Prediction of Response of Hepatocellular Carcinoma to Radioembolization: Machine Learning Using Preprocedural Clinical Factors and MR Imaging Radiomics.

PURPOSE: To create and evaluate the ability of machine learning-based models with clinicoradiomic features to predict radiologic response after transarterial radioembolization (TARE). MATERIALS AND METHODS: 82 treatment-naïve patients (65 responders and 17 nonresponders; median age: 65 years; interquartile range: 11) who underwent selective TARE were included. Treatment responses were evaluated using the European Association for the Study of the Liver criteria at 3-month follow-up. Laboratory, clinical, and procedural information were collected. Radiomic features were extracted from pretreatment contrast-enhanced T1-weighted magnetic resonance images obtained within 3 months before TARE. Feature selection consisted of intraclass correlation, followed by Pearson correlation analysis and finally, sequential feature selection algorithm. Support vector machine, logistic regression, random forest, and LightGBM models were created with both clinicoradiomic features and clinical features alone. Performance metrics were calculated with a nested 5-fold cross-validation technique. The performances of the models were compared by Wilcoxon signed-rank and Friedman tests. RESULTS: In total, 1,128 features were extracted. The feature selection process resulted in 12 features (8 radiomic and 4 clinical features) being included in the final analysis. The area under the receiver operating characteristic curve values from the support vector machine, logistic regression, random forest, and LightGBM models were 0.94, 0.94, 0.88, and 0.92 with clinicoradiomic features and 0.82, 0.83, 0.82, and 0.83 with clinical features alone, respectively. All models exhibited significantly higher performances when radiomic features were included (P = .028, .028, .043, and .028, respectively). CONCLUSIONS: Based on clinical and imaging-based information before treatment, machine learning-based clinicoradiomic models demonstrated potential to predict response to TARE.

MRI Signal Intensity Varies Along the Course of the Normal Optic Nerve.

BACKGROUND: MRI can help distinguish various causes of optic neuropathy including optic neuritis. Importantly, neuromyelitis optica spectrum disorder (NMOSD) has a propensity to cause enhancement of the prechiasmatic optic nerves. To determine whether the prechiasmatic optic nerve (PC-ON) demonstrates a different intensity from the midorbital optic nerve (MO-ON) on MRI among patients without optic neuropathy. METHODS: Data were retrospectively obtained from 75 patients who underwent brain MRI for an ocular motor nerve palsy between January 2005 and April 2021. Inclusion criteria were patients aged 18 years or older with visual acuities of at least 20/25 and no evidence of optic neuropathy on neuro-ophthalmic examination. A total of 67 right eyes and 68 left eyes were assessed. A neuroradiologist performed quantitative intensity measurements of the MO-ON and PC-ON on precontrast and postcontrast T1 axial images. Normal-appearing temporalis muscle intensity was also measured and used as a reference to calculate an intensity ratio to calibrate across images. RESULTS: The mean PC-ON intensity ratio was significantly higher than the MO-ON intensity ratio on both precontrast (19.6%, P < 0.01) and postcontrast images (14.2%, P < 0.01). Age, gender, and laterality did not independently affect measurements. CONCLUSIONS: The prechiasmatic optic nerve shows brighter intensity ratios on both precontrast and postcontrast T1 images than the midorbital optic nerve among normal optic nerves. Clinicians should recognize this subtle signal discrepancy when assessing patients with presumed optic neuropathy.

Effect of bariatric surgery procedures on serum cytokine and Nesfatin-1 levels.

BACKGROUND: and Purpose: Obesity is known to cause chronic inflammation. We aimed to evaluate the changes in Nesfatin-1 and serum cytokine levels of patients who underwent sleeve gastrectomy or gastric bypass surgery. METHODS: A total of 30 patients with BMI>35 and undergoing bariatric surgery were divided in two group, sleeve gastrectomy (SG) (group-1), Roux-en-Y gastric bypass (RYGB) (group-2). Demographic data, weight, BMI, AST, ALT, blood glucose, CRP values, and IL-1ß, IFN-α, IFN-γ, TNF-α, MCP-1, IL-6, IL-8, IL-10, IL-12p70, IL-17A, IL-18, IL-23, IL-33 cytokine, and Nesfatin-1 values were noted at the time of hospitalization and in the 6th month postoperative follow-up. RESULTS: The mean age of the patients was 37.56 ± 11.73 years, and there were 16 females and 14 males in the study. Body weight and excess body weight change were slightly higher in RYGB patients than in SG patients. In RYGB and SG patients, a significant decrease was found in glucose, AST, ALT, CRP, IL-6, IL-10, and IL-18 values compared to the preoperative period, and serum Nesfatin-1 levels were significantly increased in RYGB patients and not significantly in SG patients. There were also significant decreases in IL-1ß levels in RYGB patients. On the other hand, a decrease in cytokines was observed in both surgical methods, except for IL-17A, although it was not significant. CONCLUSION: The present study showed that there is also a regression in inflammation, which can be associated with NLRP3 inflammasome, due to weight loss after bariatric surgery, more specifically in RYGB.

