Publications by authors named "Mohit Agarwal"

66 Publications

Oligodendrocyte tethering effect on hyperelastic 3D response of axons in white matter.

J Mech Behav Biomed Mater 2022 Aug 2;134:105394. Epub 2022 Aug 2.

Mechanical and Aerospace Engineering, Rutgers University-New Brunswick, Piscataway, NJ, 08854, USA. Electronic address:

A novel finite element model is proposed to study the mechanical response of axons embedded in extracellular matrix when subjected to tensile loads under purely non-affine kinematic boundary conditions. Ogden hyperelastic material model describes the axons and the extracellular matrix material characterizations. Two axon-glia tethering scenarios in white matter are studied a single oligodendrocyte (single-OL) with multiple connections a multi-oligodendrocyte (multi-OL) one. In the multi-OL tethering configuration, resultant forces are randomly oriented as distributed glial cells arbitrarily wrap around axons in their immediate vicinity. In the single-OL setup, a centrally located oligodendrocyte myelinates multiple axons nearby. Tethering forces are directed towards this oligodendrocyte, resulting in greater directionality and farther-reaching stress distribution. The oligodendrocyte connections to axons are represented by a spring-dashpot model. The material properties of myelin served as the upper limit for the parameterization of the oligodendrocyte stiffness ("K"). The proposed FE models enable realization of connection mechanisms and their influence on axonal stiffness to determine resultant stress states accurately. Root mean square deviation analysis of stress-strain plots of different connection scenarios reveal an increasing axonal stiffness with increasing tethering, indicating the role of oligodendrocytes in stress redistribution. In single-OL submodel, for the same number of connections per axons, RMSD values increased as "K" (the oligodendrocyte spring stiffness) values were set higher. RMSD calculations reveal that for a "K" value, single-OL model yielded slightly stiffer models compared to multi-OL. The current study also addresses the potential geometrical limitations of multi-OL model by randomizing and adding connections to ensure greater responsiveness. Cyclic bending stresses noticed in both submodels suggest the risk of axonal damage accumulation and repeated load failure.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jmbbm.2022.105394DOI Listing
August 2022

Bell's palsy as a presenting feature of COVID-19.

Oman J Ophthalmol 2022 May-Aug;15(2):258-259. Epub 2022 Jun 29.

Department of Anesthesiology, All India Institute of Medical Sciences, Gorakhpur, Uttar Pradesh, India.

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4103/ojo.ojo_38_21DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC9351953PMC
June 2022

Eight pruning deep learning models for low storage and high-speed COVID-19 computed tomography lung segmentation and heatmap-based lesion localization: A multicenter study using COVLIAS 2.0.

Comput Biol Med 2022 07 21;146:105571. Epub 2022 May 21.

Ann and Robert H. Lurie Children's Hospital of Chicago, Chicago, USA.

Background: COVLIAS 1.0: an automated lung segmentation was designed for COVID-19 diagnosis. It has issues related to storage space and speed. This study shows that COVLIAS 2.0 uses pruned AI (PAI) networks for improving both storage and speed, wiliest high performance on lung segmentation and lesion localization.

Method: ology: The proposed study uses multicenter ∼9,000 CT slices from two different nations, namely, CroMed from Croatia (80 patients, experimental data), and NovMed from Italy (72 patients, validation data). We hypothesize that by using pruning and evolutionary optimization algorithms, the size of the AI models can be reduced significantly, ensuring optimal performance. Eight different pruning techniques (i) differential evolution (DE), (ii) genetic algorithm (GA), (iii) particle swarm optimization algorithm (PSO), and (iv) whale optimization algorithm (WO) in two deep learning frameworks (i) Fully connected network (FCN) and (ii) SegNet were designed. COVLIAS 2.0 was validated using "Unseen NovMed" and benchmarked against MedSeg. Statistical tests for stability and reliability were also conducted.

Results: Pruning algorithms (i) FCN-DE, (ii) FCN-GA, (iii) FCN-PSO, and (iv) FCN-WO showed improvement in storage by 92.4%, 95.3%, 98.7%, and 99.8% respectively when compared against solo FCN, and (v) SegNet-DE, (vi) SegNet-GA, (vii) SegNet-PSO, and (viii) SegNet-WO showed improvement by 97.1%, 97.9%, 98.8%, and 99.2% respectively when compared against solo SegNet. AUC > 0.94 (p < 0.0001) on CroMed and > 0.86 (p < 0.0001) on NovMed data set for all eight EA model. PAI <0.25 s per image. DenseNet-121-based Grad-CAM heatmaps showed validation on glass ground opacity lesions.

Conclusions: Eight PAI networks that were successfully validated are five times faster, storage efficient, and could be used in clinical settings.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.compbiomed.2022.105571DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC9123805PMC
July 2022

Anatomy and Diseases of the Greater Wings of the Sphenoid Bone.

Radiographics 2022 Jul-Aug;42(4):1177-1195. Epub 2022 Jun 3.

From the Department of Radiology, Instituto Nacional de Câncer (INCA), Praça Cruz Vermelha 23, Rio de Janeiro, RJ, Brazil 20230-130 (R.C., J.A., C.A.d.A., R.P.S., M.F.U., A.P.K.C., G.S.M., G.R.S.d.N., F.M., M.D.); and Department of Radiology, Medical College of Wisconsin, Milwaukee, Wis (M.A.).

