Distinguishing Intramedullary Spinal Cord Neoplasms from Non-Neoplastic Conditions by Analyzing the Classic Signs on MRI in the Era of AI

2021 ◽  
Vol 17 ◽  
Author(s):  
Ernest Junrui Lim ◽  
Natalie Wei Lyn Leong ◽  
Chi Long Ho
Keyword(s):  
Spinal Cord ◽  
Pattern Recognition ◽  
Deep Learning ◽  
Careful Analysis ◽  
Imaging Features ◽  
Intramedullary Spinal Cord ◽  
Neoplastic Processes ◽  
Wide Range ◽  
Pattern Recognition Approach ◽  
Remarkable Progress

: Intramedullary lesions can be challenging to diagnose given the wide range of possible pathologies. Each lesion has unique clinical and imaging features, which are best evaluated on magnetic resonance imaging. Radiological imaging is unique with rich, descriptive patterns and classic signs—which are often metaphorical. In this review, we present a collection of classic MRI signs, ranging from neoplastic to non-neoplastic lesions, within the spinal cord. The differential diagnosis (DD) of intramedullary lesions can be narrowed down by careful analysis of the classic signs and pattern of involvement in the spinal cord. Furthermore, the signs are illustrated memorably with emphasis on the pathophysiology, mimics and pitfalls. Artificial intelligence (AI) algorithms, particularly deep learning, have made remarkable progress in image recognition tasks. The classic signs and related illustrations can enhance a pattern recognition approach in diagnostic radiology. Deep learning can potentially be designed to distinguish neoplastic from non-neoplastic processes by pattern recognition of the classic MRI signs.

Magnetic resonance imaging of intramedullary spinal cord schwannomas

2003 ◽  
Vol 99 (1) ◽  
pp. 114-117 ◽  
Author(s):  
Cesare Colosimo ◽  
Alfonso Cerase ◽  
Luca Denaro ◽  
Giulio Maira ◽  
Romano Greco
Keyword(s):  
Spinal Cord ◽  
Magnetic Resonance ◽  
Mr Imaging ◽  
Spinal Cord Tumor ◽  
Spinal Nerve ◽  
Benign Tumors ◽  
Imaging Features ◽  
Spinal Nerve Root ◽  
Resonance Imaging ◽  
Intramedullary Spinal Cord

✓ Intramedullary spinal cord schwannomas are rare benign tumors for which resection is possible and safe. The purpose of this paper is to present the magnetic resonance (MR) imaging features in two cases of intramedullary spinal cord schwannoma to assist both neurosurgeons and pathologists in preventing misdiagnosis and resultant partial resection. The MR imaging evidence of a small- or medium-sized well-marginated intramedullary spinal cord tumor in a patient in whom no syringomyelia is present but in whom moderate edema with marked Gd enhancement can be seen should be considered in the differential diagnosis of intramedullary spinal cord schwannoma. In cases in which an associated thickened Gd-enhancing spinal nerve root is seen the diagnosis of schwannoma should be assumed.

Subependymomas: An analysis of Clinical and Imaging Features

Neurosurgery ◽  
2006 ◽  
Vol 58 (5) ◽  
pp. 881-890 ◽  
Author(s):  
Brian T. Ragel ◽  
Anne G. Osborn ◽  
Kum Whang ◽  
Jeannette J. Townsend ◽  
Randy L. Jensen ◽  
...  
Keyword(s):  
Magnetic Resonance Imaging ◽  
Spinal Cord ◽  
Magnetic Resonance ◽  
Spinal Cord Tumor ◽  
Benign Tumors ◽  
Resonance Spectroscopy ◽  
Imaging Features ◽  
Resonance Imaging ◽  
Radiographic Appearance ◽  
Intramedullary Spinal Cord