A bicentric retrospective analysis of clinical utility of 18F-fluciclovine PET in biochemically recurrent prostate cancer following primary radiation therapy: is it helpful in patients with a PSA rise less than the Phoenix criteria?

PURPOSE: 18F-Fluciclovine PET imaging has been increasingly used in the restaging of prostate cancer patients with biochemical recurrence (BCR); however, its clinical utility in patients with low prostate-specific antigen (PSA) levels following primary radiation therapy has not been well-studied. This study aims to determine the detection rate and diagnostic accuracy of 18F-fluciclovine PET and the patterns of prostate cancer recurrence in patients with rising PSA after initial radiation therapy, particularly in patients with PSA levels below the accepted Phoenix definition of BCR (PSA nadir +2 ng/mL). METHODS: This retrospective study included patients from two tertiary institutions who underwent 18F-fluciclovine PET scans for elevated PSA level following initial external beam radiation therapy, brachytherapy, and/or proton therapy. Logistic regression and receiver operating characteristic (ROC) curve analyses were performed to determine the diagnostic accuracy of 18F-fluciclovine PET and associations of PSA kinetic parameters with 18F-fluciclovine PET outcome. RESULTS: One hundred patients were included in this study. The overall detection rate on a patient-level was 79% (79/100). 18F-Fluciclovine PET was positive in 62% (23/37) of cases with PSA below the Phoenix criteria. The positive predictive value of 18F-fluciclovine PET was 89% (95% CI: 80-94%). In patients with PSA below the Phoenix criteria, the PSA velocity had the highest predictive value of 18F-fluciclovine PET outcome. PSA doubling time (PSADT) and PSA velocity were associated with the presence of extra-pelvic metastatic disease. CONCLUSION: 18F-Fluciclovine PET can identify recurrent disease at low PSA level and PSA rise below accepted Phoenix criteria in patients with suspected BCR after primary radiation therapy, particularly in patients with low PSADT or high PSA velocity. In patients with low PSADT or high PSA velocity, there is an increased probability of extra-pelvic metastases. Therefore, these patients are more likely to benefit from PET/CT or PET/MRI than pelvic MRI alone.

Ultrasound-Guided Portal Vein Access and Percutaneous Wire Placement in the Portal Vein Are Associated With Shorter Procedure Times and Lower Radiation Doses During TIPS Placement.