The greater wings of the sphenoid bone (GWS) comprise the components of the sphenoid bone that make up most of the posterior orbital wall and form the anterior and medial parts of the floor of the middle cranial fossa. Many important skull base foramina, which transmit vital neurovascular structures, are present in these paired wings on either side of the central body of the sphenoid bone. A wide variety of diseases can affect the GWS, ranging from benign osseus lesions to malignant primary and secondary bone abnormalities. The complex three-dimensional curved (winged) shape of the GWS and the wide array of pathologic entities that affect this bone can make it challenging for the radiologist to report the imaging findings accurately, especially in relation to the important skull base foramina. The authors describe a systematic approach to understanding the three-dimensional anatomy of the GWS and review important diseases, with the aid of imaging examples. Useful imaging "pearls" that can help in making specific diagnoses are provided throughout the article. RSNA, 2022.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1148/rg.210094DOI Listing
July 2022

ACR Appropriateness Criteria® Imaging of Facial Trauma Following Primary Survey.

J Am Coll Radiol 2022 05;19(5S):S67-S86

Specialty Chair, Atlanta VA Health Care System and Emory University, Atlanta, Georgia.

Maxillofacial trauma patients comprise a significant subset of patients presenting to emergency departments. Before evaluating for facial trauma, an emergency or trauma physician must perform a primary survey to ensure patient stabilization. Following this primary survey, this document discusses the following clinical scenarios for facial trauma: tenderness to palpation or contusion or edema over frontal bone (suspected frontal bone injury); pain with upper jaw manipulation or pain overlying zygoma or zygomatic deformity or facial elongation or malocclusion or infraorbital nerve paresthesia (suspected midface injury); visible nasal deformity or palpable nasal deformity or tenderness to palpation of the nose or epistaxis (suspected nasal bone injury); and trismus or malocclusion or gingival hemorrhage or mucosal hemorrhage or loose teeth or fractured teeth or displaced teeth (suspected mandibular injury). The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jacr.2022.02.013DOI Listing
May 2022

ACR Appropriateness Criteria® Sinonasal Disease: 2021 Update.

J Am Coll Radiol 2022 05;19(5S):S175-S193

Specialty Chair, Atlanta VA Health Care System and Emory University, Atlanta, Georgia.

This article presents guidelines for initial imaging utilization in patients presenting with sinonasal disease, including acute rhinosinusitis without and with suspected orbital and intracranial complications, chronic rhinosinusitis, suspected invasive fungal sinusitis, suspected sinonasal mass, and suspected cerebrospinal fluid leak. CT and MRI are the primary imaging modalities used to evaluate patients with sinonasal disease. Given its detailed depiction of bony anatomy, CT can accurately demonstrate the presence of sinonasal disease, bony erosions, and anatomic variants, and is essential for surgical planning. Given its superior soft tissue contrast, MRI can accurately identify clinically suspected intracranial and intraorbital complications, delineate soft tissue extension of tumor and distinguish mass from obstructed secretions.The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision include an extensive analysis of current medical literature from peer reviewed journals and the application of well-established methodologies (RAND/UCLA Appropriateness Method and Grading of Recommendations Assessment, Development, and Evaluation or GRADE) to rate the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where evidence is lacking or equivocal, expert opinion may supplement the available evidence to recommend imaging or treatment.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.jacr.2022.02.011DOI Listing
May 2022

Diagnosis of Lumbar Spondylolisthesis Using Optimized Pretrained CNN Models.

Comput Intell Neurosci 2022 13;2022:7459260. Epub 2022 Apr 13.

Bennett University, Greater Noida, India.

Spondylolisthesis refers to the slippage of one vertebral body over the adjacent one. It is a chronic condition that requires early detection to prevent unpleasant surgery. The paper presents an optimized deep learning model for detecting spondylolisthesis in X-ray radiographs. The dataset contains a total of 299 X-ray radiographs from which 156 images are showing the spine with spondylolisthesis and 143 images are of the normal spine. Image augmentation technique is used to increase the data samples. In this study, VGG16 and InceptionV3 models were used for the image classification task. The developed model is optimized by utilizing the TFLite model optimization technique. The experimental result shows that the VGG16 model has achieved a 98% accuracy rate, which is higher than InceptionV3's 96% accuracy rate. The size of the implemented model is reduced up to four times so it can be used on small devices. The compressed VGG16 and InceptionV3 models have achieved 100% and 96% accuracy rate, respectively. Our finding shows that the implemented models were outperformed in the diagnosis of lumbar spondylolisthesis as compared to the model suggested by Varcin et al. (which had a maximum of 93% accuracy rate). Also, the developed quantized model has achieved higher accuracy rate than Zebin and Rezvy's (VGG16 + TFLite) model with 90% accuracy. Furthermore, by evaluating the model's performance on other publicly available datasets, we have generalised our approach on the public platform.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1155/2022/7459260DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC9007141PMC
April 2022

Long-term oxygen therapy prescription in India: Evaluation of compliance, factors affecting compliance, indications, and survival.

Lung India 2022 Mar-Apr;39(2):129-138

Department of Pulmonary Medicine, ESI-PGIMSR, New Delhi, India.

Introduction: The international data shows that long-term oxygen therapy (LTOT) compliance is insufficient and variable. We conducted the first study from India on LTOT compliance, factors affecting compliance, indications, and survival through oxygen concentrator.

Materials And Methods: Our organization from Delhi had given 378 oxygen concentrators over the last 5 years. We evaluated 120 patients randomly for participating in the study. Compliance was defined as the use of LTOT for at least 15 h/day.

