RESUMEN
PURPOSE: Traumatic elbow instability (subluxation and dislocation) represents a challenging clinical entity that can be complicated by persistent intraoperative instability after attempted stabilization or recurrent postoperative instability. Both static and dynamic supplemental stabilization procedures have been described for cases where fracture fixation and ligament repair fail to restore stability. There is a paucity of prior studies involving ulnohumeral cross-pinning (UCP), which is a type of static supplemental stabilization. Our purpose was to assess complications and outcomes after UCP. METHODS: We reviewed all surgical cases involving primary and revision UCP for traumatic elbow instability at a single center from 2017-2023. Baseline demographics were recorded. Outcomes including radiographs, range of motion, patient-reported outcome measures, and surgical complications were analyzed. RESULTS: Fourteen patients undergoing UCP were included with a mean follow-up of 27 months. Five cross-pinnings (36%) were performed during revision procedures. Mean visual analog scale (VAS) pain, Disabilities of the Arm, Shoulder, and Hand (QuickDASH), and the Single Assessment Numeric Evaluation (SANE) scores were 2.4, 34, and 69, respectively. The mean flexion-extension and pronation-supination arcs were 114° and 140°, with 85% achieving at least a 100° arc for flexion and forearm rotation. Five patients (36%) had complications, all of which required reoperation. Two complications involved postoperative instability: one radial head subluxation and one radial head dislocation. Both occurred in revision UCP cases indicated for recurrent postoperative instability. CONCLUSIONS: Ulnohumeral cross-pinning for persistent and recurrent elbow instability results in maintained ulnohumeral joint alignment, functional arcs of elbow range of motion, and acceptable patient-reported outcome measures, particularly in the setting of a primary procedure indicated for persistent intraoperative instability. Ulnohumeral cross-pinning is a reasonable supplemental stabilization procedure for complex elbow instability. TYPE OF STUDY/LEVEL OF EVIDENCE: Therapeutic IV.
RESUMEN
Introduction The creation of research groups and consortiums has become more common in all medical and surgical specialities. The purpose of this investigation was to assess and describe collaborative research groups and consortiums within orthopaedic surgery. In addition, we aimed to define the demographics of the research consortium members with particular attention to female and minority members. Methods Journals with a musculoskeletal/orthopaedic focus and a few medical journals were selected to identify articles published by research groups and consortiums. Articles published from 2020 to 2022 were manually reviewed. Bibliographic information, author information and level of evidence (LOE) were recorded. For identified consortium members, sex and race were defined in a binary manner. Results A total of 92 research consortiums were identified. A list of members was identified for 77 groups (83.7%), totalling 2,260 researchers. The remaining group members were not able to be identified due to the lack of information in the included publications, research group websites or after communicating with the corresponding author for respective articles. Most researchers were male (n=1,748, 77.3%) and white (n=1,694, 75%). Orthopaedic surgeons comprised 1,613 (71.4%) identified researchers. The most common fellowship training for orthopaedic surgeons was paediatrics (n=370, 16.4%), trauma (n=266, 11.8%) and sports medicine (n=229, 10.1%). The consortiums published 261 articles: women were lead (first) authors in 23% and senior (last) authors in 11.1%. Non-white researchers were lead authors in 24.5% (n=64) and senior authors in 17.2% (n=45). The most common level of evidence was level 3, accounting for 45.6% (n=119) of all publications. Level 1 evidence accounted for 12.6% (n=33) of published articles. Discussion Representation of women in orthopaedic research consortiums exceeds their representation in almost every orthopaedic professional society. There is less publicly available data to compare the involvement of under-represented minorities (URMs) in research consortiums to general practice. Further investigations should analyse possible avenues in which gender and racial disparity could be improved within orthopaedic surgery research.
