Experimental comparison of first-pass effect between direct thromboaspiration and combined thrombectomy in the setting of distal basilar occlusion.

BACKGROUND: Studies investigating endovascular therapy in vertebro-basilar stroke have led to controversial results in the past, but recent randomized trials seem to show an effectiveness superiority of endovascular therapy versus best medical treatment. However, uncertainty remains concerning many aspects of thrombectomy in acute basilar artery occlusion, notably technical considerations. This study compared the first-pass effect of direct thromboaspiration and combined thrombectomy in the setting of distal basilar occlusion. METHODS: An in-vitro experimental set-up was used, consisting of a vascular phantom model and thrombus analogs of different consistencies to mimic human clots. Thrombus analogs were injected into the model through the vertebral artery and flowed to the basilar distal third to mimic a distal basilar occlusion. Ten procedures were performed for each thrombus analog stiffness and technique (direct thromboaspiration versus combined thrombectomy). RESULTS: Direct thromboaspiration showed an overall first-pass effect rate of 83.3% (25/30) and was particularly effective for ultra-soft and soft clot analogs, but decreased for hard clot analogs. Combined thrombectomy had an overall first-pass effect rate of 56.7% (17/30). The effect rate for ultra-soft and soft clot analogs was 60% and 50% for hard clot analogs. In the softer clot analogs, the stent-retriever device used for the combined thrombectomies tended to deviate the clot analog from a co-axial trajectory with the aspiration catheter. CONCLUSIONS: In the context of distal basilar occlusion, our in-vitro results showed that higher first-pass effect rates were achieved with direct thromboaspiration compared to combined thrombectomy in all types of thrombus analogs.

Can micro-guidewire advancement forces predict clot consistency and location to assist the first-line technique for mechanical thrombectomy?

BACKGROUND: The identification of specific clot characteristics before mechanical thrombectomy (MTB) might allow the selection of the most effective first-line technique, thus potentially improving the procedural outcome. We aimed to evaluate if the microwire push forces could extrapolate information on clot consistency and extension before MTB, based on clot mechanical properties. METHODS: We measured in vitro the forces exerted on the proximal extremity of the guidewire during the advancement and retrieval of the guidewire through clot analogs of different compositions. In addition, we analyzed the forces exerted on the guidewire to extrapolate information about the location of the proximal and distal extremities of the clot analogs. RESULTS: The maximum forces recorded during the whole penetration phase were significantly different for hard and soft clots (median values, 55.6 mN vs 15.4 mN, respectively; P<0.0001). The maximum slope of the force curves recorded during the advancement of the guidewire for the first 3 s of penetration also significantly differentiated soft from hard clot analogs (7.6 mN/s vs 23.9 mN/s, respectively; P<0.0001). In addition, the qualitative analysis of the shape of the force curves obtained during the advancement and retrieval of the guidewire showed a good potential for the identification of the proximal and distal edges of the clot analogs. CONCLUSION: Our results demonstrated that it was possible to differentiate between soft and hard clot analogs. Furthermore, force measurements could give important information about the location of the clot extremities. Such an approach might support the selection of the first-line MTB technique, with the potential to improve the outcome.

Double stent-retriever technique for mechanical thrombectomy: a systematic review and meta-analysis.