Results: Ninety-seven patients were included in the final analysis after exclusion criteria. The compliance to LTOT was seen in 45.36% (44/97). The commonest cause of noncompliance was lack of instructions (49.06%) followed by electricity issues, social stigma, and workplace constraints. A higher PaCO was associated with significantly lower compliance (PaCO 53.18 vs. 44.98 mmHg, P = 0.036). Interstitial lung disease was associated with significantly higher compliance. Oxygen prescription was titrated with arterial blood gas analysis in only 4.12%. The indications for LTOT were chronic obstructive pulmonary disease (49.48%), posttuberculous obstructive airway disease (20.6%), and interstitial lung disease (12.37%). We found a significant reduction in the mean number of exacerbations/year from 3.91 to 1.93 (P < 0.0001). 61.86% of the patients were surviving on LTOT with a median survival time of 12 months.

Conclusion: The adherence to LTOT in Indian patients is suboptimal mainly due to lack of instruction and is associated with a higher PaCO. The practice of titration needs to be followed. The development of a national registry to monitor LTOT should be the long-term target.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4103/lungindia.lungindia_445_21DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC9053926PMC
March 2022

Imaging features of cartilaginous tumors of the head and neck.

J Clin Imaging Sci 2021 4;11:66. Epub 2021 Dec 4.

Department of Radiology, Mayo Clinic Arizona, Phoenix, Arizona, United States.

There is a wide spectrum of head and neck cartilaginous lesions which include both neoplastic and nonneoplastic processes. Cartilaginous tumors of the head and neck are uncommon, posing a diagnostic challenge. Benign cartilaginous tumors that may occur in the head and neck include chondroma, chondroblastoma, chondromyxoid fibroma, osteochondroma, and synovial chondromatosis. Chondromesenchymal hamartoma is a rare non-neoplastic cartilaginous lesion that is included for the 1first time in the new WHO classification and radiologically can mimic a tumor. Malignant cartilaginous tumors include chondrosarcoma and chondroid variant of chordoma. Characteristic tumor locations, internal chondroid matrix calcification, and typical T2 hyperintense signal secondary to high-water content within the extracellular matrix of the hyaline cartilage are useful imaging features that narrow the differential diagnosis and help in diagnosing these diseases. This article presents a narrative review of the anatomy of the head and neck cartilaginous structures, discusses the current knowledge and imaging spectrum of benign and malignant cartilaginous tumors and tumor-like lesions of the head and neck.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.25259/JCIS_186_2021DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8720426PMC
December 2021

A study to analyze the pattern of synovial lesions from synovial biopsies in a tertiary care centre.

Indian J Pathol Microbiol 2021 Oct-Dec;64(4):702-706

Department of Pathology, Command Hospital Southern Command, Pune, Maharashtra, India.

Introduction: Synovium has been documented as a primary site of inflammation and a major effector organ in a variety of joint diseases. Study of simple technique like synovial biopsy can help in early diagnosis and treatment of diseases significantly improving outcome of patient in cases of rheumatoid arthritis, osteoarthritis, etc., Only limited data exist on utility of synovial biopsies.

Aim And Objectives: To analyze the pattern of synovial lesions to differentiate between different kinds of arthritis. Also, to identify early stages of arthritis so as to prevent unnecessary invasive surgical procedure.

Materials And Methods: It's a retrospective study to analyze 103 cases of synovial lesions diagnosed in last five years at a tertiary care orthopedic center. All synovial biopsies obtained mainly by open method and few by arthroscopic method, that came to the Dept of Pathology were included. Lesions were classified into four categories that is, inflammatory joint diseases, degenerative joint diseases, tumor-like conditions and tumors.

Results: Age group most affected was between 61 and 70 years, with male predominance. Osteoarthritis (OA) was the most common histopathological diagnosis. Early OA tissues showed greater lining layer thickness, vessel proliferation, and inflammation, while surface fibrin deposition along with fibrosis was noted in later stages.

Conclusion: The histo-morphological observations made in this study may have important therapeutic implications for some patients during the early evolution of arthritis and could prevent unnecessary operative intervention of later stages.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4103/IJPM.IJPM_498_20DOI Listing
February 2022

Hippocampal Volumes in Amnestic and Non-Amnestic Mild Cognitive Impairment Types Using Two Common Methods of MCI Classification.

J Int Neuropsychol Soc 2022 04 16;28(4):391-400. Epub 2021 Jun 16.

Department of Neurology, Medical College of Wisconsin, Milwaukee, WI.

Objectives: Mild cognitive impairment (MCI) types may have distinct neuropathological substrates with hippocampal atrophy particularly common in amnestic MCI (aMCI). However, depending on the MCI classification criteria applied to the sample (e.g., number of abnormal test scores considered or thresholds for impairment), volumetric findings between MCI types may change. Additionally, despite increased clinical use, no prior research has examined volumetric differences in MCI types using the automated volumetric software, Neuroreader™.

Methods: The present study separately applied the Petersen/Winblad and Jak/Bondi MCI criteria to a clinical sample of older adults (N = 82) who underwent neuropsychological testing and brain MRI. Volumetric data were analyzed using Neuroreader™ and hippocampal volumes were compared between aMCI and non-amnestic MCI (naMCI).

Results: T-tests revealed that regardless of MCI classification criteria, hippocampal volume z-scores were significantly lower in aMCI compared to naMCI (p's < .05), and hippocampal volume z-scores significantly differed from 0 (Neuroreader™ normative mean) in the aMCI group only (p's < .05). Additionally, significant, positive correlations were found between measures of delayed recall and hippocampal z-scores in aMCI using either MCI classification criteria (p's < .05).