Abstract OBJECTIVE: Subependymomas are slow-growing, benign tumors usually found incidentally in the fourth ventricle at autopsy. They are typically associated with the ventricular system and become apparent clinically only when symptoms of hydrocephalus or mass effect develop. We review clinical, histological, and contemporary radiographic presentations of 16 subependymomas, including 2 intraparenchymal tumors. METHODS: We retrospectively evaluated eight patients with pathologically proven subependymomas. Initial magnetic resonance imaging and magnetic resonance spectroscopy were reviewed when available. Imaging was also available on eight outside subependymoma cases reviewed by our radiology department. RESULTS: Twelve of these subependymomas were intraventricular, one was in the posterior fossa, two were intraparenchymal, and one was an intramedullary spinal cord tumor. These lesions were hypo- to hyperintense on T1- and T2-weighted magnetic resonance imaging, with minimal to moderate enhancement. Initial complaints included headache, seizures, tingling sensations, and weakness. Among our eight patients who underwent gross total resection with no adjuvant therapy, no recurrences have been noted on follow-up magnetic resonance imaging. CONCLUSION: Subependymomas are rare, representing only 0.51% of all central nervous system tumors operated on during an 8-year period at the University of Utah. Clinical symptoms were associated with tumor location: intracranial masses caused headaches, seizures, and neurological complaints, and spinal cord locations resulted in neurological deficit. The authors review the clinical presentation, management, and contemporary radiographic appearance of this rare tumor.

scholarly journals Automatic Multiclass Intramedullary Spinal Cord Tumor Segmentation on MRI with Deep Learning

2021 ◽  
pp. 102766
Author(s):  
Andreanne Lemay ◽  
Charley Gros ◽  
Zhizheng Zhuo ◽  
Jie Zhang ◽  
Yunyun Duan ◽  
...  
Keyword(s):  
Spinal Cord ◽  
Deep Learning ◽  
Spinal Cord Tumor ◽  
Tumor Segmentation ◽  
Intramedullary Spinal Cord ◽  
Intramedullary Spinal Cord Tumor

Magnetic Resonance Imaging Features of Intramedullary Spinal Cord Tumors with Pathological Correlations

2013 ◽  
Vol 19 (1) ◽  
pp. 51-66
Author(s):  
Jureerat Thammaroj ◽  
Amnat Kitkhandee ◽  
Parinyaporn Tumkot ◽  
Pichayen Duangtongpol ◽  
Sakda Waraosawapati
Keyword(s):  
Magnetic Resonance Imaging ◽  
Spinal Cord ◽  
Magnetic Resonance ◽  
Signal Intensity ◽  
Imaging Features ◽  
Resonance Imaging ◽  
Spinal Cord Tumors ◽  
Intramedullary Spinal Cord ◽  
Cystic Component ◽  
Spinal Cord Ependymoma

Objective: The purpose of this study was to determine characteristic imaging findings of intramedullary spinal cord tumor in magnetic resonance imaging (MRI). Material and Methods: We retrospectively analyzed MRI in 15 patients with histologicaly proven intramedullary spinal cord tumors. The demographic data, MRI findings with histological findings were recorded in terms of age, location, length, morphology, signal intensity, the presence or absence of cyst and hemorrhage, enhancement pattern, other associated findings, necrosis, vascular proliferation and WHO grading. Results: Among the 15 patients, spinal cord ependymomas were eccentric 75%, well-define border 62.5% and cervicothoracic spine located 37.5%. Spinal cord astrocytomas were eccentrically located and ill-define border 85.7%, cervicothoracic and thoracic spine located 28.5%. A cystic component was seen in 87.5% of spinal cord ependymoma and 71.5% of astrocytomas. Intratumoral hemorrhage occurred in 75% of spinal cord ependymomas, and 57.1% of astrocytomas. In 12.5% of spinal cord ependymomas, a curvilinear low T2 signal, suggesting marginal hemorrhage, was seen at the upper and/or lower margins of the tumors. Twenty-five percent of spinal cord ependymoma and 57.2% of astrocytomas showed heterogeneous enhancement, while in 12.5% of spinal cord ependymomas, enhancement was homogeneous. Conclusion: Although no statistically significant characteristic MRI feature to distinguish between ependymoma and astrocytoma is detected. By percentage we found that border, length and signal intensity of tumors may help diagnosis. With pathological correlation, all of spinal cord ependymomas are mark hypervascular tumor, but astrocytomas never showed.