BACKGROUND. TIPS placement is an effective method for treating a number of complications of portal hypertension. Although this complex procedure has been firmly established in treatment algorithms, more data are needed to determine the most efficient and safest ways to perform the procedure. OBJECTIVE. The purpose of this study was to determine the effect of three different techniques of portal vein (PV) cannulation during TIPS placement on procedure efficiency. METHODS. The medical records of patients who underwent TIPS creation between 2005 and 2019 were reviewed. On the basis of the PV access technique used, patients were grouped as follows: group 1 (G1) included patients who underwent a transabdominal ultrasound (US)-guided technique to obtain PV access, group 2 (G2) consisted of those who underwent fluoroscopically guided wedged hepatic portography, and group 3 (G3) included those who underwent percutaneous US-guided PV guidewire placement for fluoroscopic targeting. RESULTS. Of the 264 patients who underwent TIPS creation, 54 (20.5%) were in G1, 172 (65.1%) were in G2, and 38 (14.4%) were in G3. The mean (± SD) fluoroscopic time in G1 (34.8 ± 16.6 minutes) did not differ from that in either G2 (38.9 ± 20.8 minutes; p = .09) or G3 (29.5 ± 14.6 minutes; p = .06). However, G2 patients had significantly longer fluoroscopic times than G3 patients (p = .005). The mean total anesthesia time in G1 (190.2 ± 45.6 minutes) did not differ from that in G2 (199.7 ± 59.5 minutes; p = .15). However, G3 had a mean anesthesia time (162.6 ± 39.7 minutes) that was significantly shorter than that in both G1 (p = .003) and G2 (p < .001). The mean contrast volume was significantly lower in G1 than in G2 (67.9 ± 36.8 mL vs 87.1 ± 42.9 mL; p = .005). More intrahepatic needle passes were required in G2 (median, 4 passes; interquartile range [IQR], 1-7 passes) than in G1 (median, 2 passes; IQR, 1-4 passes; p = .004) and G3 (median, 2 passes; IQR, 1-7.25 passes; p = .04). When complications in G1 and G3 were pooled, this cohort had significantly fewer complications than G2 (p = .01). CONCLUSION. Ultrasound-guided PV access and percutaneous PV guidewire placement for fluoroscopic targeting during TIPS creation are associated with shorter procedure and fluoroscopic times and potentially decreased complications. CLINICAL IMPACT. The present study helps interventional radiologists understand the safest and most efficient way to access the PV, which is a key step during TIPS placement.

Gadoxetate Disodium-Enhanced Imaging of Gradenigo Syndrome in End-Stage Renal Disease.

ABSTRACT: A 65-year-old man with end-stage renal disease on peritoneal dialysis was admitted for new onset binocular double vision, failure to thrive, and debilitating right-sided headaches. Medical history was significant for tympanomastoidectomy for polymicrobial mastoiditis and treatment with IV antibiotics. MRI brain without contrast was read by radiology initially as showing expected postsurgical changes; however, given patient's history of mastoiditis infection, there was a high clinical suspicion for Gradenigo syndrome. MRI brain was repeated with gadoxetate contrast to minimize the risk of nephrogenic systemic fibrosis (NSF) in a patient with severe renal disease and this revealed an intracranial empyema involving both the trigeminal and abducens nerves confirming the diagnosis of Gradenigo syndrome. This case presented a diagnostic challenge because of suboptimal visualization on initial nonenhanced MRI. Gadoxetate was chosen because of its unique properties including high hepatobiliary excretion making it a safer form of gadolinium-based contrast agent that may not have the potential to precipitate NSF.

Effect of platelet-rich plasma on healing in laser pilonidoplasty for pilonidal sinus disease.

The aim of this study was to evaluate the effects of application of platelet-rich plasma in addition to laser pilonidoplasty for the treatment of pilonidal sinus. Twenty-five patients who were treated by laser pilonidoplasty for pilonidal sinus (group 1) and 25 patients who were treated by platelet-rich plasma in addition to laser pilonidoplasty (group 2) at this clinic were included in the study. Patients were classified according to the Irkorucu and Adana Numune's classification and treatment concept. Duration of stay of the patients in the hospital, time to start daily activities, duration of wound healing, recurrence, and complications were evaluated. Among the 50 patients included in the study, 41 (%82) were males and 9 (%18) were females. The mean age was 25.6 ± 2.4 years and 24.8 ± 3.8 years in groups 1 and 2, respectively. The locations of the pilonidal sinus were similar in the two groups. No statistically significant differences were found in the duration of hospital stay, duration of the procedure, time to return to work, and complication rates between the two groups. Nevertheless, duration of wound healing was 6.1 ± 2.3 and 4.1 ± 0.9 weeks in groups 1 and 2, respectively, and was shorter in group 2. Duration of wound healing was statistically significantly different in the two groups. We concluded in this study that application of platelet-rich plasma in addition to laser pilonidoplasty significantly shortens the time of wound healing.

Surveillance Imaging after Autologous Hematopoietic Cell Transplantation Predicts Survival in Patients with Diffuse Large B Cell Lymphoma.