Conclusions: We provide evidence of correlated neuroanatomical changes associated with memory performance for two commonly used neuropsychological MCI classification criteria. Future research should investigate the clinical utility of hippocampal volumes analyzed via Neuroreader™ in MCI.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1017/S1355617721000564DOI Listing
April 2022

Solvent Induced Helix Folding of Defined Indolenine Squaraine Oligomers.

Chemistry 2021 Jun 7;27(32):8380-8389. Epub 2021 May 7.

Institut für Organische Chemie, Universität Würzburg, Am Hubland, 97074, Würzburg, Germany.

A protecting group strategy was employed to synthesise a series of indolenine squaraine dye oligomers up to the nonamer. The longer oligomers show a distinct solvent dependence of the absorption spectra, that is, either a strong blue shift or a strong red shift of the lowest energy bands in the near infrared spectral region. This behaviour is explained by exciton coupling theory as being due to H- or J-type coupling of transition moments. The H-type coupling is a consequence of a helix folding in solvents with a small Hansen dispersity index. DOSY NMR, small angle neutron scattering (SANS), quantum chemical and force field calculations agree upon a helix structure with an unusually large pitch and open voids that are filled with solvent molecules, thereby forming a kind of clathrate. The thermodynamic parameters of the folding process were determined by temperature dependent optical absorption spectra.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/chem.202101063DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8251825PMC
June 2021

The development and efficacy of a mobile phone application to improve medication adherence for persons with epilepsy in limited resource settings: A preliminary study.

Epilepsy Behav 2021 03 10;116:107794. Epub 2021 Feb 10.

Department of Neurology, All India Institute of Medical Sciences, New Delhi, India. Electronic address:

Objective: Persons with epilepsy (PWE), especially those with limited education backgrounds from developing countries, are challenged by complicated medication regimens, debilitating seizures, and stigmatization in their daily life. Consequently, it is difficult for physicians to ensure medication adherence. This study validates a novel mobile application which was hypothesized to increase medication adherence and self-management skills in PWE. Created by medical professionals, the application included behavioral and educational components and was built to be easy-to-understand for those of socio-economically disadvantaged backgrounds.

Methods: This was a parallel, two-armed randomized controlled trial in which a total of 96 participants were enrolled from a Neurology Outpatient Department into a control standard care group and a mobile application group that used the smartphone application (app) in addition to the standard medical treatment. The app was intuitive and easy to understand for those coming from a socio-economically disadvantaged background. Medication adherence and self-efficacy were assessed with the Morisky Green and Levine Scale (MGLS) and the Epilepsy Self Efficacy Scale (ESES). Patients were reassessed 12 weeks later. Change in seizure frequency following administration of the application was a secondary outcome.

Results: In an intent-to-treat analysis, the mobile application interventional group showed over a 60% increase in the proportion of medication adherence (P < 0.0001). The mean self-efficacy score for the mobile application group was increased from 269.5 to 289.75 (P < 0.0001). The control group showed no statistically significant increases in either the proportion adherent or mean self-efficacy scores.

Significance: This study demonstrated the statistically significant performance of a mobile application in improving medication adherence and self-management skills in Indian persons with epilepsy.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.yebeh.2021.107794DOI Listing
March 2021

A narrative review on characterization of acute respiratory distress syndrome in COVID-19-infected lungs using artificial intelligence.

Comput Biol Med 2021 03 18;130:104210. Epub 2021 Jan 18.

Academic Affairs, Dudley Group NHS Foundation Trust, Dudley, UK.

COVID-19 has infected 77.4 million people worldwide and has caused 1.7 million fatalities as of December 21, 2020. The primary cause of death due to COVID-19 is Acute Respiratory Distress Syndrome (ARDS). According to the World Health Organization (WHO), people who are at least 60 years old or have comorbidities that have primarily been targeted are at the highest risk from SARS-CoV-2. Medical imaging provides a non-invasive, touch-free, and relatively safer alternative tool for diagnosis during the current ongoing pandemic. Artificial intelligence (AI) scientists are developing several intelligent computer-aided diagnosis (CAD) tools in multiple imaging modalities, i.e., lung computed tomography (CT), chest X-rays, and lung ultrasounds. These AI tools assist the pulmonary and critical care clinicians through (a) faster detection of the presence of a virus, (b) classifying pneumonia types, and (c) measuring the severity of viral damage in COVID-19-infected patients. Thus, it is of the utmost importance to fully understand the requirements of for a fast and successful, and timely lung scans analysis. This narrative review first presents the pathological layout of the lungs in the COVID-19 scenario, followed by understanding and then explains the comorbid statistical distributions in the ARDS framework. The novelty of this review is the approach to classifying the AI models as per the by school of thought (SoTs), exhibiting based on segregation of techniques and their characteristics. The study also discusses the identification of AI models and its extension from non-ARDS lungs (pre-COVID-19) to ARDS lungs (post-COVID-19). Furthermore, it also presents AI workflow considerations of for medical imaging modalities in the COVID-19 framework. Finally, clinical AI design considerations will be discussed. We conclude that the design of the current existing AI models can be improved by considering comorbidity as an independent factor. Furthermore, ARDS post-processing clinical systems must involve include (i) the clinical validation and verification of AI-models, (ii) reliability and stability criteria, and (iii) easily adaptable, and (iv) generalization assessments of AI systems for their use in pulmonary, critical care, and radiological settings.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.compbiomed.2021.104210DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7813499PMC
March 2021

Wilson disease tissue classification and characterization using seven artificial intelligence models embedded with 3D optimization paradigm on a weak training brain magnetic resonance imaging datasets: a supercomputer application.

