Artificial intelligence for X-ray scaphoid fracture detection: a systematic review and diagnostic test accuracy meta-analysis.

OBJECTIVES: Scaphoid fractures are usually diagnosed using X-rays, a low-sensitivity modality. Artificial intelligence (AI) using Convolutional Neural Networks (CNNs) has been explored for diagnosing scaphoid fractures in X-rays. The aim of this systematic review and meta-analysis is to evaluate the use of AI for detecting scaphoid fractures on X-rays and analyze its accuracy and usefulness. MATERIALS AND METHODS: This study followed the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) and PRISMA-Diagnostic Test Accuracy. A literature search was conducted in the PubMed database for original articles published until July 2023. The risk of bias and applicability were evaluated using the QUADAS-2 tool. A bivariate diagnostic random-effects meta-analysis was conducted, and the results were analyzed using the Summary Receiver Operating Characteristic (SROC) curve. RESULTS: Ten studies met the inclusion criteria and were all retrospective. The AI's diagnostic performance for detecting scaphoid fractures ranged from AUC 0.77 to 0.96. Seven studies were included in the meta-analysis, with a total of 3373 images. The meta-analysis pooled sensitivity and specificity were 0.80 and 0.89, respectively. The meta-analysis overall AUC was 0.88. The QUADAS-2 tool found high risk of bias and concerns about applicability in 9 out of 10 studies. CONCLUSIONS: The current results of AI's diagnostic performance for detecting scaphoid fractures in X-rays show promise. The results show high overall sensitivity and specificity and a high SROC result. Further research is needed to compare AI's diagnostic performance to human diagnostic performance in a clinical setting. CLINICAL RELEVANCE STATEMENT: Scaphoid fractures are prone to be missed secondary to assessment with a low sensitivity modality and a high occult fracture rate. AI systems can be beneficial for clinicians and radiologists to facilitate early diagnosis, and avoid missed injuries. KEY POINTS: • Scaphoid fractures are common and some can be easily missed in X-rays. • Artificial intelligence (AI) systems demonstrate high diagnostic performance for the diagnosis of scaphoid fractures in X-rays. • AI systems can be beneficial in diagnosing both obvious and occult scaphoid fractures.

A Prospective, Single-Arm Study to Evaluate the Safety and Efficacy of an Autologous Blood Clot Product in the Treatment of Anal Fistula.

