Assessing robustness and generalization of a deep neural network for brain MS lesion segmentation on real-world data.

OBJECTIVES: Evaluate the performance of a deep learning (DL)-based model for multiple sclerosis (MS) lesion segmentation and compare it to other DL and non-DL algorithms. METHODS: This ambispective, multicenter study assessed the performance of a DL-based model for MS lesion segmentation and compared it to alternative DL- and non-DL-based methods. Models were tested on internal (n = 20) and external (n = 18) datasets from Latin America, and on an external dataset from Europe (n = 49). We also examined robustness by rescanning six patients (n = 6) from our MS clinical cohort. Moreover, we studied inter-human annotator agreement and discussed our findings in light of these results. Performance and robustness were assessed using intraclass correlation coefficient (ICC), Dice coefficient (DC), and coefficient of variation (CV). RESULTS: Inter-human ICC ranged from 0.89 to 0.95, while spatial agreement among annotators showed a median DC of 0.63. Using expert manual segmentations as ground truth, our DL model achieved a median DC of 0.73 on the internal, 0.66 on the external, and 0.70 on the challenge datasets. The performance of our DL model exceeded that of the alternative algorithms on all datasets. In the robustness experiment, our DL model also achieved higher DC (ranging from 0.82 to 0.90) and lower CV (ranging from 0.7 to 7.9%) when compared to the alternative methods. CONCLUSION: Our DL-based model outperformed alternative methods for brain MS lesion segmentation. The model also proved to generalize well on unseen data and has a robust performance and low processing times both on real-world and challenge-based data. CLINICAL RELEVANCE STATEMENT: Our DL-based model demonstrated superior performance in accurately segmenting brain MS lesions compared to alternative methods, indicating its potential for clinical application with improved accuracy, robustness, and efficiency. KEY POINTS: • Automated lesion load quantification in MS patients is valuable; however, more accurate methods are still necessary. • A novel deep learning model outperformed alternative MS lesion segmentation methods on multisite datasets. • Deep learning models are particularly suitable for MS lesion segmentation in clinical scenarios.

Femoral intraneural ganglion cyst: the first confirmed case report.

BACKGROUND: Ganglion cysts affecting nerve are rare causes of neuropathy. The formation of intraneural ganglion cysts, once controversial, has recently been clarified. We describe the first modern description of a femoral intraneural ganglion cyst at the hip region. METHODS: A patient presented with a 1 year history of radiating pain, quadriceps weakness and anteromedial leg numbness was found to have a femoral intraneural cyst with a hip joint connection on MRI. RESULTS: Surgical disconnection of the articular branch led to improvement of the neuropathy and resolution of the cyst on postoperative MRI. CONCLUSIONS: The unifying articular (synovial) theory describes the joint origin of intraneural cysts, even when they occur in unusual locations, and their propagation into the parent nerve. Knowledge of this theory can improve outcomes; surgery needs to address the joint origin or capsulolabral defect lest recurrence ensue.

Neuralgic amyotrophy detected by magnetic resonance neurography: subclinical, bilateral, and multifocal brachial plexus involvement.

INTRODUCTION: Neuralgic amyotrophy (NA) is a painful non-traumatic peripheral nervous system condition affecting the brachial plexus. Signal abnormalities in nerves and muscles have been detected in these patients using magnetic resonance neurography (MRN). METHODS: Electronic medical records and MRN images obtained in a 3 T scanner, in 14 adult patients diagnosed with NA at our Neurological institution (Neuromuscular Disorders Section), between December 2015 and December 2019 were retrospectively reviewed. The study was first approved by our Institutional Ethics Committee. RESULTS: Subclinical, multifocal, and bilateral nerve signal anomalies were recorded in the brachial plexus of these patients. We identified four different types of nerve constriction without entrapment, which we categorized as follows: incomplete focal (type I), complete focal or hourglass (type II), multifocal or string of pearls (type III) and segmental (type IV). CONCLUSIONS: Given that MRN is an accurate diagnostic tool to detect nerve damage, we believe abnormal findings could improve early detection of NA patients.