Med Biol Eng Comput 2021 Mar 5;59(3):511-533. Epub 2021 Feb 5.

Stroke Monitoring and Diagnostic Division, AtheroPoint™, Roseville, CA, 95661, USA.

Wilson's disease (WD) is caused by copper accumulation in the brain and liver, and if not treated early, can lead to severe disability and death. WD has shown white matter hyperintensity (WMH) in the brain magnetic resonance scans (MRI) scans, but the diagnosis is challenging due to (i) subtle intensity changes and (ii) weak training MRI when using artificial intelligence (AI). Design and validate seven types of high-performing AI-based computer-aided design (CADx) systems consisting of 3D optimized classification, and characterization of WD against controls. We propose a "conventional deep convolution neural network" (cDCNN) and an "improved DCNN" (iDCNN) where rectified linear unit (ReLU) activation function was modified ensuring "differentiable at zero." Three-dimensional optimization was achieved by recording accuracy while changing the CNN layers and augmentation by several folds. WD was characterized using (i) CNN-based feature map strength and (ii) Bispectrum strengths of pixels having higher probabilities of WD. We further computed the (a) area under the curve (AUC), (b) diagnostic odds ratio (DOR), (c) reliability, and (d) stability and (e) benchmarking. Optimal results were achieved using 9 layers of CNN, with 4-fold augmentation. iDCNN yields superior performance compared to cDCNN with accuracy and AUC of 98.28 ± 1.55, 0.99 (p < 0.0001), and 97.19 ± 2.53%, 0.984 (p < 0.0001), respectively. DOR of iDCNN outperformed cDCNN fourfold. iDCNN also outperformed (a) transfer learning-based "Inception V3" paradigm by 11.92% and (b) four types of "conventional machine learning-based systems": k-NN, decision tree, support vector machine, and random forest by 55.13%, 28.36%, 15.35%, and 14.11%, respectively. The AI-based systems can potentially be useful in the early WD diagnosis. Graphical Abstract.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/s11517-021-02322-0DOI Listing
March 2021

Six artificial intelligence paradigms for tissue characterisation and classification of non-COVID-19 pneumonia against COVID-19 pneumonia in computed tomography lungs.

Int J Comput Assist Radiol Surg 2021 Mar 3;16(3):423-434. Epub 2021 Feb 3.

Stroke Monitoring and Diagnostic Division, AtheroPoint™, Roseville, CA, USA.

Background: COVID-19 pandemic has currently no vaccines. Thus, the only feasible solution for prevention relies on the detection of COVID-19-positive cases through quick and accurate testing. Since artificial intelligence (AI) offers the powerful mechanism to automatically extract the tissue features and characterise the disease, we therefore hypothesise that AI-based strategies can provide quick detection and classification, especially for radiological computed tomography (CT) lung scans.

Methodology: Six models, two traditional machine learning (ML)-based (k-NN and RF), two transfer learning (TL)-based (VGG19 and InceptionV3), and the last two were our custom-designed deep learning (DL) models (CNN and iCNN), were developed for classification between COVID pneumonia (CoP) and non-COVID pneumonia (NCoP). K10 cross-validation (90% training: 10% testing) protocol on an Italian cohort of 100 CoP and 30 NCoP patients was used for performance evaluation and bispectrum analysis for CT lung characterisation.

Results: Using K10 protocol, our results showed the accuracy in the order of DL > TL > ML, ranging the six accuracies for k-NN, RF, VGG19, IV3, CNN, iCNN as 74.58 ± 2.44%, 96.84 ± 2.6, 94.84 ± 2.85%, 99.53 ± 0.75%, 99.53 ± 1.05%, and 99.69 ± 0.66%, respectively. The corresponding AUCs were 0.74, 0.94, 0.96, 0.99, 0.99, and 0.99 (p-values < 0.0001), respectively. Our Bispectrum-based characterisation system suggested CoP can be separated against NCoP using AI models. COVID risk severity stratification also showed a high correlation of 0.7270 (p < 0.0001) with clinical scores such as ground-glass opacities (GGO), further validating our AI models.

Conclusions: We prove our hypothesis by demonstrating that all the six AI models successfully classified CoP against NCoP due to the strong presence of contrasting features such as ground-glass opacities (GGO), consolidations, and pleural effusion in CoP patients. Further, our online system takes < 2 s for inference.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/s11548-021-02317-0DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7854027PMC
March 2021

A Novel Block Imaging Technique Using Nine Artificial Intelligence Models for COVID-19 Disease Classification, Characterization and Severity Measurement in Lung Computed Tomography Scans on an Italian Cohort.

J Med Syst 2021 Jan 26;45(3):28. Epub 2021 Jan 26.

Stroke Diagnosis and Monitoring Division, AtheroPoint™, Roseville, CA, 95661, USA.