RESUMEN
Purpose: Bibliometric analysis is a common method for evaluating current trends within a scientific field. The primary aim of this study was to define and analyze the 50 most frequently cited articles in the field of elbow surgery, both of all time and those published during the 21st century. Methods: We searched the Journal Citation Report to identify articles related to elbow surgery within academic journals. Articles were sorted by total citations. The overall top 50 articles and those published since 2000 were identified, and data were collected, including title, journal of publication, publication year, country of publication, citation density, level of evidence, article type, institution, and sex of the lead and senior authors, and inclusion on the reference list for the Orthopaedic In-Training Examination within the last 5 years. Descriptive statistics were reported, and correlation analysis was performed using Spearman test. Results: For the most-cited elbow surgery articles, "fracture" was overall the most reported topic, whereas "lateral epicondylosis" and "fracture" were equal for those published since 2000. The United States was the most represented overall and for articles published since 2000. Women comprised 1/50 (2%) of lead authors overall, increasing to 8/50 (16%) for articles published during the 21st century. Most articles in during both periods contained level IV evidence, with level I evidence appearing infrequently (4%). Six percent of the most-cited articles of all time had appeared on the reference list of the Orthopaedic In-Training Examination within the past 5 years. Conclusions: The top 50 most-cited elbow surgery articles often assess fracture and lateral epicondylosis, most commonly originating from the United States. Level IV retrospective series comprises over half of the articles on this list. Women remain underrepresented as authors. Clinical Relevance: This study provides a modern reading list for upper-extremity surgeons about impactful elbow surgery articles.
RESUMEN
Purpose: To identify patient, surgeon, and injury characteristics associated with preoperative computed tomography (CT) scan utilization for operative distal radius fractures (DRF). In addition, we aimed to determine if preoperative CT was associated with treatment methods other than isolated volar-locked plating (VLP). Methods: We retrospectively reviewed all operatively treated adult DRFs within our health care system from 2016 to 2020. Baseline demographics, injury, treatment characteristics, and the fellowship training of the 44 included surgeons were recorded. We compared cases with and without a preoperative CT, and an adjusted logistic regression model was generated to determine the odds of having a preoperative CT. Results: A total of 1,204 operatively treated DRFs performed by 44 surgeons were included. CT utilization increased during the study period. Intra-articular fractures accounted for 76% of cases, and preoperative CT scans were ordered in 243 of 1240 cases (20%). Overall, isolated VLP was used in 83% of cases. Cases with a preoperative CT were more likely to be treated with an alternative method of fixation (such as dorsal plating). The adjusted logistic regression model demonstrated that male sex (OR 1.62; 95% CI: 1.16, 2.26), intra-articular fractures (OR 3.11; 95% CI: 1.87, 5.81), and associated fractures (OR 2.69; 95% CI: 1.82, 3.98) had a significantly increased odds of having a preoperative CT. Fellowship training was not associated with increased CT utilization overall, but hand surgeons were more likely to use a CT in Orthopaedic Trauma Association-C3 fractures. Conclusions: Patient and injury characteristics are associated with CT utilization in operative DRFs. Preoperative CTs are associated with alternative fixation approaches, as cases with a CT were more likely to use fixation methods other than isolated VLP. The costs and benefits of CT scans must be carefully weighed against whether this modality adds value or improves outcomes in treating DRFs. Level of evidence: Prognostic II.
RESUMEN
Introduction: The use of telemedicine (TM) for patient care greatly increased during the COVID pandemic. This study presents data from a single health system regarding physician's perspectives on TM, which could ultimately determine how it is used in the future. Methods: A questionnaire was distributed to physicians throughout the health system. Physicians were divided based on the standard level of patient interaction in each specialty, as well as practice locations and years in practice. Physician perspectives were categorized by their opinions on different aspects of telehealth visits. Results: Of 1,794 physicians, 379 (20.7%) responded to the survey. Psychiatrists used TM significantly more than other groups and project the most future use. Surgeons were least likely to incorporate TM in the future. Ability to perform a physical examination via TM differed significantly by specialty and practice environment, but not by years in practice. Frequency of being able to complete a treatment plan via TM differed significantly by specialty, but not by years in practice or practice environment. Overall, 76.3% of physicians reported feeling "satisfied" with performing TM visits. Satisfaction with TM varied significantly by specialty and practice environment, but not by years in practice. There were no significant differences regarding physician expectations on reimbursement or billing for TM visits based on specialty, age, or practice environment. Conclusions: Discrepancies exist among physicians with respect to their satisfaction and expected future use of TM. Consensus may be difficult to reach regarding reimbursement for these visits, and further work is needed to clarify the optimal practice setting for TM.
