall thickening of up to 12 mm, which developed within 24 hours in 10 patients and within the first month after TACE in 1 patient. Gallbladder wall thickening persisted in 1 patient for at least 6 months. Eleven of 12 patients showed deposition of Lipiodol in the gallbladder wall. In 1 patient, the dominant finding on CT was pericholecystic stranding that resolved on follow-up CT after 1 month. None of the patients demonstrated gas in the gallbladder wall, significant pericholecystic fluid, abdominal or liver abscesses. Blood work results revealed transient increase in white blood cell count, alkaline phosphatase level, and total bilirubin level, not different from that seen after TACE in patients without CT evidence of cholecystitis. Clinical reports documented transient right upper quadrant pain for a few days and up to 1 month in 1 case with eventual symptom relief. None of the cases required surgical or radiological intervention. All but 1 case demonstrated normal gallbladder wall thickness after 6 months. CONCLUSIONS: Acute ischemic cholecystitis is not an uncommon complication after TACE. However, it is self-limiting and does not seem to require any intervention or surgery.

Subclavian steal syndrome treated by kissing stenting of the subclavian and vertebral arteries.

The prevalence of subclavian artery (SA) stenosis is approximately 2%. The exact prevalence of extracranial vertebral artery (VA) stenosis is undetermined, with estimates ranging from 7% to 40%. Nearly 25% of ischaemic strokes involve the vertebrobasilar circulation, and arteriosclerotic disease and narrowing of the proximal VA may be the cause for up to one-fifth of these incidents. The bulk of SA stenoses occur proximally to the ostium of the VA. Vertebrobasilar ischaemia can be caused both by VA and SA stenosis. Surgical and endovascular approaches are potential treatment options for SA/VA stenosis. It has been demonstrated that endovascular intervention is considerably safer for this pathology, and with advances in device technology, angioplasty with stenting has become the preferred treatment option. We present the case of a 76-year-old man who presented with vertebrobasilar ischaemia from coexisting stenosis of the SA/VA which was treated by endovascular methods.

Preoperative Computed Tomographic Angiogram Measurement of Abdominal Muscles Is a Valuable Risk Assessment for Bulge Formation after Microsurgical Abdominal Free Flap Breast Reconstruction.

BACKGROUND: A major shortcoming associated with abdominal tissue breast reconstruction is long-term abdominal wall morbidity. Although abdominal muscle size on computed tomographic angiography has been correlated with morbidity following many abdominal operations, it has not been studied for breast reconstruction. Therefore, the authors evaluated the association between preoperative computed tomography angiography-derived measurements of abdominal core muscles and postoperative abdominal wall morbidity after abdominal tissue breast reconstruction. METHODS: In this retrospective matched case-control study of women who underwent microsurgical abdominal flap breast reconstruction at one institution between January 2011 and June 2015, the authors evaluated all cases of postoperative bulge/hernia, matched by type of abdominal flap and body mass index in a ratio of 1:2 to controls without bulge/hernia. The authors obtained morphometric measurements of abdominal core muscles on preoperative computed tomographic angiographs. Using univariable and multivariable logistic regressions, the authors examined the effects of clinical risk factors and computed tomographic angiography morphometric measurements on postoperative bulge/hernia formation. RESULTS: Of the 589 patients who underwent abdominal free flap breast reconstruction, symptomatic bulges/hernias were identified in 35 patients (5.9 percent). When compared to the 70 matched controls, multivariable analysis showed that decreased area of rectus abdominis muscle (OR, 0.18; p < 0.01) and increased inter-rectus abdominis distance (OR, 1.14; p < 0.01) on computed tomographic angiography were significant risk factors associated with postoperative bulge/hernia. CONCLUSION: Preoperative computed tomographic angiography allows objective measurements of the patient's abdominal muscle anatomy that provide valuable prognostic information on the risk of bulge/hernia formation following abdominally based microsurgical breast reconstruction. CLINICAL QUESTION/LEVEL OF EVIDENCE: Risk, III.

Incidentalomas associated with abdominal and pelvic CT angiograms for abdominal-based breast free flap reconstruction.

Computed tomography angiography (CTA) is routinely performed prior to breast reconstruction using deep inferior epigastric perforator (DIEP) flaps to provide better surgical planning and improve preoperative decision making. These investigations occasionally result in unexpected findings in otherwise asymptomatic women. Unexpected findings on imaging in a population of women with previous breast cancer or strong breast cancer risk factors can lead to undue stress and anxiety. The aim of this study is to determine the incidence of unexpected findings in preoperative CT angiograms and to correlate these with patient and breast cancer characteristics. A retrospective chart review from May 2008 to December 2012 was performed reviewing all patients who underwent DIEP flap breast reconstruction. Radiology reports of their preoperative CT angiograms, details of unexpected findings, patients' past medical and cancer history, additional radiological investigations, outcomes, and interventions were reviewed. In total, 360 patients met the inclusion criteria for the study. Sixty-four percent of the patients who underwent CTA imaging had incidental findings. Further imaging was suggested in 48% of this group. The most common incidentalomas were hepatic (47%), bone (24%), and renal (20%). "Incidentalomas" were associated with patients' underlying comorbidities (p = 0.001) and age (p = 0.01). "Radiographically suspicious incidentalomas" were associated with the underlying comorbidities (p = 0.001). The radiologists most commonly suggested investigation methods such as ultrasound (41%), another CT (28%), bone scan (21%), and magnetic resonance imaging (MRI; 14%). No incidentalomas were found to be malignant on further recommended investigation in this study and no breast cancer reconstruction was delayed as a result of the discovery of incidentalomas or their subsequent investigations. It is important to counsel patients of the possibility of incidental findings and the resulting sequelae prior to CTA investigations for breast cancer reconstruction and involve them in shared decision making if additional investigations are required.

Intra-vascular blood velocity and volumetric flow rate calculated from dynamic 4D CT angiography using a time of flight technique.

We examine a time of flight (TOF) approach for the analysis of contrast enhanced 4D volumetric CT angiography scans to derive and display blood velocity in arteries. Software was written to divide blood vessels into a series of cross sections and to track contrast bolus TOF along the central vessel axis, which was defined by a user, from 4D CT source data. Time density curves at each vessel cross section were fit with quadratic, Gaussian, and gamma variate functions to determine bolus time to peak (TTP). A straight line was used to plot TTP versus vessel path length for all three functions and the slope used to calculate intraluminal velocity. Software was validated in a simulated square channel and non-pulsatile flow phantom prior to the calculation of blood velocity in the major cerebral arteries of 8 normal patients. The TOF algorithm correctly calculates intra-luminal fluid velocity in eight flow conditions of the CT flow phantom where quadratic functions were used. Across all conditions, in phantoms and in vivo, the success of calculations depended strongly on having a sufficiently long path length to make measurements and avoiding venous contamination. Total blood flow into the brain was approximately 17 % of a normal 5 L cardiac output. The technique was explored in vivo in a patient with subclavian steal syndrome, in the pulmonary arteries and in the iliac artery from clinical 4D CT source data. Intravascular blood velocity and flow may be calculated from 4D CT angiography using a TOF approach.