Computer Tomography (CT) is currently being adapted for visualization of COVID-19 lung damage. Manual classification and characterization of COVID-19 may be biased depending on the expert's opinion. Artificial Intelligence has recently penetrated COVID-19, especially deep learning paradigms. There are nine kinds of classification systems in this study, namely one deep learning-based CNN, five kinds of transfer learning (TL) systems namely VGG16, DenseNet121, DenseNet169, DenseNet201 and MobileNet, three kinds of machine-learning (ML) systems, namely artificial neural network (ANN), decision tree (DT), and random forest (RF) that have been designed for classification of COVID-19 segmented CT lung against Controls. Three kinds of characterization systems were developed namely (a) Block imaging for COVID-19 severity index (CSI); (b) Bispectrum analysis; and (c) Block Entropy. A cohort of Italian patients with 30 controls (990 slices) and 30 COVID-19 patients (705 slices) was used to test the performance of three types of classifiers. Using K10 protocol (90% training and 10% testing), the best accuracy and AUC was for DCNN and RF pairs were 99.41 ± 5.12%, 0.991 (p < 0.0001), and 99.41 ± 0.62%, 0.988 (p < 0.0001), respectively, followed by other ML and TL classifiers. We show that diagnostics odds ratio (DOR) was higher for DL compared to ML, and both, Bispecturm and Block Entropy shows higher values for COVID-19 patients. CSI shows an association with Ground Glass Opacities (0.9146, p < 0.0001). Our hypothesis holds true that deep learning shows superior performance compared to machine learning models. Block imaging is a powerful novel approach for pinpointing COVID-19 severity and is clinically validated.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1007/s10916-021-01707-wDOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7835451PMC
January 2021

A mysterious case of an elevated dome of the right diaphragm.

Breathe (Sheff) 2020 Jun;16(2):190334

Dept of Pulmonary Medicine, ESI PGIMSR, New Delhi, India.

https://bit.ly/2UXTuw7.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1183/20734735.0334-2019DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7714547PMC
June 2020

Fronto-Orbital Variant of Supraorbital Keyhole Approach for Clipping Ruptured Anterior Circulation Aneurysms (f-Sokha).

Neurol India 2020 Sep-Oct;68(5):1019-1027

Department of Neurosurgery, All India Institute of Medical Sciences, New Delhi, India.

Objective: The following paper describes the technique and outcomes of fronto-orbital variant of supraorbital key hole approach (f-SOKHA) to clip anterior circulation aneurysms and compares the same with a similar cohort operated through pterional craniotomy (PT).

Material And Methods: Ambispective study (2012-2019); Technique applied for anterior circulation aneurysms. Contraindications included: Large hematomas, tense brain, avoided in poor grade (Hunt and Hess grade III and IV). Large frontal sinus: Relative contraindication. Procedure included a trans-ciliary skin incision, burr hole over key point, cutting of orbital roof via the burr hole, and removal of a single small fronto-orbital flap (1). This was followed by drilling of the inner table of the frontal bone (2). Both 1 and 2 resulted in expansion of the operative space by 60%. Results compared with a similar cohort of PT.

Results: n = 75 cases; most commonly used for ACom (anterior communicating: 43) followed by middle cerebral (16), internal cerebral (13), Posterior communicating (6), anterior cerebral (2), and anterior choroidal (1). Mean age: 47.9 ± 14 years; mean Hunt and Hess grade: 1.96 ± 1.35; duration of surgery: 203 ± 45 minutes, mean size of aneurysm: 6.96 ± 3.65 mm. Both blood loss and surgery duration was less (P: 0.099 and <0.001) when compared with a similar cohort with PT. It also demonstrated better cosmetic results and patient satisfaction.

Conclusions: f-SOKHA provided a larger operating corridor (60% more) as compared with the standard supra-orbital key-hole approaches while preserving the same degree of minimally invasive nature and cosmetic results.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4103/0028-3886.294827DOI Listing
May 2021

Patient-Centric Head and Neck Cancer Radiation Therapy: Role of Advanced Imaging.

Neuroimaging Clin N Am 2020 Aug 11;30(3):341-357. Epub 2020 Jun 11.

Department of Radiology, Section of Neuroradiology, Froedtert and Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, USA. Electronic address:

The traditional 'one-size-fits-all' approach to H&N cancer therapy is archaic. Advanced imaging can identify radioresistant areas by using biomarkers that detect tumor hypoxia, hypercellularity etc. Highly conformal radiotherapy can target resistant areas with precision. The critical information that can be gleaned about tumor biology from these advanced imaging modalities facilitates individualized radiotherapy. The tumor imaging world is pushing its boundaries. Molecular imaging can now detect protein expression and genotypic variations across tumors that can be exploited for tailoring treatment. The exploding field of radiomics and radiogenomics extracts quantitative, biologic and genetic information and further expands the scope of personalized therapy.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.nic.2020.04.005DOI Listing
August 2020

Dual Energy Computed Tomography in Head and Neck Imaging: Pushing the Envelope.

Neuroimaging Clin N Am 2020 Aug 10;30(3):311-323. Epub 2020 Jun 10.

Augmented Intelligence & Precision Health Laboratory, Department of Radiology, Research Institute of the McGill University Health Centre, 1001 Decarie Boulevard, Montreal, Quebec H4A 3J1, Canada; Department of Radiology, McGill University, 1650 Cedar Avenue, Montreal, Quebec H3G 1A4, Canada; Segal Cancer Centre, Lady Davis Institute for Medical Research, Jewish General Hospital, 3755 Cote Suite-Catherine Road, Montreal, Quebec H3T 1E2, Canada; Gerald Bronfman Department of Oncology, McGill University, Suite 720, 5100 Maisonneuve Boulevard West, Montreal, Quebec H4A3T2, Canada; Department of Otolaryngology, Head and Neck Surgery, Royal Victoria Hospital, McGill University Health Centre, 1001 boul. Decarie Boulevard, Montreal, Quebec H3A 3J1, Canada. Electronic address:

Multiple applications of dual energy computed tomography (DECT) have been described for the evaluation of disorders in the head and neck, especially in oncology. We review the body of evidence suggesting advantages of DECT for the evaluation of the neck compared with conventional single energy computed tomography scans, but the full potential of DECT is still to be realized. There is early evidence suggesting significant advantages of DECT for the extraction of quantitative biomarkers using radiomics and machine learning, representing a new horizon that may enable this technology to reach its full potential.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.nic.2020.04.003DOI Listing
August 2020

Radiomic Features of Multiparametric MRI Present Stable Associations With Analogous Histological Features in Patients With Brain Cancer.