BACKGROUND: Mechanical thrombectomy using a double stent-retriever technique has recently been described for the treatment of acute ischemic stroke, but its efficacy and safety are not well established. PURPOSE: The aim of this systematic review and meta-analysis was to evaluate reports of the use of double stent-retriever during the endovascular treatment of patients with ischemic stroke. DATA SOURCES: The PubMed, Embase, Web of Science and Scopus databases were searched to identify all studies (clinical trials, cohorts series and case reports) investigating the utility of double stent-retriever for the treatment of stroke. The study is reported in accordance with PRISMA 2020 guidelines and was prospectively registered in PROSPERO (BLINDED FOR PEER REVIEW). STUDY SELECTION: 17 studies involving a total of 128 patients with large vessel occlusions predominantly in the anterior circulation (93.0%) were identified. DATA ANALYSIS: Outcomes of interest were the prevalence of successful recanalization (mTICI ≥2b) and a first-pass effect following double stent-retriever, as well as complications such as iatrogenic dissections and subarachnoid hemorrhage. Data were pooled using a random-effects model. DATA SYNTHESIS: Double stent-retriever was used as a rescue strategy in occlusions refractory to conventional endovascular treatment in 68.7% (88/128) of patients and as a first-line strategy in 31.3% (40/128) patients. Double stent-retriever achieved an overall final mTICI ≥2b in 92.6% cases with a first-pass effect of 76.6%. The complication rate remained low, with 0.37% dissection and 1.56% subarachnoid hemorrhage. LIMITATIONS: Limitations of the study include (1) a large number of case reports or small series, (2) a meta-analysis of proportions with no statistical comparison to a control group, and (3) the lack of access to patient-level data. CONCLUSIONS: Our findings suggest that double stent-retriever thrombectomy may be safe and associated with good recanalization outcomes, but prospective comparative studies are needed to determine which patients may benefit from this endovascular procedure. ABBREVIATIONS: AICH = asymptomatic intracranial hemorrhage; AIS = acute ischemic stroke; DSR = double stent-retriever; FPE = first pass effect; ICH = intracranial hemorrhage; LVO = large vessel occlusion; MT = mechanical thrombectomy; SAH = subarachnoid hemorrhage; SICH = symptomatic intracranial hemorrhage; SSR = single stent-retriever.

Validating the accuracy of deep learning for the diagnosis of pneumonia on chest x-ray against a robust multimodal reference diagnosis: a post hoc analysis of two prospective studies.

BACKGROUND: Artificial intelligence (AI) seems promising in diagnosing pneumonia on chest x-rays (CXR), but deep learning (DL) algorithms have primarily been compared with radiologists, whose diagnosis can be not completely accurate. Therefore, we evaluated the accuracy of DL in diagnosing pneumonia on CXR using a more robust reference diagnosis. METHODS: We trained a DL convolutional neural network model to diagnose pneumonia and evaluated its accuracy in two prospective pneumonia cohorts including 430 patients, for whom the reference diagnosis was determined a posteriori by a multidisciplinary expert panel using multimodal data. The performance of the DL model was compared with that of senior radiologists and emergency physicians reviewing CXRs and that of radiologists reviewing computed tomography (CT) performed concomitantly. RESULTS: Radiologists and DL showed a similar accuracy on CXR for both cohorts (p ≥ 0.269): cohort 1, radiologist 1 75.5% (95% confidence interval 69.1-80.9), radiologist 2 71.0% (64.4-76.8), DL 71.0% (64.4-76.8); cohort 2, radiologist 70.9% (64.7-76.4), DL 72.6% (66.5-78.0). The accuracy of radiologists and DL was significantly higher (p ≤ 0.022) than that of emergency physicians (cohort 1 64.0% [57.1-70.3], cohort 2 63.0% [55.6-69.0]). Accuracy was significantly higher for CT (cohort 1 79.0% [72.8-84.1], cohort 2 89.6% [84.9-92.9]) than for CXR readers including radiologists, clinicians, and DL (all p-values < 0.001). CONCLUSIONS: When compared with a robust reference diagnosis, the performance of AI models to identify pneumonia on CXRs was inferior than previously reported but similar to that of radiologists and better than that of emergency physicians. RELEVANCE STATEMENT: The clinical relevance of AI models for pneumonia diagnosis may have been overestimated. AI models should be benchmarked against robust reference multimodal diagnosis to avoid overestimating its performance. TRIAL REGISTRATION: NCT02467192 , and NCT01574066 . KEY POINT: • We evaluated an openly-access convolutional neural network (CNN) model to diagnose pneumonia on CXRs. • CNN was validated against a strong multimodal reference diagnosis. • In our study, the CNN performance (area under the receiver operating characteristics curve 0.74) was lower than that previously reported when validated against radiologists' diagnosis (0.99 in a recent meta-analysis). • The CNN performance was significantly higher than emergency physicians' (p ≤ 0.022) and comparable to that of board-certified radiologists (p ≥ 0.269).

Effectiveness of the preventive administration of vasoactive amines in counterbalancing significant patient blood pressure drops following nimodipine administration during mechanical thrombectomy procedures.