Asunto(s)
COVID-19 , Prestación Integrada de Atención de Salud , Cirujanos , Telemedicina , Humanos , COVID-19/epidemiología , Pandemias , Telemedicina/métodosRESUMEN
PURPOSE: To assess the outcomes of operatively treated terrible triad (TT) elbow injuries for a single surgeon at the start of clinical practice. We aimed to define postoperative patient reported outcome measures (PROMs), range of motion (ROM), and complications during the period immediately following fellowship training, in order to describe the learning process for surgical treatment of TT. Methods: All operatively treated TTs from 2017 to 2020 were included. All cases were performed by a single, fellowship-trained upper-extremity surgeon and represented a consecutive series at the start of clinical practice. Baseline demographics, injury characteristics, and surgical details were recorded for each case. PROMs [QuickDisability of arm, shoulder, and hand (DASH) and (visual analog scale) pain scale], ROM, and complications were recorded at the time of final follow-up. A perioperative glucocorticoid protocol was used in all cases without diabetes. RESULTS: There was a total of 21 included TT cases with a mean follow-up of 20 months. The operative time averaged 89 min for the first 10 cases and 83 min for the subsequent 11 cases. The mean QuickDASH and VAS pain score at final follow-up were 19 and 2.3, respectively. The mean flexion-extension arc was 122° and two cases (9%) had < 100° arc of motion. The mean pronation-supination arc was 145°. Three cases (14%) had a postoperative complication, all of which underwent reoperation. Of the 21 included cases, these reoperations represented cases #1, #14, and #17 respectively. Conclusions: Upper-extremity surgeons at the start of clinical practice may be able to achieve outcomes similar to more experienced surgeons for operatively treated TT elbow fracture dislocations. There does not appear to be a substantial "learning curve" after fellowship training with respect to PROMs, complication rates, or operative time associated with surgical treatment of TT elbow injuries.
RESUMEN
OBJECTIVE: We assessed the readability of spine-related patient education materials on professional society websites to determine whether this had improved since last studied. We also compared the readability of these materials to a more patient-centered source, such as WebMD. METHODS: Patient education pages from the American Association of Neurologic Surgeons (AANS), North American Spine Society (NASS), and spine-related pages from the American Academy of Orthopaedic Surgeons (AAOS), and WebMD were reviewed. Readability was evaluated using the Flesch Kincaid Grade Level (FKGL) and Flesch Reading Ease (FRE) formulas. The mean FKGL and FRE scores of the societies were compared using one-way analysis of variance. The rate of a reading level at or below an eighth grade level was compared using the χ2 test. RESULTS: We analyzed a total of 156 sites. The mean FKGL score for the professional society sites was 11.4. The mean FRE score for the professional societies was 45.8, with 14.4% written at or below an eighth grade reading level. We found a significant difference in the FKGL scores and materials at or below the eighth grade level between the AAOS and AANS and AAOS and NASS. The mean FKGL and FRE scores for WebMD were 7.57 and 68.1, respectively, with a significant difference compared with the scores for the AAOS, NASS, and AANS. In addition, 80% of the WebMD materials had been written at or below the eighth grade reading level. A significant difference compared with the AANS and NASS (P < 0.0001) but not for the AAOS (P = 0.059). CONCLUSIONS: The average readability of spine-related topics exceeded the eighth grade reading level. The AAOS resources had better readability compared with the NASS and AANS. We found no improvement in readability since last studied. The readability of professional societies' materials was significantly worse than those from WebMD.