Tomography 2020 06;6(2):160-169

Radiology.

Magnetic resonance (MR)-derived radiomic features have shown substantial predictive utility in modeling different prognostic factors of glioblastoma and other brain cancers. However, the biological relationship underpinning these predictive models has been largely unstudied, and the generalizability of these models had been called into question. Here, we examine the localized relationship between MR-derived radiomic features and histology-derived "histomic" features using a data set of 16 patients with brain cancer. Tile-based radiomic features were collected on T1, post-contrast T1, FLAIR, and diffusion-weighted imaging (DWI)-derived apparent diffusion coefficient (ADC) images acquired before patient death, with analogous histomic features collected for autopsy samples coregistered to the magnetic resonance imaging. Features were collected for each original image, as well as a 3D wavelet decomposition of each image, resulting in 837 features per MR and histology image. Correlative analyses were used to assess the degree of association between radiomic-histomic pairs for each magnetic resonance imaging. The influence of several confounds was also assessed using linear mixed-effect models for the normalized radiomic-histomic distance, testing for main effects of different acquisition field strengths. Results as a whole were largely heterogeneous, but several features showed substantial associations with their histomic analogs, particularly those derived from the FLAIR and postcontrast T1W images. These features with the strongest association typically presented as stable across field strengths as well. These data suggest that a subset of radiomic features can consistently capture texture information on underlying tissue histology.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.18383/j.tom.2019.00029DOI Listing
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC7289245PMC
June 2020

Posterior quadrant disconnection for sub-hemispheric drug refractory epilepsy.

Neurol India 2020 Mar-Apr;68(2):270-273

Department of Neurosurgery, All India Institute of Medical Sciences, New Delhi, India.

The posterior quadratic epilepsy (PQE) is a form of a multilobar epilepsy, involving the temporal-parietal and occipital lobes. Basically, epilepsies with localized networks to the posterior temporal, posterior parietal, and occipital lobes can benefit from this type of surgery. Gliosis due to perinatal insult and cortical dysplasis and angiomas in Sturge Weber syndrome involving the PQ have often been cited in the literature as the etiology for PQE. However, before considering surgery, it is important to localize the epileptogenic focus through a complete pre operative work up involving; EEG (Electro-Encephalo-Graphy), video EEG, single photon emission computed tomography (SPECT), positron emission tomography (PET), and magneto encephalography (MEG). Historically, these pathologies were dealt with multi-lobar resections, which were associated with high morbidity and mortality, owing to blood loss, especially in young children, hydrocephalus, and hemosiderosis. Based on the theory of networks involved in epileptogenesis, the concept of disconnection in epilepsy surgery was introduced. Delalande and colleagues, described the technique of hemispheric disconnection (functional hemispherectomy) for pathologies like: hemimegalencephaly, rasmussens encephalitis involving the entire hemisphere. The technique has evolved with time, moving towards minimally invasive endoscopic vertical hemispherotomy, described by Chandra and colleagues. The posterior quadrant disconnection (PQD) evolved as a tailored disconnection on similar lines as hemispherotomy, for managing refractory epilepsy arising from the posterior quadrant. The technique and principles involved in the PQD surgery are similar to the those of peri-insular hemispherotomy and has been described in the literature by few authors. The technique of performing PQD will be described here in a step-wise fashion with illustrations supplemented by a surgical video.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4103/0028-3886.284358DOI Listing
March 2021

Memory Performance and Quantitative Neuroimaging Software in Mild Cognitive Impairment: A Concurrent Validity Study.

J Int Neuropsychol Soc 2020 11 28;26(10):954-962. Epub 2020 Apr 28.

Medical College of Wisconsin, 9200 W Wisconsin Ave, Milwaukee, WI53226, USA.

Objective: This study examined the relationship between patient performance on multiple memory measures and regional brain volumes using an FDA-cleared quantitative volumetric analysis program - Neuroreader™.

Method: Ninety-two patients diagnosed with mild cognitive impairment (MCI) by a clinical neuropsychologist completed cognitive evaluations and underwent MR Neuroreader™ within 1 year of testing. Select brain regions were correlated with three widely used memory tests. Regression analyses were conducted to determine if using more than one memory measures would better predict hippocampal z-scores and to explore the added value of recognition memory to prediction models.

Results: Memory performances were most strongly correlated with hippocampal volumes than other brain regions. After controlling for encoding/Immediate Recall standard scores, statistically significant correlations emerged between Delayed Recall and hippocampal volumes (rs ranging from .348 to .490). Regression analysis revealed that evaluating memory performance across multiple memory measures is a better predictor of hippocampal volume than individual memory performances. Recognition memory did not add further predictive utility to regression analyses.

Conclusions: This study provides support for use of MR Neuroreader™ hippocampal volumes as a clinically informative biomarker associated with memory performance, which is a critical diagnostic feature of MCI phenotype.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1017/S1355617720000454DOI Listing
November 2020

Longitudinal Reproducibility of MR Perfusion Using 3D Pseudocontinuous Arterial Spin Labeling With Hadamard-Encoded Multiple Postlabeling Delays.