The utility of surveillance imaging after autologous hematopoietic cell transplantation (AHCT) in relapsed/refractory diffuse large B cell lymphoma (DLBCL) remains unclear. The purpose of this study was to determine whether surveillance imaging predicts survival after AHCT. At the University of Minnesota, serial imaging for early relapse detection has been used prospectively for all consecutive AHCT recipients treated since 2010. The present analysis included 91 AHCT recipients with DLBCL who underwent 18F-fluorodeoxyglucose positron emission tomography (18F-FDG-PET) scan at day +100 post-AHCT. 18F-FDG-PET parameters included the Deauville (D) 5-point scale, peak standardized uptake values (SUVmax), total legion glycolysis (TLG), and total metabolic tumor volume (TMTV). Survival of patients with clinically symptomatic versus asymptomatic radiographically detected relapsed DLBCL after AHCT was compared. Sixty patients experienced relapse; 35% was detected on day +100 surveillance PET scan. 5-year overall survival (OS) by 18F-FDG-PET scan at day +100 post-AHCT was significantly lower in D4 and D5 patients (37%; 95% confidence interval [CI], 14% to 100% versus 25%; 95% CI, 43% to 89%) compared with patients with D1 and D2 (62%; 95% CI, 43% to 89% versus 62%; 95% CI, 46% to 84%). TLG and TMTV were not prognostic. SUVmax at day +100 varied from 1.5 (D1) to 17.9 (D5). In multivariate analysis, only SUVmax was predictive of relapse and OS; mortality increased 1.8-fold with each SUVmax doubling (hazard ratio [HR], 1.8; 95% CI, 1.3 to 2.3; P < .01). At a median follow-up of 3.3 years (range, 1 to 12 years), lymphoma-related mortality was 1.8-fold higher among patients whose relapse was detected clinically (symptomatic) versus radiographically on surveillance scan (HR, 1.8; 95% CI, .9 to 3.4; P = .08). In patients with relapsed/refractory DLBCL, a routine PET imaging at day +100 post-AHCT detects asymptomatic relapse and high SUVmax identifies patients with poor expected survival of less than 1 year. Identifying this high-risk cohort can potentially highlight patients who might benefit from preemptive interventions to prevent or delay relapse.

Role of Whole-Body 18F-FDG PET/CT in Screening for Metastases in Newly Diagnosed Sinonasal Malignancies.

OBJECTIVE. The aim of this study is to assess the utility of pretreatment whole-body 18F-FDG PET/CT in screening for distant metastasis (DM) and regional lymphatic metastasis (LM). MATERIALS AND METHODS. Eighty-nine consecutive patients with untreated sinonasal malignant lesions (32 women and 57 men; mean age, 62 years) underwent whole-body FDG PET/CT between January 2009 and August 2017. A retrospective analysis was performed to determine the presence of DM and LM. Any suspected metastases were confirmed by histopathologic analysis or clinical and imaging follow-up in the subsequent 12 months. The statistics were verified by comparing FDG PET/CT results with a reference standard. RESULTS. Overall, the frequency of DM was 24% (21/89), of which 81% (17/21) were identified by whole-body FDG PET/CT. The sensitivity and specificity of FDG PET/CT in predicting DM were 81% (95% CI, 62-95%) and 99% (95% CI, 82-100%), respectively. The most common DM sites were the lungs (n = 6; 28%) and bones (n = 5; 24%), followed by the liver (n = 2; 10%), brain (n = 1; 5%), and spinal canal (n = 1; 5%), with six patients (28%) having DMs at multiple sites. Overall, the frequency of LM according to the reference standard was 20%, of which 83% (15/18) were confirmed with FDG PET/CT. The sensitivity and specificity of FDG PET/CT in detecting LM were 83% (95% CI, 68-97%) and 96% (95% CI, 77-100%), respectively. CONCLUSION. Our study showed that whole-body FDG PET/CT can be used as a screening tool for the detection of DM and LM in sinonasal neoplasms and could be performed as part of the routine pretreatment evaluation of metastatic workup.

Performance of whole-body 18F-FDG PET/CT as a posttreatment surveillance tool for sinonasal malignancies.