Asunto(s)
Comprensión , Alfabetización en Salud , Academias e Institutos , Humanos , Internet , Educación del Paciente como Asunto , Columna Vertebral , Estados UnidosRESUMEN
BACKGROUND: Several post hoc analyses of randomized controlled trials (RCTs) suggested the importance of microsatellite instability (MSI) as a positive predictive factor to immunotherapy in patients with advanced gastric cancer (GC); however, individually these have low statistical power. METHODS: RCTs investigating treatment with or without an anti-programmed cell death protein 1 (PD-1) agent for advanced GC and providing outcome according to MSI status were selected. The hazard ratio (HR) and the odds ratio were used to compare the treatment effect on survival outcomes and tumor response, respectively, for anti-PD-1-based therapy compared with standard therapy. Evidence for treatment effect by MSI status was evaluated by a test of interaction. RESULTS: The phase III KEYNOTE-062, CheckMate-649, JAVELIN Gastric 100 and KEYNOTE-061 trials were included. A total of 2545 patients with evaluable MSI status were included and 123 (4.8%) had MSI-high cancers. The HR for overall survival benefit with anti-PD-1-based regimens was 0.34 (95% CI: 0.21-0.54) for MSI-high cancers versus 0.85 [95% confidence interval (CI): 0.71-1.00] for microsatellite stable. The treatment effect was significantly different in the two subgroups (P for interaction 0.003). In the MSI-high subgroup, the HR for progression-free survival was 0.57 (95% CI: 0.33-0.97; P = 0.04) and the odds ratio for response was 1.76 (95% CI: 1.10-2.83; P = 0.02). CONCLUSIONS: Patients with MSI-high GC should be regarded as a specific and highly immunosensitive population worthy of dedicated clinical trials.
Asunto(s)
Inestabilidad de Microsatélites , Neoplasias Gástricas , Humanos , Receptor de Muerte Celular Programada 1/genética , Supervivencia sin Progresión , Ensayos Clínicos Controlados Aleatorios como Asunto , Neoplasias Gástricas/tratamiento farmacológico , Neoplasias Gástricas/genéticaRESUMEN
PURPOSE: To evaluate the potential of imaging criteria in predicting overall survival of patients with hepatocellular carcinoma (HCC) after a first transcatheter arterial yttrium-90 radioembolization (TARE) MATERIALS AND METHODS: From October 2013 to July 2017, 37 patients with HCC were retrospectively included. There were 34 men and 3 women with a mean age of 60.5±10.2 (SD) years (range: 32.7-78.9 years). Twenty-five patients (68%) were Barcelona Clinic Liver Cancer (BCLC) C and 12 (32%) were BCLC B. Twenty-four primary index tumors (65%) were>5cm. Three radiologists evaluated tumor response on pre- and 4-7 months post-TARE magnetic resonance imaging or computed tomography examinations, using Response Evaluation Criteria in Solid Tumors (RECIST) 1.1, modified RECIST (mRECIST), European Association for Study of the Liver (EASL), volumetric RECIST (vRECIST), quantitative EASL (qEASL) and the Liver Imaging Reporting and Data System treatment response algorithm. Kaplan-Meier survival curves were used to compare responders and non-responders for each criterion. Univariate and multivariate Cox proportional hazard ratio (HR) analysis were used to identify covariates associated with overall survival. Fleiss kappa test was used to assess interobserver agreement. RESULTS: At multivariate analysis, RECIST 1.1 (HR: 0.26; 95% confidence interval [95% CI]: 0.09-0.75; P=0.01), mRECIST (HR: 0.22; 95% CI: 0.08-0.59; P=0.003), EASL (HR: 0.22; 95% CI: 0.07-0.63; P=0.005), and qEASL (HR: 0.30; 95% CI: 0.12-0.80; P=0.02) showed a significant difference in overall survival between responders and nonresponders. RECIST 1.1 had the highest interobserver reproducibility. CONCLUSION: RECIST and mRECIST seem to be the best compromise between reproducibility and ability to predict overall survival in patients with HCC treated with TARE.