BACKGROUND: Surgical treatment of complex perianal fistula is technically challenging, associated with risk of failure, and may require multiple procedures. In recent years, several biologic agents have been developed for permanently eradicating anal fistulous disease with variable success. In this study, the treatment is an autologous whole-blood product created from the patients' blood. It forms a provisional matrix that was found to be safe and effective in healing acute and chronic cutaneous wounds. OBJECTIVE: The study aimed to assess the efficacy and safety of an autologous blood clot product as a treatment for transsphincteric perianal fistulas. DESIGN: A prospective single-arm study. SETTINGS: A single tertiary medical center. PATIENTS: Patients with simple or complex transsphincteric fistulas confirmed by MRI were included in the study. Cause was either cryptoglandular or Crohn's disease related (in the absence of active luminal bowel disease). INTERVENTION: The outpatient procedure was performed under general anesthesia and consisted of: 1) physical debridement and cleansing of the fistula tract; 2) suture closure of the internal opening; and 3) instillation of the autologous blood clot product into the entire tract. MAIN OUTCOME MEASURES: Safety and efficacy at 6- and 12-months after surgery. RESULTS: Fifty-three patients (77% men) with a median age of 42 (20-72) years were included in the study. Three patients withdrew consent, and 1 patient was lost to follow-up. At the time of this interim analysis, 49 and 33 patients completed the 6- and 12-month follow-up period. Thirty-four of the 49 patients achieved complete healing (69%) at 6 months, but 20 of the 33 patients (60%) achieved healing after 1 year. All patients who achieved healing at 6 months remained healed at the 1-year mark. In a subgroup analysis of patients with Crohn's disease, 7 of 9 patients completed 1-year follow-up, with 5 patients (71%) achieving clinical remission. No major side effects or postoperative complications were noted, but 2 adverse events occurred (admission for pain control and coronavirus 2019 infection). LIMITATIONS: Noncomparative single-arm pilot study. CONCLUSIONS: Treatment with an autologous blood clot product in perianal fistular disease was found to be feasible and safe, with an acceptable healing rate in both cryptoglandular and Crohn's disease fistula-in-ano. Further comparative assessment is required to determine its potential role in the treatment paradigm of fistula-in-ano. See Video Abstract . BRAZO PARA EVALUAR LA SEGURIDAD Y EFICACIA DE RDVER, UN COGULO DE SANGRE AUTLOGO, EN EL TRATAMIENTO DE LA FSTULA ANAL: ANTECEDENTES:El tratamiento quirúrgico de la fístula perianal compleja es técnicamente desafiante, se asocia con riesgo de fracaso y puede requerir múltiples procedimientos. En los últimos años, se han desarrollado varios agentes biológicos con el fin de erradicar permanentemente la enfermedad fistulosa anal con éxito variable. El tratamiento RD2-Ver.02 es un producto de sangre total autólogo creado a partir de la sangre de los pacientes, que forma una matriz provisional que resultó segura y eficaz para curar heridas cutáneas agudas y crónicas.OBJETIVO:Evaluar la eficacia y seguridad de RD2-Ver.02 como tratamiento para las fístulas perianales transesfinterianas.DISEÑO:Un estudio prospectivo de un solo brazo.LUGARES:Un único centro médico terciario.PACIENTES:Se incluyeron en el estudio pacientes con fístulas transesfinterianas simples o complejas confirmadas mediante resonancia magnética. La etiología fue criptoglandular o relacionada con la enfermedad de Crohn (en ausencia de enfermedad intestinal luminal activa).INTERVENCIÓN:El procedimiento ambulatorio se realizó bajo anestesia general y consistió en: 1) desbridamiento físico y limpieza del trayecto fistuloso; 2) cierre con sutura de la abertura interna; y 3) instilación de RD2-Ver.02 en todo el tracto.PRINCIPALES MEDIDAS DE VALORACIÓN:Seguridad y eficacia a los 6 y 12 meses después de la cirugía.RESULTADOS:Se incluyeron en el estudio 53 pacientes (77% varones) con una mediana de edad de 42 (20-72) años. Tres pacientes retiraron su consentimiento y un paciente se perdió durante el seguimiento. En el momento de este análisis intermedio, 49 y 33 pacientes completaron el período de seguimiento de 6 y 12 meses, respectivamente. Treinta y cuatro (34) pacientes lograron una curación completa (69%) a los 6 meses, mientras que 20 de 33 pacientes (60%) lograron una curación después de un año. Todos los pacientes que lograron la curación a los 6 meses permanecieron curados al año. En un análisis de subgrupos de pacientes con enfermedad de Crohn, 7/9 pacientes completaron un seguimiento de un año y 5 pacientes (71%) alcanzaron la remisión clínica. No se observaron efectos secundarios importantes ni complicaciones postoperatorias, mientras que ocurrieron 2 eventos adversos (ingreso para control del dolor e infección por COVID-19).LIMITACIONES:Estudio piloto no comparativo de un solo brazo.CONCLUSIONES:Se encontró que el tratamiento con RD2-Ver.02 en la enfermedad fístula perianal es factible y seguro, con una tasa de curación aceptable tanto en la fístula criptoglandular como en la de Crohn en el ano. Se requiere una evaluación comparativa adicional para determinar su papel potencial en el paradigma de tratamiento de la fístula anal. (Pre-proofed version ).

Deep learning in voice analysis for diagnosing vocal cord pathologies: a systematic review.

OBJECTIVES: With smartphones and wearable devices becoming ubiquitous, they offer an opportunity for large-scale voice sampling. This systematic review explores the application of deep learning models for the automated analysis of voice samples to detect vocal cord pathologies. METHODS: We conducted a systematic literature review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines. We searched MEDLINE and Embase databases for original publications on deep learning applications for diagnosing vocal cord pathologies between 2002 and 2022. Risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2). RESULTS: Out of the 14 studies that met the inclusion criteria, data from a total of 3037 patients were analyzed. All studies were retrospective. Deep learning applications targeted Reinke's edema, nodules, polyps, cysts, unilateral cord paralysis, and vocal fold cancer detection. Most pathologies had detection accuracy above 90%. Thirteen studies (93%) exhibited a high risk of bias and concerns about applicability. CONCLUSIONS: Technology holds promise for enhancing the screening and diagnosis of vocal cord pathologies. While current research is limited, the presented studies offer proof of concept for developing larger-scale solutions.

Disparities in Receipt of Adjuvant Therapy After Upfront Surgical Resection for Pancreatic Ductal Adenocarcinoma.