BACKGROUND: Focal vasospasm (FV) of the occluded vessel can occur during the endovascular treatment of acute ischemic stroke (AIS). Nimodipine is commonly used to treat vasospasm and can play a role in distinguishing it from artery narrowing due to iatrogenic dissection or residual clot. However, nimodipine administration can result in arterial hypotension and subsequent enlargement of the ischemic core. OBJECTIVE: To assess the efficacy of preventive and continuous vasoactive amine infusion to counterbalance nimodipine-induced hypotension. METHODS: We reviewed data from a prospective registry of patients treated for AIS between January 2019 and January 2022 who were administered nimodipine. All patients were equipped with an arterial cannula for invasive blood pressure measurement and given vasoactive amines preemptively before general anesthesia and throughout the procedure. Data obtained from invasive monitoring of mean arterial blood pressure (MABP) in a time-lapse of 25 min before and after nimodipine administration were analyzed. RESULTS: MABP significantly decreased after nimodipine administration but remained within the recommended range (81.79 ± 0.49 mmHg). Nimodipine was effective in reducing FV caused by stent retriever passage in 76.3% of cases. Furthermore, it proved valuable in diagnosing iatrogenic dissection (9.2%), residual clot (10.5%), or intracranial stenosis (4%). CONCLUSIONS: Infusion of vasoactive amines effectively counteracted the intraarterial nimodipine effect, thus avoiding frank arterial hypotension during endovascular treatment. Nimodipine has been useful in differentiating the diagnosis of FV resulting from mechanical thrombectomy and other potential causes, such as iatrogenic dissection or residual clot.

Incidence and outcome of perforations during medium vessel occlusion compared with large vessel occlusion thrombectomy.

BACKGROUND: Vessel perforation during thrombectomy is a severe complication and is hypothesized to be more frequent during medium vessel occlusion (MeVO) thrombectomy. The aim of this study was to compare the incidence and outcome of patients with perforation during MeVO and large vessel occlusion (LVO) thrombectomy and to report on the procedural steps that led to perforation. METHODS: In this multicenter retrospective cohort study, data of consecutive patients with vessel perforation during thrombectomy between January 1, 2015 and September 30, 2022 were collected. The primary outcomes were independent functional outcome (ie, modified Rankin Scale 0-2) and all-cause mortality at 90 days. Binomial test, chi-squared test and t-test for unpaired samples were used for statistical analysis. RESULTS: During 25 769 thrombectomies (5124 MeVO, 20 645 LVO) in 25 stroke centers, perforation occurred in 335 patients (1.3%; mean age 72 years, 62% female). Perforation occurred more often in MeVO thrombectomy (2.4%) than in LVO thrombectomy (1.0%, p<0.001). More MeVO than LVO patients with perforation achieved functional independence at 3 months (25.7% vs 10.9%, p=0.001). All-cause mortality did not differ between groups (overall 51.6%). Navigation beyond the occlusion and retraction of stent retriever/aspiration catheter were the two most common procedural steps that led to perforation. CONCLUSIONS: In our cohort, perforation was approximately twice as frequent in MeVO than in LVO thrombectomy. Efforts to optimize the procedure may focus on navigation beyond the occlusion site and retraction of stent retriever/aspiration catheter. Further research is necessary in order to identify thrombectomy candidates at high risk of intraprocedural perforation and to provide data on the effectiveness of endovascular countermeasures.

Benchtop evaluation of a double stent retriever thrombectomy technique for acute ischemic stroke treatment.

BACKGROUND AND PURPOSE: A mechanical thrombectomy technique using a double stent retriever approach has been reported for the treatment of patients with acute ischemic stroke. The purpose of this study was to perform a benchtop evaluation of the mechanism of action and efficacy of a double-stent retriever approach compared to a single-stent retriever approach. MATERIALS AND METHODS: In vitro, mechanical thrombectomy procedures were performed in a vascular phantom reproducing an M1-M2 occlusion with two different clot analog consistencies (soft and hard). We compared the double stent retriever approach to the single stent retriever approach and recorded the recanalization rate, distal embolization, and retrieval forces of each mechanical thrombectomy procedure. RESULTS: The double stent retriever approach achieved a higher recanalization rate and lower embolic complications compared to the single stent retriever approach. This seems to stem from two facts: the greater probability of targeting the correct artery with two stents in the case of bifurcation occlusion, and an improved clot capture mechanism using the double stent retriever approach. However, the double stent retriever was associated with an increased initial retrieval force. CONCLUSIONS: In vitro evaluation of the mechanism of action of the double stent retriever provided explanations that appear to support the high efficacy of such an approach in patient cohorts and could help operators when selecting the optimal mechanical thrombectomy strategy in cases of arterial occlusions difficult to treat with a single stent retriever.