Asunto(s)
Carcinoma Hepatocelular , Quimioembolización Terapéutica , Neoplasias Hepáticas , Adulto , Anciano , Carcinoma Hepatocelular/diagnóstico por imagen , Carcinoma Hepatocelular/radioterapia , Femenino , Humanos , Neoplasias Hepáticas/diagnóstico por imagen , Neoplasias Hepáticas/radioterapia , Masculino , Persona de Mediana Edad , Reproducibilidad de los Resultados , Estudios Retrospectivos , Tomografía Computarizada por Rayos X , Resultado del Tratamiento , Radioisótopos de Itrio/uso terapéuticoRESUMEN
OBJECTIVES: To assess the value of hepatospecific MR contrast agent uptake on hepatobiliary phase (HBP) images to detect marked activation of the ß-catenin pathway in hepatocellular adenomas (HCAs). METHODS: This multicentric retrospective IRB-approved study included all patients with a pathologically proven HCA who underwent gadobenate dimeglumine-enhanced liver MRI with HBP. Tumor signal intensity on HBP was first assessed visually, and lesions were classified into three distinct groups-hypointense, isointense, or hyperintense-according to the relative signal intensity to liver. Uptake was then quantified using the lesion-to-liver contrast enhancement ratio (LLCER). Finally, the accuracy of HBP analysis in depicting marked ß-catenin activation in HCA was evaluated. RESULTS: A total of 124 HCAs were analyzed including 12 with marked ß-catenin activation (HCA B+). Visual analysis classified 94/124 (76%), 12/124 (10%), and 18/124 (14%) HCAs as being hypointense, isointense, and hyperintense on HBP, respectively. Of these, 1/94 (1%), 3/12 (25%), and 8/18 (44%) were HCA B+, respectively (p < 0.001). The LLCER of HCA B+ was higher than that of HCA without marked ß-catenin activation in the entire cohort (means 4.9 ± 11.8% vs. - 19.8 ± 11.4%, respectively, p < 0.001). A positive LLCER, i.e., LLCER ≥ 0%, had 75% (95% CI 43-95%) sensitivity and 97% (95% CI 92-99%) specificity, with a LR+ of 28 (95% CI 8.8-89.6) for the diagnosis of HCA B+. CONCLUSIONS: Hepatospecific contrast uptake on hepatobiliary phase is strongly associated with marked activation of the ß-catenin pathway in hepatocellular adenoma, and its use might improve hepatocellular adenoma subtyping on MRI. KEY POINTS: ⢠Tumor uptake on hepatobiliary phase in both the visual and quantitative analyses had a specificity higher than 90% for the detection of marked ß-catenin activation in hepatocellular adenoma. ⢠However, the sensitivity of visual analysis alone is inferior to that of LLCER quantification on HBP due to the high number of HCAs with signal hyperintensity on HBP, especially those developed on underlying liver steatosis.
Asunto(s)
Adenoma de Células Hepáticas , Carcinoma Hepatocelular , Neoplasias Hepáticas , Adenoma de Células Hepáticas/diagnóstico por imagen , Biomarcadores , Medios de Contraste , Gadolinio DTPA , Humanos , Neoplasias Hepáticas/diagnóstico por imagen , Imagen por Resonancia Magnética , Estudios Retrospectivos , Sensibilidad y Especificidad , beta CateninaRESUMEN
PURPOSE: The purpose of this study was to build and train a deep convolutional neural networks (CNN) algorithm to segment muscular body mass (MBM) to predict muscular surface from a two-dimensional axial computed tomography (CT) slice through L3 vertebra. MATERIALS AND METHODS: An ensemble of 15 deep learning models with a two-dimensional U-net architecture with a 4-level depth and 18 initial filters were trained to segment MBM. The muscular surface values were computed from the predicted masks and corrected with the algorithm's estimated bias. Resulting mask prediction and surface prediction were assessed using Dice similarity coefficient (DSC) and root mean squared error (RMSE) scores respectively using ground truth masks as standards of reference. RESULTS: A total of 1025 individual CT slices were used for training and validation and 500 additional axial CT slices were used for testing. The obtained mean DSC and RMSE on the test set were 0.97 and 3.7 cm2 respectively. CONCLUSION: Deep learning methods using convolutional neural networks algorithm enable a robust and automated extraction of CT derived MBM for sarcopenia assessment, which could be implemented in a clinical workflow.
