Artificial intelligence and point-of-care ultrasound: Benefits, limitations, and implications for the future.

The utilization of artificial intelligence (AI) in medical imaging has become a rapidly growing field as a means to address contemporary demands and challenges of healthcare. Among the emerging applications of AI is point-of-care ultrasound (POCUS), in which the combination of these two technologies has garnered recent attention in research and clinical settings. In this Controversies paper, we will discuss the benefits, limitations, and future considerations of AI in POCUS for patients, clinicians, and healthcare systems.

Comparison of artificial intelligence versus real-time physician assessment of pulmonary edema with lung ultrasound.

BACKGROUND: Lung ultrasound can evaluate for pulmonary edema, but data suggest moderate inter-rater reliability among users. Artificial intelligence (AI) has been proposed as a model to increase the accuracy of B line interpretation. Early data suggest a benefit among more novice users, but data are limited among average residency-trained physicians. The objective of this study was to compare the accuracy of AI versus real-time physician assessment for B lines. METHODS: This was a prospective, observational study of adult Emergency Department patients presenting with suspected pulmonary edema. We excluded patients with active COVID-19 or interstitial lung disease. A physician performed thoracic ultrasound using the 12-zone technique. The physician recorded a video clip in each zone and provided an interpretation of positive (≥3 B lines or a wide, dense B line) or negative (<3 B lines and the absence of a wide, dense B line) for pulmonary edema based upon the real-time assessment. A research assistant then utilized the AI program to analyze the same saved clip to determine if it was positive versus negative for pulmonary edema. The physician sonographer was blinded to this assessment. The video clips were then reviewed independently by two expert physician sonographers (ultrasound leaders with >10,000 prior ultrasound image reviews) who were blinded to the AI and initial determinations. The experts reviewed all discordant values and reached consensus on whether the field (i.e., the area of lung between two adjacent ribs) was positive or negative using the same criteria as defined above, which served as the gold standard. RESULTS: 71 patients were included in the study (56.3% female; mean BMI: 33.4 [95% CI 30.6-36.2]), with 88.3% (752/852) of lung fields being of adequate quality for assessment. Overall, 36.1% of lung fields were positive for pulmonary edema. The physician was 96.7% (95% CI 93.8%-98.5%) sensitive and 79.1% (95% CI 75.1%-82.6%) specific. The AI software was 95.6% (95% CI 92.4%-97.7%) sensitive and 64.1% (95% CI 59.8%-68.5%) specific. CONCLUSION: Both the physician and AI software were highly sensitive, though the physician was more specific. Future research should identify which factors are associated with increased diagnostic accuracy.

Ultrasound for the diagnosis of necrotizing fasciitis: A systematic review of the literature.

BACKGROUND: Necrotizing fasciitis (NF) is a deadly disorder that can be challenging to diagnose on history and examination alone. Point-of-care ultrasound (POCUS) is widely available and has been increasingly used for diagnosing skin and soft tissue infections. We performed a systematic review to determine the accuracy of POCUS for diagnosing NF with subgroup analyses of the accuracy of specific POCUS examination components. METHODS: PubMed, Scopus, CINAHL, LILACS, the Cochrane databases, Google Scholar, and bibliographies of selected articles were assessed for all retrospective, prospective, and randomized control trials evaluating the accuracy of POCUS for diagnosing NF. Data were dual extracted into a predefined worksheet and quality analysis was performed with the QUADAS-2 tool. Data were summarized and an overall summary was completed. RESULTS: We identified three papers (n = 221 patients; 33% NF) that met our inclusion criteria. The overall sensitivity ranged from 85.4%-100% while the specificity ranged from 44.7% to 98.2%. Fluid accumulation along the fascial plane was the most sensitive (85.4%; 95% CI 72.2% - 93.9%), while subcutaneous emphysema was the most specific (100%; 95% CI 92.5% - 100%). CONCLUSIONS: POCUS has good sensitivity and specificity for the diagnosis of NF. POCUS should be considered as an adjunct to the initial clinical decision making for the diagnosis of NF.

Diagnostic accuracy of ultrasound to confirm endotracheal tube depth.