Endovascular treatment strategies for simultaneous bilateral anterior circulation occlusions in acute ischemic stroke: A single-center experience and systematic review of the literature.

BACKGROUND: Bilateral and simultaneous occlusion of the anterior circulation is a rare event in patients with acute ischemic stroke. Although endovascular treatment is feasible and safe, the endovascular strategy to be used remains a subject of debate. OBJECTIVE: To assess the different endovascular strategies proposed for the treatment of a bilateral, simultaneous anterior circulation occlusion following acute ischemic stroke. METHODS: We present a retrospective study of the clinical and radiological records of all patients with a bilateral, simultaneous anterior circulation occlusion treated at our center between January 2019 and December 2022. Following the PRISMA guidelines, we also conducted a systematic review of the literature. RESULTS: Two patients with a bilateral and simultaneous middle cerebral artery occlusion were treated at our center during the study period. A TICI score ≥2b was obtained in 4 out of 4 occlusions. Modified Rankin Scale (mRS) at 90 days was 0 and 4, respectively. The literature review retrieved reports on 22 patients. The most frequent bilateral occlusion sites were internal carotid artery-middle cerebral artery. The clinical presentation was severe in most patients. A combined thrombectomy technique proved to have the highest number of first-pass recanalization. A TICI ≥2b was obtained in 95% of patients and an mRS ≤2 was found in 31.8% of patients. CONCLUSIONS: In patients with bilateral and simultaneous occlusion of the anterior circulation, endovascular treatment using a combined technique appears to be rapid and effective. The clinical evolution of this patient population strongly depends on the severity of the onset symptoms.

Three-dimensional rotational angiography improves mechanical thrombectomy recanalization rate for acute ischaemic stroke due to middle cerebral artery M2 segment occlusions.

BACKGROUND: Occlusions of the middle cerebral artery (MCA) M2 segments can be difficult to address with mechanical thrombectomy (MTB) using standard projections and this can affect the final recanalization. Three-dimensional rotational angiography (3D-RA) allows to obtain a 3D model of cerebral vessels in a few seconds and to determine the best two-dimensional (2D) projections to be selected to evaluate and treat cerebrovascular diseases, such as aneurysms or vascular malformations. We aimed to determine if 3D-RA could be applied also in MTB. METHODS: A retrospective review of two patient cohorts treated during two time periods of 12 months before and after the introduction of 3D-RA use at our institution for MTB in M2 occlusions. Analyses were conducted to compare the two groups for procedural characteristics, such as timing, recanalization rate and complications and clinical outcome. RESULTS: One hundred acute ischaemic stroke (AIS) patients (3D-RA group = 57; controls = 43) underwent MTB for an M2 occlusion during the two study periods. Recanalization rates were significantly higher in cases treated with 3D-RA. The mean 3D technique thrombectomy time was compared to that of non-3D cases (47 vs. 49 min, respectively). CONCLUSIONS: Our findings showed that 3D-RA is a useful tool to select specific working projections to AIS patients presenting an M2 occlusion by improving final recanalization compared to standard projections, without increasing the overall procedural time.

In vitro evaluation of how the presence of the stent retriever and microcatheter influences aspiration parameters in thrombectomy according to their position inside the aspiration catheter.