MRI pattern in acute optic neuritis: Comparing multiple sclerosis, NMO and MOGAD.

BACKGROUND: Several MRI findings of optic neuritis (ON) have been described and correlated with specific underlying etiologies. Specifically, optic nerve enhancement is considered an accurate biomarker of acute ON. OBJECTIVE: To identify differences in MRI patterns of optic nerve enhancement in certain demyelinating etiologies presenting with acute ON. METHODS: Retrospective analysis of enhancement patterns on fat-suppressed T1-weighted images from patients presenting clinical and radiological acute ON, treated at our institution between January 2014 and June 2022. Location and extension of enhancing optic nerve segments, as well as presence of perineural enhancement were evaluated in three predetermined demyelinating conditions. Fisher's exact test and chi2 were calculated. RESULTS: Fifty-six subjects met eligibility criteria. Mean age was 31 years (range 6-79) and 70% were females. Thirty-four (61%) patients were diagnosed with multiple sclerosis (MS), 8 (14%) with neuromyelitis optica (NMO), and 14 (25%) with anti-myelin oligodendrocyte glycoprotein disease (MOGAD). Bilateral involvement was more frequent in MOGAD, compared to MS and NMO (43 vs 3% and 12.5% respectively, p = 0.002). MS patients showed shorter optic nerve involvement, whereas MOGAD showed more extensive lesions (p = 0.006). Site of involvement was intraorbital in 63% MS, 89% NMO, 90% MOGAD (p = 0.051) and canalicular in 43% MS, 33% NMO and 75% MOGAD (p = 0.039). Intracranial or chiasmatic involvement and presence of perineural enhancement were not statistically different between entities. CONCLUSION: In the setting of acute ON, patients presenting MOGAD were more likely to show bilateral, longitudinally extended and anterior (intraorbital and canalicular) optic nerve involvement compared to patients with MS or NMO.

Childhood focal compressive mononeuropathies during the COVID-19 pandemic in Buenos Aires, Argentina.

INTRODUCTION/AIMS: Focal peripheral neuropathies are infrequently seen in pediatric patients. The COVID-19 pandemic has disrupted normal life for many people, including complete lockdowns and school closing for long periods of time in many countries, which prompted children to stay at home. Our aim is to assess whether there has been an increased incidence of focal compressive peripheral neuropathies in the pediatric population during COVID-19-associated lockdown. METHODS: Clinical, electrophysiological, and imaging characteristics were reviewed for patients referred to the electrodiagnostic (EDx) laboratory with suspicion of a focal neuropathy. The incidence of focal compressive peripheral neuropathies seen during the period of March to September 2020 was compared with the same time period in 2019. RESULTS: An increased incidence of focal neuropathies was seen in 2020 (31%) compared with 2019 (6.8%). During 2020, 7 fibular (peroneal) mononeuropathies and 2 ulnar neuropathies were diagnosed. Most patients with focal neuropathies were underweight and acknowledged prolonged screen time periods. Electrophysiological findings consisted of mostly demyelinating lesions with an overall good clinical outcome. DISCUSSION: In this study we raise awareness about a possible increased incidence of focal compressive peripheral neuropathies in children during COVID-19-associated lockdown, which may be prevented with changing positions during sedentary activities.

Brain volumes quantification from MRI in healthy controls: Assessing correlation, agreement and robustness of a convolutional neural network-based software against FreeSurfer, CAT12 and FSL.