BACKGROUND: A multimodal approach of surgery and chemotherapy, with or without radiation, is the mainstay of therapy with curative-intent for resectable pancreatic ductal adenocarcinoma (PDAC). This study compared utilization trends and outcomes of upfront surgery with and without adjuvant therapy. METHODS: The National Cancer Database was queried for patients with a diagnosis of stage 1 or 2 PDAC who underwent upfront resection. Multivariable regression was applied to identify factors associated with initiation of adjuvant therapy. RESULTS: Of the 39,128 patients in the study, 67% initiated adjuvant therapy after resection, whereas 33% received upfront surgery alone. Receipt of adjuvant multimodal therapy increased from 59% in 2006 to 69% in 2017 (P < 0.0001). Non-white race was associated with lower odds of receiving adjuvant therapy after adjustment for income status, education attainment, and other variables (Hispanic/Spanish [odds ratio {OR}, 0.77; 95% confidence interval {CI}, 0.69-0.86] and non-Hispanic black [OR 0.84; 95% CI 0.78-0.91 vs non-Hispanic white; P < 0.001). The variables that contributed to receipt of adjuvant therapy were place of residence in high versus low education attainment area (OR 1.30; 95% CI 1.18-1.44; P < 0.0001) and lower odds for initiation of adjuvant therapy with increasing distance from the treating facility (> 50 miles [OR 0.51; 95% CI 0.47-0.54] vs <12.5 miles; P < 0.0001). The median unadjusted overall survival (OS) time was 18.2 months (95% CI 17.7-18.8 months) for upfront surgery alone and 25.3 months (95% CI 24.9-25.8 months) for surgery with adjuvant therapy. CONCLUSIONS: The patients who underwent upfront surgical resection for PDAC showed wide socioeconomic disparities in the use of adjuvant therapy independent of insurance status, facility type, or travel distance.

Deep learning for embryo evaluation using time-lapse: a systematic review of diagnostic test accuracy.

OBJECTIVE: This study aimed to investigate the accuracy of convolutional neural network models in the assessment of embryos using time-lapse monitoring. DATA SOURCES: A systematic search was conducted in PubMed and Web of Science databases from January 2016 to December 2022. The search strategy was carried out by using key words and MeSH (Medical Subject Headings) terms. STUDY ELIGIBILITY CRITERIA: Studies were included if they reported the accuracy of convolutional neural network models for embryo evaluation using time-lapse monitoring. The review was registered with PROSPERO (International Prospective Register of Systematic Reviews; identification number CRD42021275916). METHODS: Two reviewer authors independently screened results using the Covidence systematic review software. The full-text articles were reviewed when studies met the inclusion criteria or in any uncertainty. Nonconsensus was resolved by a third reviewer. Risk of bias and applicability were evaluated using the QUADAS-2 tool and the modified Joanna Briggs Institute or JBI checklist. RESULTS: Following a systematic search of the literature, 22 studies were identified as eligible for inclusion. All studies were retrospective. A total of 522,516 images of 222,998 embryos were analyzed. Three main outcomes were evaluated: successful in vitro fertilization, blastocyst stage classification, and blastocyst quality. Most studies reported >80% accuracy, and embryologists were outperformed in some. Ten studies had a high risk of bias, mostly because of patient bias. CONCLUSION: The application of artificial intelligence in time-lapse monitoring has the potential to provide more efficient, accurate, and objective embryo evaluation. Models that examined blastocyst stage classification showed the best predictions. Models that predicted live birth had a low risk of bias, used the largest databases, and had external validation, which heightens their relevance to clinical application. Our systematic review is limited by the high heterogeneity among the included studies. Researchers should share databases and standardize reporting.

Thromboprophylaxis Timing After Blunt Solid Organ Injury: A Systematic Review and Meta-analysis.

INTRODUCTION: Trauma patients with blunt abdominal solid organ injuries are at high risk for venous thromboembolism (VTE), but the optimal time to safely administer chemical thromboprophylaxis is controversial, especially for patients who are managed nonoperatively due to increased risk of hemorrhage. We sought to compare failure of nonoperative management (NOM) and VTE events based on timing of chemical thromboprophylaxis initiation. METHODS: A systematic review was conducted in PubMed and Embase databases. Studies were included if they evaluated timing of initiation of chemical thromboprophylaxis in trauma patients who underwent NOM of blunt solid organ injuries. Outcomes included failure of NOM and incidence of VTE. A random-effects meta-analysis was performed comparing patients who received late (>48 h) versus early thromboprophylaxis initiation. RESULTS: Twelve retrospective cohort studies, comprising 21,909 patients, were included. Three studies, including 6375 patients, provided data on adjusted outcomes. Pooled adjusted analysis demonstrated no difference in failure of NOM in patients receiving late versus early thromboprophylaxis (odds ratio [OR] 0.92, 95% confidence interval [CI]:0.4-2.14). When including all unadjusted studies, even those at high risk of bias, there remained no difference in failure of NOM (OR 1.16, 95% CI:0.72-1.86). In the adjusted analysis for VTE events, which had 6259 patients between two studies, patients receiving late chemical thromboprophylaxis had a higher risk of VTE compared with those who received early thromboprophylaxis (OR 1.89, 95% CI:1.15-3.12). CONCLUSIONS: Based on current observational evidence, initiation of prophylaxis before 48 h is associated with lower VTE rates without higher risk of failure of NOM.