BACKGROUND: Several variations of the combined thrombectomy technique for acute ischemic stroke using a stent retriever and aspiration catheter have been described. The aim of our study was to assess how the presence of the microcatheter and stent retriever affect the basic aspiration parameters, namely, flow rate and aspiration force, depending on their position within the aspiration catheter. METHODS: Two experimental set-ups were designed to assess changes in flow rate and aspiration force according to the position of the stent retriever and microcatheter within the aspiration catheter. RESULTS: The transition of the stent retriever and microcatheter from the distal to proximal position resulted in a progressive increase in the flow rate, but with no impact on aspiration force. Additionally, the size of the stent retriever had no significant effect on flow rate changes and the reduction in flow rate was related to the microcatheter diameter. Negative pressure generated inside the aspiration catheter impacted on its distal segment located beyond the radiopaque marker, thus leading to its partial collapse. As a consequence, the measured aspiration force was lower than the theoretical aspiration force level for all tested aspiration catheters. CONCLUSIONS: In our experimental model, the position of the stent retriever and microcatheter within the aspirator catheter affected the flow rate, but not the aspiration force. Negative pressure generated within the aspiration catheter appeared to determine a partial collapse of the distal segment that resulted in a less effective aspiration force than the theoretical aspiration force level.

Cerebrovascular Complications and Vessel Wall Imaging in COVID-19 Encephalopathy-A Pilot Study.

BACKGROUND AND PURPOSE: Coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is associated with several complications of the central nervous system (CNS), including acute encephalopathy. METHODS: In this pilot study, we report a series of 39 patients (66.5 ± 9.2 years; 10.3% female) with acute encephalopathy, who underwent a standard brain magnetic resonance imaging (MRI) at 1.5 T during the acute symptomatic phase. In addition to diffusion-weighted imaging, MR angiography and susceptibility-weighted images, high-resolution vascular black blood sequences (in 34 cases) were used to investigate the vasculature of the brain. RESULTS: In 29 out of 34 patients with COVID-19 encephalopathy (85%) with high-resolution vessel wall imaging, we found a circular enhancement and thickening of the basilar and vertebral arteries, without any correlation with ischemia or microbleeds (reported in 21% and 59%, respectively). CONCLUSION: We report a high prevalence of vascular changes suggestive of endotheliitis as reported in other organs. This could suggest an inflammatory mechanism underlying this encephalopathy.

Acute Stenting and Concomitant Tirofiban Administration for the Endovascular Treatment of Acute Ischemic Stroke Related to Intracranial Artery Dissections: A Single Center Experience and Systematic Review of the Literature.

BACKGROUND: Intracranial artery dissection is an uncommon cause of acute ischemic stroke. Although acute stenting of the dissected arterial segment is a therapeutic option, the associated antiplatelet regimen remains a matter of debate. OBJECTIVES: To evaluate the efficacy and safety of acute intracranial stenting together with concomitant intravenous administration of tirofiban and to perform a systematic review of the literature. MATERIALS AND METHODS: A single-center, retrospective study of the clinical and radiological records of all patients treated at our center by intracranial stenting in the setting of acute ischemic stroke between January 2010 and December 2020. A systematic review of the literature was conducted according to the PRISMA-P guidelines for relevant publications from January 1976 to December 2020 on intracranial artery dissection treated by stent. RESULTS: Seven patients with intracranial artery dissections underwent acute stenting with concomitant tirofiban during the study period. Mid-term follow-up showed parent artery patency in 6/7 cases (85.7%). The modified Rankin Score was ≤ 0-2 at 3 months in 5/7 cases (71.4%). The literature review identified 22 patients with intracranial artery dissection treated with acute stenting in association with different antithrombotic therapies. Complete revascularization was obtained in 86.3% of cases with a modified Rankin Score of ≤ 0-2 in 68% of patients at 3-month follow-up. CONCLUSIONS: Acute intracranial stenting together with intravenous tirofiban administration could be a therapeutic option in patients with intracranial artery dissection and a small ischemic core.

How to Improve the Management of Acute Ischemic Stroke by Modern Technologies, Artificial Intelligence, and New Treatment Methods.

Stroke remains one of the leading causes of death and disability in Europe. The European Stroke Action Plan (ESAP) defines four main targets for the years 2018 to 2030. The COVID-19 pandemic forced the use of innovative technologies and created pressure to improve internet networks. Moreover, 5G internet network will be helpful for the transfer and collecting of extremely big databases. Nowadays, the speed of internet connection is a limiting factor for robotic systems, which can be controlled and commanded potentially from various places in the world. Innovative technologies can be implemented for acute stroke patient management soon. Artificial intelligence (AI) and robotics are used increasingly often without the exception of medicine. Their implementation can be achieved in every level of stroke care. In this article, all steps of stroke health care processes are discussed in terms of how to improve them (including prehospital diagnosis, consultation, transfer of the patient, diagnosis, techniques of the treatment as well as rehabilitation and usage of AI). New ethical problems have also been discovered. Everything must be aligned to the concept of "time is brain".