INTRODUCTION: Endotracheal intubation is commonly performed in the Emergency Department. Traditional measures for estimating and confirming the endotracheal tube (ETT) depth may be inaccurate or lead to delayed recognition. Ultrasound may offer a rapid tool to confirm ETT depth at the bedside. METHODS: This was a randomized trial assessing the diagnostic accuracy of ultrasound to confirm ETT depth. Three cadavers were intubated in a random sequence with the ETT placed high (directly below the vocal cords), middle (2 cm above the carina), or deep (ETT at the carina). Seven blinded sonographers assessed the depth of the ETT using ultrasound. Outcomes included diagnostic accuracy of sonographer identification, time to identification, and operator confidence based upon ETT location. A subgroup analysis was performed to assess diagnostic accuracy by operator confidence. RESULTS: 441 total assessments were performed (154 high, 154 middle, and 133 deep ETT placements). Overall accuracy was 84.8% (95% CI 81.1% to 88.0%). When placed high, ultrasound was 82.5% sensitive (95% CI 75.5% to 88.1%) and 92.3% specific (95% CI 88.6% to 95.1%) with a mean time to identification of 15.3 s (95% CI 13.6-17.0) and a mean operator confidence of 3.9/5.0 (95% CI 3.7-4.1). When the ETT was placed in the middle, ultrasound was 83.8% sensitive (95% CI 77.0% to 89.2%) and 92.3% specific (95% CI 88.6% to 95.1%) with a mean time to identification of 16.7 s (95% CI 14.6-18.8) and a mean operator confidence of 3.7/5.0 (95% CI 3.5-3.9). When the ETT was placed deep, ultrasound was 88.0% sensitive (95% CI 81.2% to 93.0%) and 92.2% specific (95% CI 88.6% to 94.6%) with a mean time to identification of 19.0 s (95% CI 17.3-20.7) and a mean operator confidence of 3.4/5.0 (95% CI 3.2-3.6). Sonographers were significantly more accurate when they reported a higher confidence score. CONCLUSION: Ultrasound was moderately accurate for identifying the ETT location in a cadaveric model and was more accurate when sonographers felt confident with their visualization. Future research should determine the accuracy of combining transtracheal ultrasound with lung sliding and other modifications to improve the accuracy.

Covered or uncovered: A randomized control trial of Tegaderm versus no Tegaderm for ocular ultrasound.

BACKGROUND: Studies on ocular point-of-care ultrasound vary on whether gel should be directly applied to the eye or on top of an adhesive membrane (i.e., Tegaderm™). However, there are currently no data regarding which approach has better image quality and the impact of patient preference. In this study, we sought to address this gap by assessing the difference in image quality and patient preference between Tegaderm™ versus no Tegaderm™ for ocular ultrasound in the emergency department. METHODS: Patients were randomized to have a Tegaderm™ placed on either their right or left eye. The other eye served as a comparator with no Tegaderm™. Ultrasound was performed on the right eye followed by the left eye in all instances. After performing each ultrasound, the sonographer asked the patient to rate their maximal discomfort from the ultrasound of that eye using a Likert scale (0 = no discomfort; 10 = severe discomfort). The sonographer then asked the patient which side (Tegaderm™ vs no Tegaderm™) they preferred. Finally, images were reviewed by an experienced ultrasound fellowship-trained sonographer blinded to allocation and rated from 1 to 5. Continuous data were analyzed using descriptive statistics with mean and standard deviation. A paired samples t-test was performed to assess for differences between groups. Categorical data were presented as frequency and percentage. RESULTS: The mean image score was significantly worse with Tegaderm™ compared with no Tegaderm™ (mean difference: 0.94/5.00; 95% CI 0.79-1.08; p < 0.001). This was consistent in both the transverse and the sagittal plane subgroups. The percentage of acceptable images was also higher in the no Tegaderm™ group compared with the Tegaderm™ group (97.8% versus 82.8%). There was no statistically significant difference in patient discomfort with the Tegaderm™ versus no Tegaderm™ group. When asked to compare the two approaches, 54.4% of patients preferred Tegaderm™, 30.0% preferred no Tegaderm™, and 15.6% had no preference. CONCLUSIONS: Tegaderm™ was associated with reduced image quality and no significant difference in patient discomfort when utilized for ocular ultrasound. This study suggests that ocular ultrasound may be better performed without the use of Tegaderm™. Future research should evaluate the impact of Tegaderm™ vs. no Tegaderm™ among more novice users.

Comparison of saline versus air for identifying endotracheal intubation with ultrasound.

INTRODUCTION: After intubation has been performed, it is important to rapidly confirm the correct location of the endotracheal tube (ETT). Multiple techniques have been described, each with different limitations. Ultrasound has been increasingly recognized as an alternate modality for identifying the ETT location. However, it can be challenging to visualize the air-filled ETT cuff. Saline insufflation of the ETT cuff has been suggested to improve visualization of the ETT but data are limited. Our study sought to compare the diagnostic accuracy of air versus saline ETT cuff inflation on the diagnostic accuracy of intubation. METHODS: This was a randomized trial comparing air versus saline cuff inflation using a cadaver model. Adult cadavers were intubated in a random sequence with respect to both the location of intubation (i.e., tracheal vs esophageal) and air versus saline. Blinded sonographers assessed the location of the ETT using the static technique. Outcomes included accuracy of sonographer identification, time to identification, and operator confidence. RESULTS: 480 total assessments were performed. When using air, ultrasound was 95.8% sensitive (95% CI 90.5% to 98.6%) and 100% specific (95% CI 97.0% to 100%) with a mean time to confirmation of 8.5 s (95% CI 7.6 s to 9.4 s) and a mean operator confidence of 4.32/5.0 (95% CI 4.21 to 4.42). When using saline, ultrasound was 100% sensitive (95% CI 97.0% to 100%) and 100% specific (95% CI 97.0% to 100%) with a mean time to confirmation of 6.3 s (95% CI 5.9 s to 6.8 s) and a mean operator confidence of 4.52/5.0 (95% CI 4.44 to 4.60). CONCLUSION: There was no statistically significant difference between air versus saline for intubation confirmation. However, saline was associated with fewer false negatives. Additionally, time to confirmation was faster and operator confidence was higher with the saline group. Further studies should determine if the outcomes would change with more novice sonographers or in specific patient populations.