The value of CA125 in predicting acute complicated colonic diverticulitis.

BACKGROUND: CA125 is a widely used serum marker for epithelial ovarian cancer which levels may also rise in benign conditions involving peritoneal irritation. We aimed to determine if serum CA125 levels can predict disease severity in patients presenting with acute diverticulitis. METHODS: We conducted a single-center prospective observational study, analyzing CA125 serum levels in patients who presented to the emergency department with computerized tomography-proven acute left-sided colonic diverticulitis. Univariate, multivariate, and receiver operating characteristic (ROC) analyses were used to correlate CA125 serum levels at time of initial presentation with the primary outcome (complicated diverticulitis) and secondary clinical outcomes (need for urgent intervention, length of hospital stay (LOS) and readmission rates). RESULTS: One hundred and fifty-one patients were enrolled between January 2018 and July 2020 (66.9% females, median age 61 years). Twenty-five patients (16.5%) presented with complicated diverticulitis. CA125 levels were significantly higher among patients with complicated (median: 16 (7-159) u/ml) vs. uncomplicated (8 (3-39) u/ml) diverticulitis (p < 0.001) and also correlated with the Hinchey severity class (p < 0.001). Higher CA125 levels upon admission were associated with a longer LOS and a greater chance to undergo invasive procedure during the hospitalization. In patients with a measurable intra-abdominal abscess (n = 24), CA125 levels were correlated with the size of the abscess (Spearman's r = 0.46, p = 0.02). On ROC analysis to predict complicated diverticulitis, the area under the curve (AUC) for CA125 (AUC = 0.82) was bigger than for the leukocyte count (AUC = 0.53), body temperature (AUC = 0.59), and neutrophil-lymphocyte ratio (AUC = 0.70) - all p values < 0.05. On multivariate analysis of factors available at presentation, CA125 was found to be the only independent predictor of complicated diverticulitis (OR 1.12 (95% CI 1.06-1.19), p < 0.001). CONCLUSIONS: The results from this feasibility study suggest that CA125 may accurately discriminate between simple and complicated diverticulitis, meriting further prospective investigation.

Neutrophil to Lymphocyte Ratio in Patients Who Received Neoadjuvant Treatment before Gastrectomy.

BACKGROUND: The neutrophil to lymphocyte ratio (NLR) has demonstrated prognostic value in various malignant conditions, including gastric adenocarcinoma. However, chemotherapy may affect NLR. OBJECTIVES: To evaluate the prognostic value of NLR as an accessory decision-making tool in terms of operating patients after neoadjuvant chemotherapy in patients with resectable gastric cancer. METHODS: We collected oncologic, perioperative, and survival data of patients with gastric adenocarcinoma who underwent curative intent gastrectomy and D2 lymphadenectomy between 2009 and 2016. The NLR was calculated from preoperative laboratory tests and classified as high (> 4) and low (≤ 4). The t-test, chi-square, Kaplan-Meier analysis, and Cox multivariate regression models were used to assess associations of clinical, histologic, and hematological variables with survival. RESULTS: For 124 patients the median follow-up was 23 months (range 1-88). High NLR was associated with greater rate of local complication (r=0.268, P < 0.01). The rate of major complications (Clavien-Dindo ≥ 3) was higher in the high NLR group (28% vs. 9%, P = 0.022). Among the 53 patients who received neoadjuvant chemotherapy, those with low NLR had significantly improved disease-free survival (DFS) (49.7 vs. 27.7 months, P = 0.025). Low NLR was not significantly associated with overall survival (mean survival, 51.2 vs. 42.3 months, P = 0.19). Multivariate regression identified NLR group (P = 0.013), male gender (P = 0.04), and body mass index (P = 0.026) as independently associated with DFS. CONCLUSIONS: Among gastric cancer patients planned for curative intent surgery who underwent neoadjuvant chemotherapy, NLR may have prognostic value, particularly regarding DFS and postoperative complications.