Experimental evaluation of direct thromboaspiration efficacy according to the angle of interaction between the aspiration catheter and the clot.

BACKGROUND: Successful direct thromboaspiration (DTA) is related to several factors such as clot consistency, size, and location. It has also been demonstrated recently that the angle of interaction (AOI) formed by the aspiration catheter and the clot is related to DTA efficacy. The aims of this study were three-fold: (a) to confirm the clinical finding that the AOI formed by the aspiration catheter and the clot influence DTA efficacy; (b) to evaluate to what extent this influence varies according to differences in clot consistency and size; and (c) to validate stent retriever thrombectomy as an effective rescue treatment after DTA failure in the presence of an unfavorable AOI. METHODS: A rigid vascular phantom designed to reproduce a middle cerebral artery trifurcation anatomy with three M2 segments forming different angles with M1 and thrombus analog of different consistencies and sizes was used. RESULTS: DTA was highly effective for AOIs >125.5°, irrespective of thrombus analog features. However, its efficacy decreased for acute AOIs. Rescue stent retriever thrombectomy was effective in 92.6% of cases of DTA failure. CONCLUSIONS: This in vitro study confirmed that the AOI formed by the aspiration catheter and the thrombus analog influenced DTA efficacy, with an AOI >125.5° related to an effective DTA. Stent retriever thrombectomy was an effective rescue treatment after DTA failure, even in the presence of an unfavorable AOI.

Clot-Based Radiomics Predict a Mechanical Thrombectomy Strategy for Successful Recanalization in Acute Ischemic Stroke.

BACKGROUND AND PURPOSE: Mechanical thrombectomy (MTB) is a reference treatment for acute ischemic stroke, with several endovascular strategies currently available. However, no quantitative methods are available for the selection of the best endovascular strategy or to predict the difficulty of clot removal. We aimed to investigate the predictive value of an endovascular strategy based on radiomic features extracted from the clot on preinterventional, noncontrast computed tomography to identify patients with first-attempt recanalization with thromboaspiration and to predict the overall number of passages needed with an MTB device for successful recanalization. METHODS: We performed a study including 2 cohorts of patients admitted to our hospital: a retrospective training cohort (n=109) and a prospective validation cohort (n=47). Thrombi were segmented on noncontrast computed tomography, followed by the automatic computation of 1485 thrombus-related radiomic features. After selection of the relevant features, 2 machine learning models were developed on the training cohort to predict (1) first-attempt recanalization with thromboaspiration and (2) the overall number of passages with MTB devices for successful recanalization. The performance of the models was evaluated on the prospective validation cohort. RESULTS: A small subset of radiomic features (n=9) was predictive of first-attempt recanalization with thromboaspiration (receiver operating characteristic curve-area under the curve, 0.88). The same subset also predicted the overall number of passages required for successful recanalization (explained variance, 0.70; mean squared error, 0.76; Pearson correlation coefficient, 0.73; P<0.05). CONCLUSIONS: Clot-based radiomics have the ability to predict an MTB strategy for successful recanalization in acute ischemic stroke, thus allowing a potentially better selection of the MTB strategy, as well as patients who are most likely to benefit from the intervention.

Radiomics and Machine Learning Differentiate Soft-Tissue Lipoma and Liposarcoma Better than Musculoskeletal Radiologists.