Asunto(s)
Músculos Abdominales , Aprendizaje Profundo , Sarcopenia , Tomografía Computarizada por Rayos X , Músculos Abdominales/diagnóstico por imagen , Algoritmos , Humanos , Redes Neurales de la Computación , Sarcopenia/diagnóstico por imagenRESUMEN
PURPOSE: The second edition of the artificial intelligence (AI) data challenge was organized by the French Society of Radiology with the aim to: (i), work on relevant public health issues; (ii), build large, multicentre, high quality databases; and (iii), include three-dimensional (3D) information and prognostic questions. MATERIALS AND METHODS: Relevant clinical questions were proposed by French subspecialty colleges of radiology. Their feasibility was assessed by experts in the field of AI. A dedicated platform was set up for inclusion centers to safely upload their anonymized examinations in compliance with general data protection regulation. The quality of the database was checked by experts weekly with annotations performed by radiologists. Multidisciplinary teams competed between September 11th and October 13th 2019. RESULTS: Three questions were selected using different imaging and evaluation modalities, including: pulmonary nodule detection and classification from 3D computed tomography (CT), prediction of expanded disability status scale in multiple sclerosis using 3D magnetic resonance imaging (MRI) and segmentation of muscular surface for sarcopenia estimation from two-dimensional CT. A total of 4347 examinations were gathered of which only 6% were excluded. Three independent databases from 24 individual centers were created. A total of 143 participants were split into 20 multidisciplinary teams. CONCLUSION: Three data challenges with over 1200 general data protection regulation compliant CT or MRI examinations each were organized. Future challenges should be made with more complex situations combining histopathological or genetic information to resemble real life situations faced by radiologists in routine practice.
Asunto(s)
Inteligencia Artificial , Imagen por Resonancia Magnética , Tomografía Computarizada por Rayos X , Humanos , RadiólogosAsunto(s)
Diagnóstico por Imagen , Radiología/educación , Francia , Humanos , Invenciones , Medicina PreventivaAsunto(s)
Neoplasias de la Lengua/cirugía , Neoplasias de la Lengua/terapia , Lengua/cirugía , Adulto , Carcinoma de Células Escamosas/cirugía , Inhibidor p16 de la Quinasa Dependiente de Ciclina/metabolismo , Supervivencia sin Enfermedad , Femenino , Humanos , Ganglios Linfáticos/cirugía , Masculino , Márgenes de Escisión , Disección del Cuello/métodos , Recurrencia Local de Neoplasia , Estadificación de Neoplasias , Análisis de Regresión , Estudios Retrospectivos , Análisis de Supervivencia , Proteína p53 Supresora de Tumor/metabolismoRESUMEN
OBJECTIVES: To evaluate the lesion-to-liver visual signal intensity ratio (SIR) before and at the hepatobiliary phase MRI (HBP-MRI) after gadobenate dimeglumine (Gd-BOPTA) injection, using several T1-weighted images (T1-WI), for the characterization of benign hepatocellular lesions. METHODS: Patients with histologically proven focal nodular hyperplasia (FNH) and hepatocellular adenoma (HCA), who underwent Gd-BOPTA-enhanced HBP-MRI from 2009 to 2017, were retrospectively identified. The lesion-to-liver SIR was visually assessed by two radiologists on HBP (post-HBP analysis) and compared with that of unenhanced sequences (pre/post-HBP analysis) on T1-WI in-phase (T1-IP), out-of-phase (T1-OP), and fat suppression (T1-FS). Lesions were classified as hyper-, iso-, or hypointense on post-HBP, and as decreasing, stable, or increasing SIR on pre/post-HBP analyses. The performance of the different T1-WI sequences for the diagnostic of FNH was evaluated on post-HBP analysis. RESULTS: Twenty-nine FNHs and 33 HCAs were analyzed. On post-HBP analysis, FNHs appeared hyper-/isointense in 89.7% of all T1-WI. HCAs appeared hypointense in 93.9%, 63.6%, and 69.7% of T1-IP, T1-OP, and T1-FS, respectively. FNHs exhibited an increasing SIR in 55.2-58.6%, a stable SIR in 44.8-58.6%, and a decreasing SIR in 0%, whereas HCAs exhibited a decreasing SIR in 66.7-93.9%, a stable SIR in 6.1-33.3%, and an increasing SIR in 0% (p < 0.0001). The specificity of T1-IP was significantly higher than that of T1-OP (p = 0.015) and T1-FS (p = 0.042). CONCLUSION: T1-IP is the most reliable sequence due to misleading tumor/liver signal ratio in the case of fatty liver when using T1-FS or T1-OP. The pre/post-HBP lesion-to-liver SIR is accurate to classify benign hepatocellular lesions and contributes to avoid biopsy. KEY POINTS: â¢The T1-weighted images in-phase should be systematically included in the HBP-MRI protocol, as it is the most reliable sequence especially in the case of fatty liver. â¢The comparison between lesion-to-liver signal intensity ratios on unenhanced and at the hepatobiliary phase sequences is useful to classify benign hepatocellular lesions in three categories without misclassification: FNH (increasing signal intensity ratio), HCA (decreasing signal intensity ration), and indeterminate lesions (stable signal intensity ratio).