J Magn Reson Imaging 2020 06 30;51(6):1846-1853. Epub 2019 Nov 30.

Department of Radiology, Medical College of Wisconsin, Milwaukee, Wisconsin, USA.

Background: Arterial spin labeling (ASL) can be confounded by varying arterial transit times (ATT) across the brain and with disease. Hadamard encoding schemes can be applied to 3D pseudocontinuous ASL (pCASL) to acquire ASL data with multiple postlabeling delays (PLDs) to estimate ATT and then correct cerebral blood flow (CBF).

Purpose: To assess the longitudinal reproducibility of 3D pCASL with Hadamard-encoded multiple PLDs.

Study Type: Prospective, longitudinal.

Population: Fifty-two healthy, right-handed male subjects who underwent imaging at four timepoints over 45 days.

Field Strength/sequence: A Hadamard-encoded 3D pCASL sequence was acquired at 3.0T with seven PLDs from 1.0-3.7 sec.

Assessment: ATT and corrected CBF (cCBF) were computed. Conventional uncorrected CBF (unCBF) was also estimated. Within- and between-subject coefficient of variation (wCV and bCV, respectively) and intraclass correlation coefficient (ICC) were evaluated across four time intervals: 7, 14, 30, and 45 days, in gray matter and 17 independent regions of interest (ROIs). A power analysis was also conducted.

Statistical Tests: A repeated-measures analysis of variance (ANOVA) was used to compare ATT, cCBF, and unCBF across the four scan sessions. A paired two-sample t-test was used to compare cCBF and unCBF. Pearson's correlation was used to examine the relationship between the cCBF and unCBF difference and ATT. Power calculations were completed using both the cCBF and unCBF variances.

Results: ATT showed the lowest wCV and bCV (3.3-4.4% and 6.0-6.3%, respectively) compared to both cCBF (10.5-11.7% and 20.6-22.2%, respectively) and unCBF (12.0-13.6% and 22.7-23.7%, respectively). wCV and bCV were lower for cCBF vs. unCBF. A significant difference between cCBF and unCBF was found in most regions (P = 5.5 × 10 -3.8 × 10 in gray matter) that was highly correlated with ATT (R = 0.79-0.86). A power analysis yielded acceptable power at feasible sample sizes using cCBF.

Data Conclusion: ATT and ATT-corrected CBF were longitudinally stable, indicating that ATT and CBF changes can be reliably evaluated with Hadamard-encoded 3D pCASL with multiple PLDs.

Level Of Evidence: 1 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2020;51:1846-1853.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1002/jmri.27007DOI Listing
June 2020

Perineural Tumor Spread in Head and Neck Malignancies.

Semin Roentgenol 2019 Jul 6;54(3):258-275. Epub 2019 Mar 6.

University of Pittsburgh Medical Center, Department of Radiology, Pittsburgh, PA. Electronic address:

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1053/j.ro.2019.03.003DOI Listing
July 2019

Sinonasal Neoplasms.

Semin Roentgenol 2019 Jul 9;54(3):244-257. Epub 2019 Mar 9.

Department of Radiology, Division of Neuroradiology, University of Iowa, Iowa City, IA.

View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.1053/j.ro.2019.03.001DOI Listing
July 2019

Craniovertebral junction evaluation by computed tomography in asymptomatic individuals in the Indian population.

Neurol India 2018 May-Jun;66(3):797-803

Department of Neurosurgery, All Institute of Medical Sciences, New Delhi, India.

Background: The available literature on the anatomy and imaging of the craniovertebral junction (CVJ) focusses on the osteometric indices described for the detection of abnormal relationships between the components of CVJ. However, a knowledge of the normal osteometry of this region in the Indian population is critically important for the operating surgeon as it may influence the surgical technique as well as the choice, size and configurations of the implants. It is also important to determine whether critical differences exist between the osteometric data of Indians and the rest of the world for this part of the anatomy. Accordingly, the present study is an attempt to quantitate the osteometric indices for the anatomically normal CVJ in Indian subjects.

Materials And Methods: We retrospectively studied the imaging data of 49 consecutive adult patients (31 males, 18 females) who underwent a computed tomographic (CT) angiogram for suspected vascular conditions unrelated to the craniovertebral junction. Several parameters related to the atlanto-dental relationship, foramen magnum, atlas and axis vertebrae were recorded, including the dimensions of the commonly instrumented bony regions and also the indices related to the CVJ bony relationships. The data was also compared between the two genders, statistically through the Student's t-test using the statistical program "R".

Results: No patient had an atlanto dens interval >2.5 mm. The mean distance of the odontoid tip from the McRae line in this series was 5.11 mm and no patient had the odontoid tip above the McRae line. Female subjects had significantly smaller diameters of C1 lateral masses and odontoid screw trajectory length when compared to males. Additionally, in the Indian population, the length range of odontoid screw trajectory and the thickness of the narrowest part of the C2 pedicles was smaller with respect to similar data from other geographical regions. However, the rest of the parameters resembled the data from studies conducted on populations with other ethnicities.

Conclusion: The osteometric parameters of the CVJ in the Indian population are largely similar to those described globally. However, there are some important differences too which can influence the design of surgical implants suited to the Indian population.
View Article and Find Full Text PDF

Download full-text PDF

Source
http://dx.doi.org/10.4103/0028-3886.232288DOI Listing
September 2019
-->