BACKGROUND AND PURPOSE: There are instances in which an estimate of the brain volume should be obtained from MRI in clinical practice. Our objective is to calculate cross-sectional robustness of a convolutional neural network (CNN) based software (Entelai Pic) for brain volume estimation and compare it to traditional software such as FreeSurfer, CAT12 and FSL in healthy controls (HC). MATERIALS AND METHODS: Sixteen HC were scanned four times, two different days on two different MRI scanners (1.5 T and 3 T). Volumetric T1-weighted images were acquired and post-processed with FreeSurfer v6.0.0, Entelai Pic v2, CAT12 v12.5 and FSL v5.0.9. Whole-brain, grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF) volumes were calculated. Correlation and agreement between methods was assessed using intraclass correlation coefficient (ICC) and Bland Altman plots. Robustness was assessed using the coefficient of variation (CV). RESULTS: Whole-brain volume estimation had better correlation between FreeSurfer and Entelai Pic (ICC (95% CI) 0.96 (0.94-0.97)) than FreeSurfer and CAT12 (0.92 (0.88-0.96)) and FSL (0.87 (0.79-0.91)). WM, GM and CSF showed a similar trend. Compared to FreeSurfer, Entelai Pic provided similarly robust segmentations of brain volumes both on same-scanner (mean CV 1.07, range 0.20-3.13% vs. mean CV 1.05, range 0.21-3.20%, p = 0.86) and on different-scanner variables (mean CV 3.84, range 2.49-5.91% vs. mean CV 3.84, range 2.62-5.13%, p = 0.96). Mean post-processing times were 480, 5, 40 and 5 min for FreeSurfer, Entelai Pic, CAT12 and FSL respectively. CONCLUSION: Based on robustness and processing times, our CNN-based model is suitable for cross-sectional volumetry on clinical practice.

COVID-19 pneumonia accurately detected on chest radiographs with artificial intelligence.

PURPOSE: To investigate the diagnostic performance of an Artificial Intelligence (AI) system for detection of COVID-19 in chest radiographs (CXR), and compare results to those of physicians working alone, or with AI support. MATERIALS AND METHODS: An AI system was fine-tuned to discriminate confirmed COVID-19 pneumonia, from other viral and bacterial pneumonia and non-pneumonia patients and used to review 302 CXR images from adult patients retrospectively sourced from nine different databases. Fifty-four physicians blind to diagnosis, were invited to interpret images under identical conditions in a test set, and randomly assigned either to receive or not receive support from the AI system. Comparisons were then made between diagnostic performance of physicians working with and without AI support. AI system performance was evaluated using the area under the receiver operating characteristic (AUROC), and sensitivity and specificity of physician performance compared to that of the AI system. RESULTS: Discrimination by the AI system of COVID-19 pneumonia showed an AUROC curve of 0.96 in the validation and 0.83 in the external test set, respectively. The AI system outperformed physicians in the AUROC overall (70% increase in sensitivity and 1% increase in specificity, p < 0.0001). When working with AI support, physicians increased their diagnostic sensitivity from 47% to 61% (p < 0.001), although specificity decreased from 79% to 75% (p = 0.007). CONCLUSIONS: Our results suggest interpreting chest radiographs (CXR) supported by AI, increases physician diagnostic sensitivity for COVID-19 detection. This approach involving a human-machine partnership may help expedite triaging efforts and improve resource allocation in the current crisis.

HNRNPDL-related muscular dystrophy: expanding the clinical, morphological and MRI phenotypes.

Autosomal dominant limb girdle muscular dystrophy D3 HNRNPDL-related is a rare dominant myopathy caused by mutations in HNRNPDL. Only three unrelated families have been described worldwide, a Brazilian and a Chinese carrying the mutation c.1132G>A p.(Asp378Asn), and one Uruguayan with the mutation c.1132G>C p. (Asp378His), both mutations occurring in the same codon. The present study enlarges the clinical, morphological and muscle MRI spectrum of AD-HNRNPDL-related myopathies demonstrating the significant particularities of the disease. We describe two new unrelated Argentinean families, carrying the previously reported c.1132G>C p.(Asp378His) HNRNPDL mutation. There was a wide phenotypic spectrum including oligo-symptomatic cases, pure limb girdle muscle involvement or distal lower limb muscle weakness. Scapular winging was the most common finding, observed in all patients. Muscle MRIs of the thigh, at different stages of the disease, showed particular involvement of adductor magnus and vastus besides a constant preservation of the rectus femoris and the adductor longus muscles, defining a novel MRI pattern. Muscle biopsy findings were characterized by the presence of numerous rimmed vacuoles, cytoplasmic bodies, and abundant autophagic material at the histochemistry and ultrastructural levels. HNRNPDL-related LGMD D3 results in a wide range of clinical phenotypes from the classic proximal form of LGMD to a more distal phenotype. Thigh MRI suggests a specific pattern. Codon 378 of HNRNPDL gene can be considered a mutation hotspot for HNRNPDL-related myopathy. Pathologically, the disease can be classified among the autophagic rimmed vacuolar myopathies as with the other multisystem proteinopathies.