[SURGICAL TREATMENT USING ROBOTIC APPROACH FOR OBSTRUCTED DEFECATION SYNDROME (ODS)].

INTRODUCTION: Obstructed Defecation Syndrome (ODS) is a complex surgical condition which involves structural and functional problems which significantly affects quality of life. There is limited information about the Da-Vinci Robotic System use in the treatment of this condition. This study examines the outcomes of robotic-assisted rectopexy. The primary outcome is recovery from surgery, with secondary outcomes including post-surgical complications, length of hospital stay, rehospitalization rate and recurrence after surgery. METHODS: A retrospective analysis was conducted of prospectively collected data for patients who underwent robotic assisted surgery for ODS between 2011-2022. A colorectal surgeon performed all surgeries at the Sheba Medical Center using the Da Vinci™ robotic system. This analysis uses descriptive statistics and presents the results as medians and ranges. RESULTS: Out of 33 patients included, 26 (84.9%) were female. Median age was 67 years (Range:19-85 years). Median American Society of Anesthesiology (ASA) score was 2 (1-3). Median Charlson's comorbidity score was 3 (0-4). Median patients' Body Mass Index (BMI) was 23.2 (15.6-33.4) kg/m2. Eight patients (24.4%) underwent previous procedure for ODS. Most (23) patients included (69.7%) underwent robotic assisted ventral rectopexy. Other interventions included combined anterior and posterior rectopexy (9.1%), combined ventral rectopexy and sacrocolpopexy (12.1%) and posterior rectopexy (9.1%). No cases of conversion to laparoscopic /open techniques were recorded. Median operation time was 135 minutes (70-270). One intra-operative complication recorded was an injury to the rectum during anterior dissection (3%). No significant blood loss was recorded. A total of 27 patients (81.8%) were operated using the Da Vinci Si system, and the rest (6) using the 6 Da Vinci Xi system. Two patients had post-operative complications. Median length of stay (LOS) was 4 days (2-6 days). Readmission rate within 30 days was 9.1%. Two patients (6.1%) had recurrence of rectal prolapse. Median follow-up was 60 (4-116) months. CONCLUSIONS: Robotic-assisted surgery for obstructed defecation syndrome is safe, with fast recovery of the patient and it is efficient during long-term follow-up.

Perioperative Pharmacologic Prophylaxis of Venous Thromboembolism: A Professional Liability Analysis.

BACKGROUND: Fear of medical liability is a major driver for broad administration of perioperative prophylactic anticoagulation, despite the persistently low rates of clinically symptomatic venous thromboembolism events (VTE) postoperatively. This study was undertaken to evaluate the medicolegal landscape of perioperative VTE and its pharmaceutical prophylaxis. METHODS: The Westlaw legal database was retrospectively searched for verdicts in medical professional liability cases in the United States between 2009 and 2020. One search strategy focused on perioperative VTE, and a second on claims of hemorrhagic complications in patients receiving perioperative anticoagulation. RESULTS: The search for VTE revealed 129 cases, and the search for hemorrhagic complications identified 24 cases. Almost half of the VTE cases were brought following orthopedic surgery (49%), and 29% following general surgery. The most common claims were failure to diagnose and treat during hospital stay or after discharge (74%), and failure to prescribe/administer anticoagulation (46%). Verdict for the health care professional (i.e., the defendant) was reached in 75% of cases. The median payout for patient verdicts was 1,213,644 USD (interquartile range 1,014,100; 150,000-7,700,000). Of hemorrhagic complication cases, 42% occurred in patients receiving VTE prophylaxis. In these cases, 82% resulted in a defendant verdict. CONCLUSIONS: Reasons for in-court medical professional liability claims involving perioperative VTE were mainly failure to diagnose VTE and rescue patients from complications postoperatively. The high rate of defendant verdicts supports the notion that decisions on VTE prophylaxis should not be influenced by fear of liability.

Challenges of virtual interviewing for surgical fellowships: a qualitative analysis of applicant experiences.