Ultrasound for the diagnosis of shoulder dislocation and reduction: A systematic review and meta-analysis.

BACKGROUND: Shoulder dislocations are a common injury prompting presentation to the emergency department. Point-of-care ultrasound (POCUS) is a diagnostic tool for shoulder dislocations, which has the potential to reduce time to diagnosis and reduction, radiation exposure, and health care costs. This systematic review sought to evaluate the diagnostic accuracy of POCUS for diagnosing shoulder dislocations. METHODS: We searched PubMed, Scopus, CINAHL, LILACS, the Cochrane databases, Google Scholar, and bibliographies of selected articles for all prospective and randomized controlled trials evaluating the diagnostic accuracy of POCUS for identifying shoulder dislocations. We dual-extracted data into a predefined worksheet and performed quality analysis using the QUADAS-2 tool. We performed a meta-analysis with subgroup analyses by technique and transducer type. As a secondary outcome, we assessed the diagnostic accuracy of identifying associated fractures. RESULTS: Ten studies met our inclusion criteria, comprising 1,836 assessments with 636 dislocations (34.6%). Overall, POCUS was 100% (95% confidence interval [CI], 85.6%-100%) sensitive and 100% (95% CI, 79.4%-100%) specific for the diagnosis of shoulder dislocation with a LR+ of 11,254.8 (95% CI, 3.9-3.3e7) and a LR- of <0.1 (95% CI, < 0.1-0.2). When compared with the anterior/lateral technique, the posterior technique had greater sensitivity but no difference in specificity. There was no difference between transducer types. POCUS was also 96.8% (95% CI, 92.6%-98.7%) sensitive and 99.7% (95% CI, 92.5%-100%) specific for the diagnosis of associated fractures. CONCLUSIONS: POCUS is a sensitive and specific tool for the rapid identification of shoulder dislocations and reductions, as well as for the detection of associated fractures. POCUS should be considered as an alternate diagnostic tool for the diagnosis and management of shoulder dislocations.

Patient and System Factors Related to Missed Opportunities for Screening in an Electronic Medical Record-driven, Opt-out HIV Screening Program in the Emergency Department.

OBJECTIVE: Emergency departments (EDs) have implemented HIV screening using a variety of strategies. This study investigates how specific patient and health system factors in the ED impact who is and is not screened in a combined targeted and nontargeted, electronic medical record (EMR)-driven, opt-out, HIV screening program. METHODS: This was a retrospective, cross-sectional study of ED visits where patients were determined eligible for HIV screening by an EMR algorithm between November 18, 2014, and July 15, 2015. The HIV screening workflow included three sequential events, all of which were required to get screened for HIV at the ED visit. The events were having a blood draw, being informed of the HIV screening policy by an ED nurse at the point of blood draw, and the patient consenting to the HIV test. Each event represented a dichotomous outcome and its association with six patient factors (age, sex, race/ethnicity, marital status, preferred language, and Emergency Severity Index [ESI]) and two health system factors (ED crowding and program phase) was investigated using multivariable modeling. RESULTS: A total of 15,918 ED visits were analyzed. Blood was drawn in 8,388 of 15,918 visits (53%). Of 8,388 visits where blood was drawn, there were 5,947 (71%) visits where ED nurses documented informing patients of the HIV screening policy. Of those visits, patient consent to the HIV test was documented at 3,815 (64%) visits. Patients between 13 and 19 years of age were significantly less likely to have blood drawn, to be informed of the screening policy, and to consent to the HIV test compared to other age groups. Both ED crowding and a patient's ESI were associated with decreased odds of having a blood draw and being informed of HIV screening by an ED nurse, but showed no association with patients consenting to the HIV test. CONCLUSION: Many patients, particularly adolescents and young adults, are missed in ED HIV screening programs that require blood draw and depend on providers to obtain consent for testing. To ensure that these patients are reached, future ED screening programs should strive to develop innovative workflows that allow for blood draws for HIV screening only and streamline the processes of obtaining informed consent and ordering tests for all eligible patients.

Retrograde dissection of the axillary artery during arteriogram via arteriovenous access.

INTRODUCTION: Retrograde arterial dissection is a recognized complication of endovascular intervention but has not been well reported especially in the context of dialysis arteriovenous access procedures. The management of this complication is also not well defined. CASE PRESENTATION: We report the case of an 80-year-old female with end-stage renal disease (ESRD) on dialysis who developed an asymptomatic retrograde dissection of the left axillary artery during an arteriogram performed via a retrograde approach as part of access dysfunction evaluation. The condition was managed conservatively with no other intervention and close follow-up. No surgical intervention or stenting was needed in this case. DISCUSSION: Conservative management of retrograde arterial dissection in the arm may be a viable option especially in asymptomatic patients due to the nature of the injury. The force of blood flow in the antegrade direction limits the expansion of the false lumen and likely promotes the spontaneous healing of the vessel.