Asunto(s)
Adenoma de Células Hepáticas/patología , Carcinoma Hepatocelular/patología , Hiperplasia Nodular Focal/patología , Neoplasias Hepáticas/patología , Adulto , Biopsia , Medios de Contraste , Hígado Graso/patología , Femenino , Humanos , Imagen por Resonancia Magnética/métodos , Masculino , Meglumina/análogos & derivados , Persona de Mediana Edad , Compuestos Organometálicos , Estudios Retrospectivos , Sensibilidad y Especificidad , Adulto JovenRESUMEN
PURPOSE: The purpose of this study was to create an algorithm that simultaneously detects and characterizes (benign vs. malignant) focal liver lesion (FLL) using deep learning. MATERIALS AND METHODS: We trained our algorithm on a dataset proposed during a data challenge organized at the 2018 Journées Francophones de Radiologie. The dataset was composed of 367 two-dimensional ultrasound images from 367 individual livers, captured at various institutions. The algorithm was guided using an attention mechanism with annotations made by a radiologist. The algorithm was then tested on a new data set from 177 patients. RESULTS: The models reached mean ROC-AUC scores of 0.935 for FLL detection and 0.916 for FLL characterization over three shuffled three-fold cross-validations performed with the training data. On the new dataset of 177 patients, our models reached a weighted mean ROC-AUC scores of 0.891 for seven different tasks. CONCLUSION: This study that uses a supervised-attention mechanism focused on FLL detection and characterization from liver ultrasound images. This method could prove to be highly relevant for medical imaging once validated on a larger independent cohort.
Asunto(s)
Aprendizaje Profundo , Neoplasias Hepáticas/diagnóstico por imagen , Hígado/diagnóstico por imagen , Algoritmos , Conjuntos de Datos como Asunto , Humanos , UltrasonografíaRESUMEN
AIM: Predicting surgical difficulty is a critical factor in the management of locally advanced rectal cancer (LARC). This study evaluates the accuracy and external validity of a recently published morphometric score to predict surgical difficulty and additionally proposes a new score to identify preoperatively LARC patients with a high risk of having a difficult surgery. METHODS: This is a retrospective study based on the European MRI and Rectal Cancer Surgery (EuMaRCS) database, including patients with mid/low LARC who were treated with neoadjuvant chemoradiation therapy and laparoscopic total mesorectal excision (L-TME) with primary anastomosis. For all patients, pretreatment and restaging MRI were available. Surgical difficulty was graded as high and low based upon a composite outcome, including operative (e.g. duration of surgery) and postoperative variables (e.g. hospital stay). Score accuracy was assessed by estimating sensitivity, specificity and area under the receiver operating characteristic curve (AROC). RESULTS: In a total of 136 LARC patients, 17 (12.5%) were graded as high surgical difficulty. The previously published score (calculated on body mass index, intertuberous distance, mesorectal fat area, type of anastomosis) showed low predictive value (sensitivity 11.8%; specificity 92.4%; AROC 0.612). The new EuMaRCS score was developed using the following significant predictors of surgical difficulty: body mass index > 30, interspinous distance < 96.4 mm, ymrT stage ≥ T3b and male sex. It demonstrated high accuracy (AROC 0.802). CONCLUSION: The EuMaRCS score was found to be more sensitive and specific than the previous score in predicting surgical difficulty in LARC patients who are candidates for L-TME. However, this score has yet to be externally validated.