BACKGROUND: The COVID-19 pandemic forced surgical fellowship programs to transition from in-person to remote applicant interviews; the virtual interviewing format presented new and unique challenges. We sought to understand applicants' perceived challenges to virtual interviewing for a surgical fellowship program. METHOD: A grounded theory-based qualitative study was performed utilizing semi-structured interviews with fellowship applicants from the 2020 fellowship match. All Fellowship Council-registered applicants were eligible. We purposefully sampled participants to balance across gender, specialty-choice, and academic versus community-program affiliation. Interviews were inductively analyzed by two researchers for prominent themes. RESULTS: Fifteen interviews were conducted. Participants were 60% male (n = 9), with 33% (n = 5) from non-academic institutions. They applied for the following fellowships: Advanced Gastrointestinal/Minimal Invasive (55%), Bariatric (30%), Hepatopancreatobiliary (10%) and Surgical Oncology (5%). Four main themes emerged to describe virtual interview process challenges: (1) perceived data deficiency, (2) superficial personal connections, (3) magnification of non-professionalism, and (4) logistical frustrations. Applicants recommend program directors provide more information about the fellowship prior to interview day and offer informal independent interactions with current and previous fellows. CONCLUSIONS: According to fellowship applicants, virtual interviews resulted in a lack of information for rank-list decision making ultimately requiring them to rely on other information avenues to base their decisions. These applicants have offered advice to fellowship program directors and future applicants to better optimize this process.

Analysis of in court malpractice litigation following pancreatic surgery.

BACKGROUND: Pancreatic operations are technically complex with a significant risk of morbidity and mortality. Analysis of litigation following surgery can disclose avoidable errors. METHODS: Two legal databases, Westlaw and Lexis Nexis, were queried for jury verdicts, settlements and appeal cases in the United States related to pancreatic surgery and malpractice between 1980 and 2020. RESULTS: Thirty-four cases were analyzed. Pancreaticoduodenectomy (n = 22, 65%) was the most common procedure litigated. Claims most commonly involved malpractice in the postoperative (n = 16) setting. The most common claims were failure to diagnose or treat postoperative complications (n = 16), lack of informed consent (n = 9), and wrongful indication for surgery (n = 8). Seven cases involved incorrect diagnosis of pancreatic cancer. Attending surgeons were the main provider named in the medical malpractice claim in 25 (74%) of the cases. Resident surgeons were named in 5 cases (15%). Half of the cases (n = 17) involved patient deaths. The court ruled in favor of the defendant in 75% of the cases, and the plaintiff in 12.5%. Median payout for settlements and plaintiff verdicts was 783,304 USD (interquartile range (IQR) 1,034,046). Mean time from incident to final disposition was 6.4 years (±3.3). CONCLUSIONS: Obtaining an accurate preoperative diagnosis may decrease malpractice litigation following pancreatic surgery. By raising awareness to unsafe practices and identifying vulnerable periods of care, these data may serve to enhance provider performance as well as improve patient safety.

Deep learning for noninvasive liver fibrosis classification: A systematic review.

BACKGROUND AND AIMS: While biopsy is the gold standard for liver fibrosis staging, it poses significant risks. Noninvasive assessment of liver fibrosis is a growing field. Recently, deep learning (DL) technology has revolutionized medical image analysis. This technology has the potential to enhance noninvasive fibrosis assessment. We systematically examined the application of DL in noninvasive liver fibrosis imaging. METHODS: Embase, MEDLINE, Web of Science, and IEEE Xplore databases were used to identify studies that reported on the accuracy of DL for classification of liver fibrosis on noninvasive imaging. The search keywords were "liver or hepatic," "fibrosis or cirrhosis," and "neural or DL networks." Risk of bias and applicability were evaluated using the QUADAS-2 tool. RESULTS: Sixteen studies were retrieved. Imaging modalities included ultrasound (n = 10), computed tomography (n = 3), and magnetic resonance imaging (n = 3). The studies analyzed a total of 40 405 radiological images from 15 853 patients. All but two of the studies were retrospective. In most studies the "ground truth" reference was the METAVIR score for pathological staging (n = 9.56%). The majority of the studies reported an accuracy >85% when compared to histopathology. Fourteen studies (87.5%) had a high risk of bias and concerns regarding applicability. CONCLUSIONS: Deep learning has the potential to play an emerging role in liver fibrosis classification. Yet, it is still limited by a relatively small number of retrospective studies. Clinicians should facilitate the use of this technology by sharing databases and standardized reports. This may optimize the noninvasive evaluation of liver fibrosis on a large scale.

Development and Utilization of a Medical Student Surgery Podcast During COVID-19.