Acute Flaccid Myelitis Associated with Enterovirus D68 in Children, Argentina, 2016.

After a 2014 outbreak of severe respiratory illness caused by enterovirus D68 in the United States, sporadic cases of acute flaccid myelitis have been reported worldwide. We describe a cluster of acute flaccid myelitis cases in Argentina in 2016, adding data to the evidence of association between enterovirus D68 and this polio-like illness.

Lumbosacral plexus root thickening: Establishing normal root dimensions using magnetic resonance neurography.

Disorders affecting the lumbosacral plexus (LSP) can alter root diameter. Our aim was to determine normal LSP nerve root dimensions using magnetic resonance neurography (MRN). Eleven asymptomatic patients (ages: 18-53, mean: 34 years) underwent MRN of the LSP on a 3 T scanner with an 8-channel torso-PA coil. IDEAL T2-weighted images were acquired and nerve root dimensions were measured from the second lumbar (L2) to the first sacral (S1) vertebrae on the coronal plane, 5 mm from the dorsal root ganglion (DRG). Root size was recorded by three separate groups of radiologists with different levels of expertise. Additional LSP-MRN images were acquired from a fresh-frozen cadaver specimen using the same scanner and parameters identical to those described above. Subsequently, two experienced anatomists dissected and measured the LSP roots at exactly the same distance from the DRG, using an electronic caliper. Mean root size values recorded (± standard deviation) in the asymptomatic patients were as follows: L2: 3.12 mm (±0.92), L3: 4.29 mm (±0.95), L4: 5.13 mm (±0.79), L5: 5.29 mm (±0.9), and S1: 5.38 mm (±0.7). The correlation coefficients were 0.72 between the patient and cadaver MRN results and 0.79 between the patient and dissected cadaver MRN results. Inter-observer agreements were 0.73 among the radiologist groups and 0.87 between the anatomists conducting dissections. We believe MRN provides reliable assessments of LSP root thickness. More extensive studies should be conducted to confirm the results described here. Clin. Anat. 31:782-787, 2018. © 2018 Wiley Periodicals, Inc.

[Magnetic resonance neurography for the identification of pudendal neuralgia]. / Neurografía por resonancia magnética para la identificación de la neuralgia del nervio pudendo.

The pudendal nerve entrapment is an entity understudied by diagnosis imaging. Various causes are recognized in relation to difficult labors, rectal, perineal, urological and gynecological surgery, pelvic trauma fracture, bones tumors and compression by tumors or pelvic pseudotumors. Pudendal neuropathy should be clinically suspected, and confirmed by different methods such as electrofisiological testing: evoked potentials, terminal motor latency test and electromyogram, neuronal block and magnetic resonance imaging. The radiologist should be acquainted with the complex anatomy of the pelvic floor, particularly on the path of pudendal nerve studied by magnetic resonance imaging. High resolution magnetic resonance neurography should be used as a complementary diagnostic study along with clinical and electrophysiological examinations in patients with suspected pudendal nerve neuralgia.

Neurografía por resonancia magnética para la identificación de la neuralgia del nervio pudendo / Magnetic resonance neurography for the identification of pudendal neuralgia

La neuralgia del nervio pudendo (NP) es una entidad poco estudiada por imágenes. Se reconocen varias causas, tales como compresión a través de su paso por estructuras ligamentarias; estiramiento por partos laboriosos; lesiones secundarias a cirugías rectales, perineales, urológicas y ginecológicas, traumatismos con o sin fractura de huesos pelvianos; procesos inflamatorios/autoinmunes; tumores del NP, y, compresión/desplazamiento por tumores o seudotumores de pelvis. El diagnóstico de neuralgia del NP se sospecha por la clínica y se confirma por diferentes métodos, tales como las pruebas electrofisiolológicas: potenciales evocados, test de latencia motora terminal y electromiograma, y, a través de bloqueos neurales y resonancia magnética. La neurografía por resonancia magnética de alta resolución, debería ser empleada como estudio diagnóstico complementario junto a la clínica y exámenes electrofisiológicos, en los pacientes con sospecha de neuralgia del NP.

