Background: To develop a deep learning model to classify primary bone tumors from preoperative radiographs and compare performance with radiologists. Methods: A total of 1356 patients (2899 images) with histologically confirmed primary bone tumors and preoperative radiographs were identified from five institutions' pathology databases. Manual cropping was performed by radiologists to label the lesions. Binary discriminatory capacity (benign versus not-benign and malignant versus not-malignant) and three-way classification (benign versus intermediate versus malignant) performance of our model were evaluated. The generalizability of our model was investigated on data from external test set. Final model performance was compared with interpretation from five radiologists of varying level of experience using the Permutations tests. Findings: For benign vs. not benign, model achieved area under curve (AUC) of 0894 and 0877 on cross-validation and external testing, respectively. For malignant vs. not malignant, model achieved AUC of 0907 and 0916 on cross-validation and external testing, respectively. For three-way classification, model achieved 721% accuracy vs. 746% and 721% for the two subspecialists on cross-validation (p = 003 and p = 052, respectively). On external testing, model achieved 734% accuracy vs. 693%, 734%, 731%, 679%, and 634% for the two subspecialists and three junior radiologists (p = 014, p = 089, p = 093, p = 002, p < 001 for radiologists 1À5, respectively). Interpretation: Deep learning can classify primary bone tumors using conventional radiographs in a multiinstitutional dataset with similar accuracy compared to subspecialists, and better performance than junior radiologists.
Background Radiologists have difficulty distinguishing benign from malignant bone lesions because these lesions may have similar imaging appearances. The purpose of this study was to develop a deep learning algorithm that can differentiate benign and malignant bone lesions using routine magnetic resonance imaging (MRI) and patient demographics. Methods 1,060 histologically confirmed bone lesions with T1- and T2-weighted pre-operative MRI were retrospectively identified and included, with lesions from 4 institutions used for model development and internal validation, and data from a fifth institution used for external validation. Image-based models were generated using the EfficientNet-B0 architecture and a logistic regression model was trained using patient age, sex, and lesion location. A voting ensemble was created as the final model. The performance of the model was compared to classification performance by radiology experts. Findings The cohort had a mean age of 30±23 years and was 58.3% male, with 582 benign lesions and 478 malignant. Compared to a contrived expert committee result, the ensemble deep learning model achieved (ensemble vs. experts): similar accuracy (0·76 vs. 0·73, p=0·7), sensitivity (0·79 vs. 0·81, p=1·0) and specificity (0·75 vs. 0·66, p=0·48), with a ROC AUC of 0·82. On external testing, the model achieved ROC AUC of 0·79. Interpretation Deep learning can be used to distinguish benign and malignant bone lesions on par with experts. These findings could aid in the development of computer-aided diagnostic tools to reduce unnecessary referrals to specialized centers from community clinics and limit unnecessary biopsies. Funding This work was funded by a Radiological Society of North America Research Medical Student Grant (#RMS2013) and supported by the Amazon Web Services Diagnostic Development Initiative.
An I2-promoted, metal-free domino protocol for one-pot synthesis of 1,3,4-oxadiazoles has been developed via oxidative cleavage of C(sp(2))-H or C(sp)-H bonds, followed by cyclization and deacylation. In this reaction, the use of K2CO3 as a base is found to be an essential factor in the cyclization and the C-C bond cleavage. This procedure proceeded smoothly in moderate to high yields with good functional group compatibility.
A hydroxyl‐assisted, organophotoredox/cobalt dual catalyzed annulation of 2‐propynolphenols to form 2‐hydroxymethyl‐benzo[b]furans was developed by employing 1,2,3,5‐tetrakis(carbazol‐9‐yl)‐4,6‐dicyanobenzene (4CzIPN) as photosensitizer and CoCl2(PPh3)2/5,5′‐dimethyl‐2,2′‐bipyridine as cobalt catalytic precursor. Various substrates and functional groups were tolerated. The practical applications of this reaction were further demonstrated by enlarged gram‐scale and various derivations for complex heterocycles. Primary mechanistic studies suggested the involvement of cobalt‐hydride mediated hydrogen atom transfer (HAT) process.
A concise and efficient protocol for PhI(OAc)2 oxidation halogenation of quinoline at the C5 position was developed, affording the desired remote C–H activation products in moderate to excellent yields.
Background:The aim of this study was to compare the biomechanical characteristics between bilateral and unilateral lumbo-iliac fixation in unilateral comminuted sacral fractures (USF) by finite element analysis.Methods:A 3-dimensional finite element model of unilateral sacral fractures was simulated. Three kinds of implants were instrumented into the model, including the unilateral lumbopelvic fixation (ULF), bilateral lumbopelvic fixation (BLF), and unilateral iliac fixation with bilateral lumbar pedicle screws (UBF). Loads of compression and rotation were distributed to the superior endplate of L3. To evaluate the biomechanical properties, the construct stiffness, the micromotion of the fractures, the stress distribution of implants, and the balance of hemilumbar vertebra are recorded and analyzed.Results:The highest construct stiffness was provided by BLF. In BLF model, the displacement between iliums was only 0.009 mm (compressional) and 0.001 mm (rotational), which was less than that under normal condition (0.02 mm). The maximum von Misses stress of implants appeared on the UBF. By using unilateral fixation, the L4 endured obvious imbalance on bilateral hemivertebra. A marked difference was exposed in BLF and UBF models, and the equilibrium of stress and activity was shown.Conclusion:From the finite element view, the stability of ULF is insufficient to reconstruct the posterior pelvic ring. Furthermore, the unilateral fixation may lead to imbalance of lumbar vertebra and pelvis. On the contrary, the BLF can provide satisfied stability and lumbar balance.
Detecting small target of ship at sea is very important to missile guidance. In this paper, infrared images are used to capture the information of small ship at sea, an edge detector combining with a median filter are used to seek for the sky-sea line and the objects in the images. According to the features of ship targets in sky-sea background, a criterion is set up to distinguish the small ship target from other object. A complete algorithm is introduced, it was simulated in computer with a set of 700 images. The results showed that this algorithm could locate precisely the positions of the small ship targets.
An ovelp rotocol for the synthesis of 3arylated quinolined erivatives has been developed using silver carbonate (Ag 2 CO 3 )a nd dioxygen( O 2 ) as the oxidants.I nt his method, quinolines acting as the parentr eagentr eact with arenes under palladium acetate [Pd(OAc) 2 ]c atalysisa long with 1-adamantanecarboxylic acid (Adm). Thec ombination of the 1-adamantanecarboxylic acid ligand and the palladium catalyst is found to be an essential factor for achieving high activity and selectivity,a nd allows aw ide range of 3-arylquinolined erivatives to be obtained in high yields. Moreover, ar elevant mechanismi sp roposed, revealing that un der the palladium(II) catalysisamigratory insertion to C-3 À Ho fq uinolines and insertion to the C À Ho f arenes are key steps in this reaction. Mild reaction conditions and high selectivity of the reactives ite provide potential for promising applications in drug discoverya nd functional materials.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.