scholarly journals Inflammatory Myelopathy Presenting as a Cystic Intramedullary Spinal Cord Lesion

1984 ◽  
Vol 11 (3) ◽  
pp. 399-401
Author(s):  
B.I. Tranmer ◽  
T.A. Gray ◽  
W.J. Horsey ◽  
C.G. Gonsalves
Keyword(s):  
Spinal Cord ◽  
Cervical Spinal Cord ◽  
Correct Diagnosis ◽  
Careful Analysis ◽  
Intramedullary Tumor ◽  
Cervical Cord ◽  
Intramedullary Spinal Cord ◽  
Cystic Tumour ◽  
Computed Tomographic ◽  
Inflammatory Changes

ABSTRACT:A case of subacute progressive spinal tetraparesis had myelographic evidence of cervical spinal cord swelling and a delayed metrizamide computed tomographic myelogram (MCTM) suggested cavitation within the swollen spinal cord. Surgical exploration of the cervical cord revealed inflammatory changes only. No syrinx or intramedullary tumour was found. The accumulation of metrizamide within the spinal cord, as demonstrated by MCTM, did not’respresent a syrinx or a cystic tumour, but more likely an area of inflammation. Because inflammatory myelopathy may simulate an intramedullary tumor or syrinx, careful analysis of all clinical and radiological information is necessary to help make a correct diagnosis.

Levenshtein Augmentation Improves Performance of SMILES Based Deep-Learning Synthesis Prediction

2020 ◽  
Author(s):  
Dean Sumner ◽  
Jiazhen He ◽  
Amol Thakkar ◽  
Ola Engkvist ◽  
Esben Jannik Bjerrum
Keyword(s):  
Neural Networks ◽  
Pattern Recognition ◽  
Deep Learning ◽  
Recurrent Neural Networks ◽  
Data Augmentation ◽  
State Of The Art ◽  
Sequence Similarity ◽  
Learning Models ◽  
Underlying Network

<p>SMILES randomization, a form of data augmentation, has previously been shown to increase the performance of deep learning models compared to non-augmented baselines. Here, we propose a novel data augmentation method we call “Levenshtein augmentation” which considers local SMILES sub-sequence similarity between reactants and their respective products when creating training pairs. The performance of Levenshtein augmentation was tested using two state of the art models - transformer and sequence-to-sequence based recurrent neural networks with attention. Levenshtein augmentation demonstrated an increase performance over non-augmented, and conventionally SMILES randomization augmented data when used for training of baseline models. Furthermore, Levenshtein augmentation seemingly results in what we define as <i>attentional gain </i>– an enhancement in the pattern recognition capabilities of the underlying network to molecular motifs.</p>

scholarly journals A Generalization Performance Study Using Deep Learning Networks in Embedded Systems

Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1031
Author(s):  
Joseba Gorospe ◽  
Rubén Mulero ◽  
Olatz Arbelaitz ◽  
Javier Muguerza ◽  
Miguel Ángel Antón
Keyword(s):  
Deep Learning ◽  
Embedded Systems ◽  
Embedded System ◽  
General Purpose ◽  
Learning Networks ◽  
Performance Study ◽  
Learning Techniques ◽  
Wide Range ◽  
Learning Architectures