PURPOSE: To determine the diagnostic utility of posttreatment surveillance whole-body 18F-FDG PET/CT in detecting local tumor recurrence (R), regional lymph-node metastasis (LM), and distant metastasis (DM) in asymptomatic sinonasal cancer patients. METHODS: Eighty consecutive patients (53 men, 27 women; mean age, 60 years; range, 28-92 years) who had undergone 197 posttreatment whole-body 18F-FDG PET/CT examinations for sinonasal malignancies between January 2009 and August 2017 were retrospectively reviewed. 18F-FDG PET/CT findings were categorized as positive or negative for R, LM, and DM, separately. Outcomes of 18F-FDG PET/CT scans were compared with the final diagnosis confirmed by histological analysis or follow-up period for a minimum 12 months. The diagnostic accuracy of scans was calculated for each site using contingency tables. Impact on the management of 18F-FDG PET/CT examinations was additionally evaluated. RESULTS: 18F-FDG PET/CT scans identified 37/44 of local recurrences, 21/23 of LMs, and 30/37 of DMs. For local recurrence, sensitivity, specificity, positive predictive value, and negative predictive value were 84% (68-97%), 95% (80-100%), 84% (68-97%), and 95% (80-100%), respectively. For LM, the respective values were 91% (75-100%), 99% (83-100%), 91% (75-100%), and 99% (83-100%). For DM, the values were 81% (64-97%), 99% (85-100%), 97% (81-100%), and 96% (81-100%), respectively. 18F-FDG PET/CT accounted for a change in management of 85% patients with recurrences. CONCLUSIONS: Whole-body 18F-FDG PET/CT is a suitable surveillance tool for sinonasal malignancies in detecting locoregional and distant recurrences in asymptomatic patients without any evidence of recurrence on regular follow-up and endoscopy during the posttreatment period.

Utility of FDG PET/CT in the Characterization of Sinonasal Neoplasms: Analysis of Standardized Uptake Value Parameters.

OBJECTIVE: We aimed to evaluate the contribution of different standardized uptake value (SUV) parameters generated from pretreatment 18F-FDG PET/CT in the characterization of sinonasal neoplasms with histopathologic correlations. MATERIALS AND METHODS: This retrospective study included 97 consecutive patients (58 men, 39 women; age range, 20-93 years; mean age, 62 years) with pathologically proven untreated sinonasal neoplasms who underwent FDG PET/CT from February 2010 to August 2017. Semiquantitative analysis of primary tumors were performed to evaluate the maximum standardized uptake value (SUVmax), mean standardized uptake value (SUVmean), and the ratio of the SUVmax of the primary tumor to the SUVmean of mediastinal blood pool, which we refer to here as " SUVratio." Various sinonasal tumor histopathologic subgroups (n = 14) were analyzed. The Kruskal-Wallis test was used to compare the SUVmax, SUVmean, and SUVratio with the histopathologic diagnosis. RESULTS: Mean values of SUVmax, SUVmean, and SUVratio for the sinonasal neoplasms were 16.6 ± 9.7 (SD), 8.6 ± 5.1, and 5.9 ± 3.7, respectively, and each parameter was significantly different between histopathologic types (p < 0.05). Mean values of SUVmax, SUVmean, and SUVratio were higher in sinonasal undifferentiated carcinoma (SNUC) than in olfactory neuroblastoma, metastasis, and adenoid cystic carcinoma (p < 0.05). Mean values of SUVmax and SUVmean were higher in squamous cell carcinoma (SCC) than in olfactory neuroblastoma and metastasis (p < 0.05). Also, mean SUVmax was higher in SCC and SNUC than in poorly differentiated carcinoma (p < 0.05). Mean SUVratio was higher in SCC than in small cell carcinoma, olfactory neuroblastoma, and adenoid cystic carcinoma (p < 0.05). CONCLUSION: We conclude that different SUV parameters from FDG PET/CT can be used as so-called "metabolic biopsy" to categorize sinonasal neoplasms into different histopathologic subgroups because it can help in the characterization of some of the more common subgroups of sinonasal neoplasms. However, we found that there is overlap in FDG uptake values among some of the rare histologic subgroups; hence, surgical biopsy is still needed for differentiation of histologic subtypes of aggressive sinonasal masses.

Quantitative Assessment of Optic Nerve With Diffusion Tensor Imaging in Patients With Thyroid Orbitopathy.