Distinguishing lipoma from liposarcoma is challenging on conventional MRI examination. In case of uncertain diagnosis following MRI, further invasive procedure (percutaneous biopsy or surgery) is often required to allow for diagnosis based on histopathological examination. Radiomics and machine learning allow for several types of pathologies encountered on radiological images to be automatically and reliably distinguished. The aim of the study was to assess the contribution of radiomics and machine learning in the differentiation between soft-tissue lipoma and liposarcoma on preoperative MRI and to assess the diagnostic accuracy of a machine-learning model compared to musculoskeletal radiologists. 86 radiomics features were retrospectively extracted from volume-of-interest on T1-weighted spin-echo 1.5 and 3.0 Tesla MRI of 38 soft-tissue tumors (24 lipomas and 14 liposarcomas, based on histopathological diagnosis). These radiomics features were then used to train a machine-learning classifier to distinguish lipoma and liposarcoma. The generalization performance of the machine-learning model was assessed using Monte-Carlo cross-validation and receiver operating characteristic curve analysis (ROC-AUC). Finally, the performance of the machine-learning model was compared to the accuracy of three specialized musculoskeletal radiologists using the McNemar test. Machine-learning classifier accurately distinguished lipoma and liposarcoma, with a ROC-AUC of 0.926. Notably, it performed better than the three specialized musculoskeletal radiologists reviewing the same patients, who achieved ROC-AUC of 0.685, 0.805, and 0.785. Despite being developed on few cases, the trained machine-learning classifier accurately distinguishes lipoma and liposarcoma on preoperative MRI, with better performance than specialized musculoskeletal radiologists.

Generative Adversarial Networks Improve the Reproducibility and Discriminative Power of Radiomic Features.

PURPOSE: To assess the contribution of a generative adversarial network (GAN) to improve intermanufacturer reproducibility of radiomic features (RFs). MATERIALS AND METHODS: The authors retrospectively developed a cycle-GAN to translate texture information from chest radiographs acquired using one manufacturer (Siemens) to chest radiographs acquired using another (Philips), producing fake chest radiographs with different textures. The authors prospectively evaluated the ability of this texture-translation cycle-GAN to reduce the intermanufacturer variability of RFs extracted from the lung parenchyma. This study assessed the cycle-GAN's ability to fool several machine learning (ML) classifiers tasked with recognizing the manufacturer on the basis of chest radiography inputs. The authors also evaluated the cycle-GAN's ability to mislead radiologists who were asked to perform the same recognition task. Finally, the authors tested whether the cycle-GAN had an impact on radiomic diagnostic accuracy for chest radiography in patients with congestive heart failure (CHF). RESULTS: RFs, extracted from chest radiographs after the cycle-GAN's texture translation (fake chest radiographs), showed decreased intermanufacturer RF variability. Using cycle-GAN-generated chest radiographs as inputs, ML classifiers categorized the fake chest radiographs as belonging to the target manufacturer rather than to a native one. Moreover, cycle-GAN fooled two experienced radiologists who identified fake chest radiographs as belonging to a target manufacturer class. Finally, reducing intermanufacturer RF variability with cycle-GAN improved the discriminative power of RFs for patients without CHF versus patients with CHF (from 55% to 73.5%, P < .001). CONCLUSION: Both ML classifiers and radiologists had difficulty recognizing the chest radiographs' manufacturer. The cycle-GAN improved RF intermanufacturer reproducibility and discriminative power for identifying patients with CHF. This deep learning approach may help counteract the sensitivity of RFs to differences in acquisition.Supplemental material is available for this article.© RSNA, 2020See also the commentary by Alderson in this issue.

Increased Reactivity of the Mesolimbic Reward System after Ketamine Injection in Patients with Treatment-resistant Major Depressive Disorder.