Deep learning techniques are being increasingly used in the scientific community as a consequence of the high computational capacity of current systems and the increase in the amount of data available as a result of the digitalisation of society in general and the industrial world in particular. In addition, the immersion of the field of edge computing, which focuses on integrating artificial intelligence as close as possible to the client, makes it possible to implement systems that act in real time without the need to transfer all of the data to centralised servers. The combination of these two concepts can lead to systems with the capacity to make correct decisions and act based on them immediately and in situ. Despite this, the low capacity of embedded systems greatly hinders this integration, so the possibility of being able to integrate them into a wide range of micro-controllers can be a great advantage. This paper contributes with the generation of an environment based on Mbed OS and TensorFlow Lite to be embedded in any general purpose embedded system, allowing the introduction of deep learning architectures. The experiments herein prove that the proposed system is competitive if compared to other commercial systems.

scholarly journals Data augmentation for computed tomography angiography via synthetic image generation and neural domain adaptation

2020 ◽  
Vol 6 (1) ◽  
Author(s):  
Malte Seemann ◽  
Lennart Bargsten ◽  
Alexander Schlaefer
Keyword(s):  
Computed Tomography ◽  
Neural Networks ◽  
Deep Learning ◽  
Medical Imaging ◽  
Computed Tomography Angiography ◽  
Data Augmentation ◽  
Domain Adaptation ◽  
Synthetic Image ◽  
Wide Range ◽  
The Impact

AbstractDeep learning methods produce promising results when applied to a wide range of medical imaging tasks, including segmentation of artery lumen in computed tomography angiography (CTA) data. However, to perform sufficiently, neural networks have to be trained on large amounts of high quality annotated data. In the realm of medical imaging, annotations are not only quite scarce but also often not entirely reliable. To tackle both challenges, we developed a two-step approach for generating realistic synthetic CTA data for the purpose of data augmentation. In the first step moderately realistic images are generated in a purely numerical fashion. In the second step these images are improved by applying neural domain adaptation. We evaluated the impact of synthetic data on lumen segmentation via convolutional neural networks (CNNs) by comparing resulting performances. Improvements of up to 5% in terms of Dice coefficient and 20% for Hausdorff distance represent a proof of concept that the proposed augmentation procedure can be used to enhance deep learning-based segmentation for artery lumen in CTA images.

scholarly journals Uncertainty-Aware Deep Learning-Based Cardiac Arrhythmias Classification Model of Electrocardiogram Signals

Computers ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 82
Author(s):  
Ahmad O. Aseeri
Keyword(s):  
Deep Learning ◽  
Cardiac Arrhythmias ◽  
Large Scale ◽  
Clinical Decision Making ◽  
Probabilistic Approach ◽  
Classification Model ◽  
Gating Mechanism ◽  
Uncertainty Estimates ◽  
Wide Range

Deep Learning-based methods have emerged to be one of the most effective and practical solutions in a wide range of medical problems, including the diagnosis of cardiac arrhythmias. A critical step to a precocious diagnosis in many heart dysfunctions diseases starts with the accurate detection and classification of cardiac arrhythmias, which can be achieved via electrocardiograms (ECGs). Motivated by the desire to enhance conventional clinical methods in diagnosing cardiac arrhythmias, we introduce an uncertainty-aware deep learning-based predictive model design for accurate large-scale classification of cardiac arrhythmias successfully trained and evaluated using three benchmark medical datasets. In addition, considering that the quantification of uncertainty estimates is vital for clinical decision-making, our method incorporates a probabilistic approach to capture the model’s uncertainty using a Bayesian-based approximation method without introducing additional parameters or significant changes to the network’s architecture. Although many arrhythmias classification solutions with various ECG feature engineering techniques have been reported in the literature, the introduced AI-based probabilistic-enabled method in this paper outperforms the results of existing methods in outstanding multiclass classification results that manifest F1 scores of 98.62% and 96.73% with (MIT-BIH) dataset of 20 annotations, and 99.23% and 96.94% with (INCART) dataset of eight annotations, and 97.25% and 96.73% with (BIDMC) dataset of six annotations, for the deep ensemble and probabilistic mode, respectively. We demonstrate our method’s high-performing and statistical reliability results in numerical experiments on the language modeling using the gating mechanism of Recurrent Neural Networks.

Sign in / Sign up

Export Citation Format

Share Document