PURPOSE: To detect abnormalities of the optic nerve in patients with thyroid orbitopathy using diffusion tensor MRI. METHODS: Twenty-eight patients with Graves orbitopathy prospectively underwent diffusion tensor imaging scanning. A full ophthalmic examination including visual acuity, intraocular pressure, fundoscopy, and visual field analysis was performed. Clinical activity scores were also calculated. Fractional anisotropy (FA) and mean diffusivity values of the patients were compared with age and sex-matched healthy control subjects. RESULTS: The mean FA values were decreased and mean diffusivity values were increased significantly in patients with Graves orbitopathy compared with the control subjects (p < 0.001). There was a strong reverse correlation between the FA levels and the visual fields in 4 quadrants of the optic nerve. In addition, there was a strong correlation between the degree of proptosis and the FA values in both eyes. The mean diffusivity levels were also correlated with changes in the visual field and the degree of proptosis. CONCLUSION: FA and mean diffusivity levels measured with the diffusion tensor imaging of the thyroid orbitopathy patients were affected. The changes in diffusion tensor imaging were also correlated with the ophthalmologic tests of the patients.

Enhancing bariatric surgery safety for patients refusing blood transfusions: a specialized protocol with comprehensive technical measures.

Bariatric surgery has become a leading treatment for obesity, with techniques such as Laparoscopic Sleeve Gastrectomy (LSG) and Laparoscopic Roux-en-Y Gastric Bypass (LRYGB) demonstrating notable success in sustained weight loss and improved quality of life. Technological advancements and improved techniques have enhanced the safety of these procedures. The surgical procedures of Jehovah's Witnesses, who refuse blood transfusions as part of their beliefs, pose unique challenges and have rarely been addressed in the context of bariatric surgery. This report aimed to investigate the safety of bariatric surgery in patients who refuse blood transfusion, with an established protocol to minimize the risk of bleeding. We examined the prospectively collected data of Jehovah's Witness patients who underwent bariatric surgery from 2019 to 2023. The surgeries were conducted following a protocol that included specific measures to prevent bleeding. Data were reviewed for demographics, anthropometrics, comorbidities, preoperative medications, operative time, blood loss, length of hospital stay, hemoglobin level, drainage volume, tranexamic acid use, and postoperative 30-day complications. Eleven Jehovah's Witness patients underwent bariatric surgery, including 10 LSG and 1 LRYGB. A patient with iron deficiency anemia underwent intravenous iron treatment before the surgery. There were no intraoperative complications or major postoperative complications. All patients maintained stable hemodynamics postoperatively. Only one patient encountered nausea-vomiting, classified as a minor complication. One patient experienced a small amount of hemorrhagic drainage, which transitioned to serous after tranexamic acid infusion. Bariatric surgery can be performed safely with established protocols in patients who refuse blood transfusions.

Establishing Optic Nerve Diameter Threshold Sensitive and Specific for Optic Atrophy Diagnosis.

PURPOSE: To determine a potential threshold optic nerve diameter (OND) that could reliably differentiate healthy nerves from those affected by optic atrophy (OA) and to determine correlations of OND in OA with retinal nerve fiber layer (RNFL) thickness, visual acuity (VA), and visual field mean deviation (VFMD). METHODS: This was a retrospective case control study. Magnetic resonance (MR) images were reviewed from individuals with OA aged 18 years or older with vision loss for more than 6 months and an OA diagnosis established by a neuro-ophthalmologist. Individuals without OA who underwent MR imaging of the orbit for other purposes were also collected. OND was measured on coronal T2-weighted images in the midorbital section, 1cm posterior to the optic disc. Measurements of mean RNFL thickness, VA and VFMD were also collected. RESULTS: In this study 47 OA subjects (63% women, 78 eyes) and 75 normal subjects (42.7% women, 127 eyes) were assessed. Healthy ONDs (mean 2.73 ± 0.24 mm) were significantly greater than OA nerve diameters (mean 1.94 ± 0.32 mm; P < 0.001). A threshold OND of ≤2.3 mm had a sensitivity of 0.92 and a specificity of 0.93 in predicting OA. Mean RNFL (r = 0.05, p = 0.68), VA (r = 0.17, p = 0.14), and VFMD (r = 0.18, p = 0.16) were not significantly associated with OND. CONCLUSION: ONDs are significantly reduced in patients with OA compared with healthy nerves. A threshold OND of ≤2.3 mm is highly sensitive and specific for a diagnosis of OA. OND was not significantly correlated with RNFL thickness, VA, or VFMD.

Correct Stapling Technique in Laparoscopic Sleeve Gastrectomy: Are We Increasing the Bleeding? A Prospective Cohort Study.