WHAT WE ALREADY KNOW ABOUT THIS TOPIC: The antidepressant effect of ketamine is associated with increased activity in the reward circuitry of the brain and a suppression of circuitry that mediates perceptual processing of negative emotions. The duration of ketamine effect on these brain structures remains to be defined. WHAT THIS ARTICLE TELLS US THAT IS NEW: As expected, ketamine administration led to an improvement in mood and global vigilance. The improvement in mood was accompanied by an increased recruitment of the orbitofrontal cortex, ventral striatum, medial substantial nigra and ventral tegmental area, structures that are part of the reward circuitry.Responses in the mesolimbic structures (amygdala, medial substantial nigra and ventral tegmental area, orbitofrontal cortex) to negative stimuli were decreased after ketamine administration.The data are consistent with the premise that ketamine induces sustained changes in the mesolimbic neural circuits to reset pathological reward and emotional processing. BACKGROUND: Ketamine rapidly improves maladaptive mood states in major depressive disorder, and some of the neural substrates underlying this therapeutic effect have been identified. This study aimed to identify functional changes within neural networks that may underlie the impact of ketamine on both reward and emotional processing in patients with treatment-resistant major depression. METHODS: Ten adult patients with a Montgomery-Åsberg Depression Rating Scale score above 25 were enrolled to receive a single intravenous administration of ketamine (0.5 mg/kg). Patients' performance along with related neural network activations were analyzed in a game-like reward task and in an emotional judgment task using functional magnetic resonance imaging 1 day before and 1 and 7 days after ketamine administration. RESULTS: A significant correlation (R = 0.46, P = 0.03) between the improvement of depression scores and the enhanced reaction time for positive items was found in the game-like reward task 1 day after ketamine administration. This enhanced sensitivity for rewarded items was accompanied by increased activity of reward-related brain regions, including the orbitofrontal cortex, ventral striatum, and the ventral tegmental area, an effect that persisted up to 1 week after ketamine injection. In the emotional judgment task, it was found that ketamine rapidly modified local brain activities in response to emotionally negative, positive, or neutral stimuli in the amygdala, insula, anterior cingulate cortex, and in the ventral tegmental area. CONCLUSIONS: Single bolus ketamine administration rapidly triggers lasting changes in mesolimbic neural networks to improve pathologic reward and emotional processing in patients with major depressive disorder.

Differentiating kidney stones from phleboliths in unenhanced low-dose computed tomography using radiomics and machine learning.

OBJECTIVES: Distinguishing between kidney stones and phleboliths can constitute a diagnostic challenge in patients undergoing unenhanced low-dose CT (LDCT) for acute flank pain. We sought to investigate the accuracy of radiomics and a machine-learning classifier in differentiating between kidney stones and phleboliths on LDCT. METHODS: Radiomics features were extracted following a semi-automatic segmentation of kidney stones and phleboliths for two independent consecutive cohorts of patients undergoing LDCT for acute flank pain. Radiomics features from the first cohort of patients (n = 369) were ultimately used to train a machine-learning model designed to distinguish kidney stones (n = 211) from phleboliths (n = 201). Classification performance was assessed on the second independent cohort (i.e., testing set) (kidney stones n = 24; phleboliths n = 23) using positive and negative predictive values (PPV and NPV), area under the receiver operating curves (AUC), and permutation testing. RESULTS: Our machine-learning classification model trained on radiomics features achieved an overall accuracy of 85.1% on the independent testing set, with an AUC of 0.902, PPV of 81.5%, and NPV of 90.0%. Classification accuracy was significantly better than chance on permutation testing (p < 0.05, permutation p value). CONCLUSION: Radiomics and machine learning enable accurate differentiation between kidney stones and phleboliths on LDCT in patients presenting with acute flank pain. KEY POINTS: • Combining a machine-learning algorithm with radiomics features extracted for abdominopelvic calcification on LDCT offers a highly accurate method for discriminating phleboliths from kidney stones. • Our radiomics and machine-learning model proved robust for CT acquisition and reconstruction protocol when tested in comparison with an external independent cohort of patients with acute flank pain. • The high performance of the radiomics-based automatic classification model in differentiating phleboliths from kidney stones indicates its potential as a future diagnostic tool for equivocal abdominopelvic calcifications in the setting of suspected renal colic.

Confidence of emotion expression recognition recruits brain regions outside the face perception network.

Metacognitive beliefs about emotions expressed by others are crucial to social life, yet very little studied. To what extent does our confidence in emotion expression recognition depend on perceptual or other non-perceptual information? We obtained behavioral and magnetic resonance imaging measures while participants judged either the emotion in ambiguous faces or the size of two lines flanking these faces, and then rated their confidence on decision accuracy. Distinct behavioral and neural mechanisms were identified for confidence and perceptual decision in both tasks. Participants overestimated their emotion recognition (ER) accuracy, unlike visual size judgments. Whereas expression discrimination recruited several areas in the face-processing network, confidence for ER uniquely engaged the bilateral retrosplenial/posterior cingulate complex and left parahippocampal gyrus. Further, structural white matter connectivity of the former region predicted metacognitive sensitivity. These results highlight a key role for brain mechanisms integrating perception with contextual mnemonic information in the service of confidence during ER.