The pudendal nerve entrapment is an entity understudied by diagnosis imaging. Various causes are recognized in relation to difficult labors, rectal, perineal, urological and gynecological surgery, pelvic trauma fracture, bones tumors and compression by tumors or pelvic pseudotumors. Pudendal neuropathy should be clinically suspected, and confirmed by different methods such as electrofisiological testing: evoked potentials, terminal motor latency test and electromyogram, neuronal block and magnetic resonance imaging. The radiologist should be acquainted with the complex anatomy of the pelvic floor, particularly on the path of pudendal nerve studied by magnetic resonance imaging. High resolution magnetic resonance neurography should be used as a complementary diagnostic study along with clinical and electrophysiological examinations in patients with suspected pudendal nerve neuralgia.

Analysis of the Revision Process by American Journal of Roentgenology Reviewers and Section Editors: Metrics of Rejected Manuscripts and Their Final Disposition.

OBJECTIVE: The objective of this study was to evaluate metrics related to manuscripts rejected by AJR with and without review during 2014 and to determine their final disposition: no record of eventual publication, eventually published, published with modified authors and title, published with the same title but modified authors, and published with modified title but the same authors. MATERIALS AND METHODS: A total of 1245 unsolicited manuscripts submitted from January to December 2014 were included in this retrospective analysis. Data were extracted from the AJR's manuscript submission system. Standard statistical analysis was used to assess the fate of a sample of 200 rejected manuscripts. RESULTS: Of the 200 manuscripts studied, 117 (59%) were published in other scientific journals (61 with revision, 56 without revision; Χ2 = 0.329, p = 0.566). Thirty-two of the 61 manuscripts (52%) rejected after peer review were later published in other journals without changes in their titles or authors, 16 (26%) with changes only in authors, 10 (16%) with changes only in their titles, and three (5%) with changes in authors and titles. Twenty-six of the 56 manuscripts (46%) rejected without peer review were published without changes in their titles or authors, 17 (30%) with changes in authors, 11 (20%) with changes only in their titles, and two (4%) with changes in both authors and titles (p = 0.686). Ten articles were published in open access journals. Of the remaining articles, those that had been reviewed were published in journals with a mean impact factor ± SD of 2.37 ± 1.30, and those that had not been reviewed were published in journals with a mean impact factor of 2.04 ± 1.06. Analysis of the 25th and 75th percentiles revealed that values were also higher for the group rejected with review (Wilcoxon rank sum test: W = 1679, p = 0.1127). Out of 61 articles rejected for publication with review, 52.5% were published with changes to their abstracts, whereas the remaining 47.5% were unchanged. CONCLUSION: This analysis found that manuscripts submitted to AJR that were rejected after review were published in journals with higher impact factors than those rejected without review. The commentaries provided by AJR reviewers and section editors appear to improve the quality of rejected manuscripts and thus contribute to the scientific community.

Muscle MRI in pediatrics: clinical, pathological and genetic correlation.

Pediatric myopathies comprise a very heterogeneous group of disorders that may develop at different ages and affect different muscle groups. Its diagnosis is sometimes difficult and must be confirmed by muscle biopsy and/or genetic analysis. In recent years, muscle involvement patterns observed on MRI have become a valuable tool, aiding clinical diagnosis and enriching pathological and genetic assessments. We selected eight myopathy cases from our institutional database in which the pattern of muscle involvement observed on MRI was almost pathognomonic and could therefore contribute to establishing diagnosis. Muscle biopsy, genetic diagnosis or both confirmed all cases.

Conflictos éticos en el editorialismo científico: el autoplagio / Ethical conflicts in scientific publishing: self-plagiarism

No disponible