Introduction: Laparoscopic sleeve gastrectomy (LSG) has become the most common bariatric surgery procedure. Bleeding is the most common complication of this surgery and mostly occurs from the staple line. The aim of this study was to evaluate whether waiting between compression and firing during the stapling phase reduces the postoperative bleeding. Methods: A total of 325 patients who underwent LSG between April and July 2022 were analyzed prospectively. In terms of postoperative bleeding, the two groups, which we waited 30 seconds between staple firings and the no wait group, were compared. Results: The mean age of patients was 37.36 (±11.12) years and mean body mass index was 45.18 (±3.1) kg/m2. Eleven patients needed transfusion. The rate of haemorrhagic complications was 3.38% (Group 1% 6.21 and Group 2% 1.11) (P = .012). The duration of surgery was ∼10 minutes longer in the study group, which we waited (P = .0001). Conclusions: During the stapling stage in LSG, waiting between compression and firing can help reduce postoperative bleeding.

Machine Learning Insights: Predicting Hepatic Encephalopathy After TIPS Placement.

PURPOSE: To develop and assess machine learning (ML) models' ability to predict post-procedural hepatic encephalopathy (HE) following transjugular intrahepatic portosystemic shunt (TIPS) placement. MATERIALS AND METHODS: In this retrospective study, 327 patients who underwent TIPS for hepatic cirrhosis between 2005 and 2019 were analyzed. Thirty features (8 clinical, 10 laboratory, 12 procedural) were collected, and HE development regardless of severity was recorded one month follow-up. Univariate statistical analysis was performed with numeric and categoric data, as appropriate. Feature selection is used with a sequential feature selection model with fivefold cross-validation (CV). Three ML models were developed using support vector machine (SVM), logistic regression (LR) and CatBoost, algorithms. Performances were evaluated with nested fivefold-CV technique. RESULTS: Post-procedural HE was observed in 105 (32%) patients. Patients with variceal bleeding (p = 0.008) and high post-porto-systemic pressure gradient (p = 0.004) had a significantly increased likelihood of developing HE. Also, patients having only one indication of bleeding or ascites were significantly unlikely to develop HE as well as Budd-Chiari disease (p = 0.03). The feature selection algorithm selected 7 features. Accuracy ratios for the SVM, LR and CatBoost, models were 74%, 75%, and 73%, with area under the curve (AUC) values of 0.82, 0.83, and 0.83, respectively. CONCLUSION: ML models can aid identifying patients at risk of developing HE after TIPS placement, providing an additional tool for patient selection and management.

Machine Learning Models in Prediction of Treatment Response After Chemoembolization with MRI Clinicoradiomics Features.

PURPOSE: To evaluate machine learning models, created with radiomics and clinicoradiomics features, ability to predict local response after TACE. MATERIALS AND METHODS: 188 treatment-naïve patients (150 responders, 38 non-responders) with HCC who underwent TACE were included in this retrospective study. Laboratory, clinical and procedural information were recorded. Local response was evaluated by European Association for the Study of the Liver criteria at 3-months. Radiomics features were extracted from pretreatment pre-contrast enhanced T1 (T1WI) and late arterial-phase contrast-enhanced T1 (CE-T1) MRI images. After data augmentation, data were split into training and test sets (70/30). Intra-class correlations, Pearson's correlation coefficients were analyzed and followed by a sequential-feature-selection (SFS) algorithm for feature selection. Support-vector-machine (SVM) models were trained with radiomics and clinicoradiomics features of T1WI, CE-T1 and the combination of both datasets, respectively. Performance metrics were calculated with the test sets. Models' performances were compared with Delong's test. RESULTS: 1128 features were extracted. In feature selection, SFS algorithm selected 18, 12, 24 and 8 features in T1WI, CE-T1, combined datasets and clinical features, respectively. The SVM models area-under-curve was 0.86 and 0.88 in T1WI; 0.76, 0.71 in CE-T1 and 0.82, 0.91 in the combined dataset, with and without clinical features, respectively. The only significant change was observed after inclusion of clinical features in the combined dataset (p = 0.001). Higher WBC and neutrophil levels were significantly associated with lower treatment response in univariant analysis (p = 0.02, for both). CONCLUSION: Machine learning models created with clinical and MRI radiomics features, may have promise in predicting local response after TACE. LEVEL OF EVIDENCE: Level 4, Case-control study.