BACKGROUND: The 2019 coronavirus disease (COVID-19) pandemic drastically reduced learning opportunities for medical students. We sought to determine the cost and success of implementation of a podcast for a surgical department in a large academic hospital. METHODS: We created a podcast series for Israeli medical students during the COVID-19 epidemic based on the Medical Student Core Curriculum of the American College of Surgeons / Association for Surgical Education. Episodes were available for free download or streaming on a designated website and popular podcast platforms. Podcast analytics were used to measure public listeners and uptake. RESULTS: Total development time was 90 hours at an estimated cost of $7091 USD. A total of 10 episodes were released between March 21, 2020 and August 31, 2020. An average of 9 ± 1.26 h (range 2-6) was required to generate each episode, including 3.4 ± 1.26 h (2-6) for content review and 5.6 ± 2 h (4-10) for audio production. An average episode ran for 35.9 ± 4.3 min (28-42). Podcasts recorded a total of 5678 downloads, with an average of 228 and 336 downloads per episode in the first 30 and 90 days, respectively. The average daily downloads before the students returned to clinical rotations (March 21-April 30) was 48 ± 58.3 (7-283;) compared to 16 ± 7.4 after their return (1-38; P< 0.01). Estimated costs to produce a video-based education series would have been significantly more. CONCLUSION: Podcasts can serve as a cost-effective and quickly produced instructional tool to supplement online learning. Further research is required to determine the efficacy of podcasts versus video-based education modules.

Low psoas muscle area is associated with postoperative complications in Crohn's disease.

BACKGROUND AND AIMS: Crohn's disease (CD) is associated with increased postoperative morbidity. Sarcopenia correlates with increased morbidity and mortality in various medical conditions. We assessed correlations of the lean body mass marker and psoas muscle area (PMA), with postoperative outcomes in CD patients undergoing gastrointestinal surgery. METHODS: We included patients with CD who underwent gastrointestinal surgery between June 2009 and October 2018 and had CT/MRI scans within 8 weeks preoperatively. PMA was measured bilaterally on perioperative imaging. RESULTS: Of 121 patients, the mean age was 35.98 ± 15.07 years; 51.2% were male. The mean BMI was 21.56 ± 4 kg/m2. The mean PMA was 95.12 ± 263.2cm2. Patients with postoperative complications (N = 31, 26%) had significantly lower PMA compared with patients with a normal postoperative recovery (8.5 ± 2.26 cm2 vs. 9.85 ± 2.68 cm2, P = 0.02). A similar finding was noted comparing patients with anastomotic leaks to those without anastomotic leaks (7.48 ± 0.1 cm2 vs. 9.6 ± 2.51 cm2, P = 0.04). PMA correlated with the maximum degree of complications per patient, according to the Clavien-Dindo classification (Spearman's coefficient = -0.26, P = 0.004). Patients with major postoperative complications (Clavien-Dindo ≥ 3) had lower mean PMA (8.12 ± 2.75 cm2 vs. 9.71 ± 2.57 cm2, P = 0.03). Associations were similar when stratifying by gender and operation urgency. On multivariate analysis, PMA (HR = 0.72/cm2, P = 0.02), operation urgency (HR = 3.84, P < 0.01), and higher white blood cell count (HR = 1.14, P = 0.02) were independent predictive factors for postoperative complications. CONCLUSION: PMA is an easily measured radiographic parameter associated with postoperative complications in patients with CD undergoing bowel resection.

Deep learning visual analysis in laparoscopic surgery: a systematic review and diagnostic test accuracy meta-analysis.

BACKGROUND: In the past decade, deep learning has revolutionized medical image processing. This technique may advance laparoscopic surgery. Study objective was to evaluate whether deep learning networks accurately analyze videos of laparoscopic procedures. METHODS: Medline, Embase, IEEE Xplore, and the Web of science databases were searched from January 2012 to May 5, 2020. Selected studies tested a deep learning model, specifically convolutional neural networks, for video analysis of laparoscopic surgery. Study characteristics including the dataset source, type of operation, number of videos, and prediction application were compared. A random effects model was used for estimating pooled sensitivity and specificity of the computer algorithms. Summary receiver operating characteristic curves were calculated by the bivariate model of Reitsma. RESULTS: Thirty-two out of 508 studies identified met inclusion criteria. Applications included instrument recognition and detection (45%), phase recognition (20%), anatomy recognition and detection (15%), action recognition (13%), surgery time prediction (5%), and gauze recognition (3%). The most common tested procedures were cholecystectomy (51%) and gynecological-mainly hysterectomy and myomectomy (26%). A total of 3004 videos were analyzed. Publications in clinical journals increased in 2020 compared to bio-computational ones. Four studies provided enough data to construct 8 contingency tables, enabling calculation of test accuracy with a pooled sensitivity of 0.93 (95% CI 0.85-0.97) and specificity of 0.96 (95% CI 0.84-0.99). Yet, the majority of papers had a high risk of bias. CONCLUSIONS: Deep learning research holds potential in laparoscopic surgery, but is limited in methodologies. Clinicians may advance AI in surgery, specifically by offering standardized visual databases and reporting.

Management of Acute Appendicitis during the COVID-19 Pandemic: A Single Tertiary Center Experience.

BACKGROUND: The coronavirus disease-2019 (COVID-19) outbreak had an effect on healthcare. OBJECTIVES: To evaluate the presentation and management of patients with acute appendicitis. METHODS: A retrospective study was conducted of all patients presenting with acute appendicitis to the emergency department of a large tertiary center during March and April 2020. Clinical features, diagnostic workup, and management were compared. RESULTS: Seventy-four patients presented with acute appendicitis during the pandemic compared to 60 patients during the same time the year before. There were no significant differences in patient demographics: age (P = 0.65), gender (P = 0.73), smoking status (P = 0.48). During COVID-19 patients were more likely to complain of right lower quadrant pain (100% vs. 78.3%, P < 0.01). Rates of surgical treatment was similar (83.8% vs. 81.7%, P = 1); mean operative time was longer during COVID-19 (63 ± 23 vs. 52 ± 26 minutes, P = 0.03). There were no significant differences in intra-operative findings including the presence of appendiceal perforation (16.3% vs. 14.5%, P = 0.8), abscess (6.1% vs. 9.7%, P = 0.73), or involvement of cecum or terminal ileum (14.28% vs. 19.63%, P = 1). Postoperative treatment with antibiotics was more prevalent during COVID-19 (37.1% vs. 18%, P = 0.04). Length of stay (1.82 ± 2.04 vs. 2.74 ± 4.68, P = 0.2) and readmission rates (6% vs. 11.3%, P =0.51) were similar. CONCLUSIONS: The COVID-19 pandemic did not significantly affect the presentation, clinical course, management, and outcomes of patients presenting with acute appendicitis.

The Effect of the COVID-19 Outbreak on Trauma-Related Visits to a Tertiary Hospital Emergency Department.

BACKGROUND: The novel coronavirus disease (COVID-19) pandemic changed medical environments worldwide. OBJECTIVES: To evaluate the impact of the COVID-19 pandemic on trauma-related visits to the emergency department (ED). METHODS: A single tertiary center retrospective study was conducted that compared ED attendance of patients with injury-related morbidity between March 2020 (COVID-19 outbreak) and pre-COVID-19 periods: February 2020 and the same 2 months in 2018 and 2019. RESULTS: Overall, 6513 patients were included in the study. During the COVID-19 outbreak, the daily number of patients visiting the ED for acute trauma declined by 40% compared to the average in previous months (P < 0.01). A strong negative correlation was found between the number of trauma-related ED visits and the log number of confirmed cases of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in Israel (Pearson's r = -0.63, P < 0.01). In the COVID-19 period there was a significant change in the proportion of elderly patients (7% increase, P = 0.002), admissions ratio (12% increase, P < 0.001), and patients brought by emergency medical services (10% increase, P < 0.001). The number of motor vehicle accident related injury declined by 45% (P < 0.01). CONCLUSIONS: A significant reduction in the number of trauma patients presenting to the ED occurred during the COVID-19 pandemic, yet trauma-related admissions were on the rise.

BETA-2 score is an early predictor of graft decline and loss of insulin independence after pancreatic islet allotransplantation.

This study aimed to evaluate whether the BETA-2 score is a reliable early predictor of graft decline and loss of insulin independence after islet allotransplantation. Islet transplant procedures were stratified into 3 groups according to clinical outcome: long-term insulin independence without islet graft decline (group 1, N = 9), initial insulin independence with subsequent islet graft decline and loss of insulin independence (group 2, N = 13), and no insulin independence (group 3, N = 13). BETA-2 was calculated on day 75 and multiple times afterwards for up to 145 months posttransplantation. A BETA-2 score cut-off of 17.4 on day 75 posttransplantation was discerned between group 1 and groups 2 and 3 (area under the receiver operating characteristic 0.769, P = .005) with a sensitivity and negative predictive value of 100%. Additionally, BETA-2 ≥ 17.4 at any timepoint during follow-up reflected islet function required for long-term insulin independence. While BETA-2 did not decline below 17.4 for each of the 9 cases from group 1, the score decreased below 17.4 for all transplants from group 2 with subsequent loss of insulin independence. The reduction of BETA-2 below 17.4 predicted 9 (1.5-21) months in advance subsequent islet graft decline and loss of insulin independence (P = .03). This finding has important implications for posttransplant monitoring and patient care.