Machine learning analyses can differentiate meningioma grade by features on magnetic resonance imaging DOI Open Access
Andrew T. Hale, David P. Stonko, Li Wang

et al.

Neurosurgical FOCUS, Journal Year: 2018, Volume and Issue: 45(5), P. E4 - E4

Published: Nov. 1, 2018

OBJECTIVEPrognostication and surgical planning for WHO grade I versus II meningioma requires thoughtful decision-making based on radiographic evidence, among other factors. Although conventional statistical models such as logistic regression are useful, machine learning (ML) algorithms often more predictive, have higher discriminative ability, can learn from new data. The authors used an array of ML to predict atypical radiologist-interpreted preoperative MRI findings. goal this study was compare the performance standard methods when predicting grade.METHODSThe cohort included patients aged 18-65 years with (n = 94) 34) in whom obtained between 1998 2010. A board-certified neuroradiologist, blinded histological grade, interpreted all MR images tumor volume, degree peritumoral edema, presence necrosis, location, a draining vein, patient sex. trained validated several binary classifiers: k-nearest neighbors models, support vector machines, naïve Bayes classifiers, artificial neural networks well grade. area under curve-receiver operating characteristic curve comparison across within model classes. All analyses were performed MATLAB using MacBook Pro.RESULTSThe 6 imaging demographic variables: sex, vein construct models. outperformed true-positive false-positive (receiver characteristic) space (area 0.8895).CONCLUSIONSML powerful computational tools that great accuracy.

Language: Английский

Will traditional biopsy be substituted by radiomics and liquid biopsy for breast cancer diagnosis and characterisation? DOI
Filippo Pesapane, Matteo B. Suter, Anna Rotili

et al.

Medical Oncology, Journal Year: 2020, Volume and Issue: 37(4)

Published: March 16, 2020

Language: Английский

Citations

50
A Comprehensive Review on Radiomics and Deep Learning for Nasopharyngeal Carcinoma Imaging DOI Creative Commons
Song Li,

Yuqin Deng,

Zhiling Zhu

et al.

Diagnostics, Journal Year: 2021, Volume and Issue: 11(9), P. 1523 - 1523

Published: Aug. 24, 2021

Nasopharyngeal carcinoma (NPC) is one of the most common malignant tumours head and neck, improving efficiency its diagnosis treatment strategies an important goal. With development combination artificial intelligence (AI) technology medical imaging in recent years, increasing number studies have been conducted on image analysis NPC using AI tools, especially radiomics neural network methods. In this review, we present a comprehensive overview research based deep learning. These depict promising prospect for NPC. The deficiencies current potential learning are discussed. We conclude that future should establish large-scale labelled dataset images focused screening necessary.

Language: Английский

Citations

41
Development and Validation of a Deep Learning Model for Brain Tumor Diagnosis and Classification Using Magnetic Resonance Imaging DOI Creative Commons

Peiyi Gao,

Wei Shan,

Yue Guo

et al.

JAMA Network Open, Journal Year: 2022, Volume and Issue: 5(8), P. e2225608 - e2225608

Published: Aug. 8, 2022

Importance

Deep learning may be able to use patient magnetic resonance imaging (MRI) data aid in brain tumor classification and diagnosis.

Objective

To develop clinically validate a deep system for automated identification of 18 types tumors from MRI data.

Design, Setting, Participants

This diagnostic study was conducted using collected between 2000 2019 37 871 patients. A segmentation intracranial based on T1- T2-weighted images T2 contrast sequences developed tested. The accuracy the tested 1 internal 3 external independent sets. clinical value assessed by comparing neuroradiologists with vs without assistance proposed separate test set. Data were analyzed March through February 2020.

Main Outcomes Measures

Changes neuroradiologist scans evaluated.

Results

trained among patients (mean [SD] age, 41.6 [11.4] years; 519 women [48.9%]). It achieved mean area under receiver operating characteristic curve 0.92 (95% CI, 0.84-0.99) 1339 4 centers’ sets diagnosis tumors. Higher outcomes found compared sensitivity similar specificity (for 300 Tiantan Hospital set: accuracy, 73.3% [95% 67.7%-77.7%] 60.9% 46.8%-75.1%]; sensitivity, 88.9% 85.3%-92.4%] 53.4% 41.8%–64.9%]; specificity, 96.3% 94.2%-98.4%] 97.9%; 97.3%-98.5%]). With system, 1166 increased 12.0 percentage points, 63.5% 60.7%-66.2%) 75.5% 73.0%-77.9%) assistance.

Conclusions Relevance

These findings suggest that system–based associated improved neuroradiologists.

Language: Английский

Citations

33
Deep-learning models for image-based gynecological cancer diagnosis: a systematic review and meta- analysis DOI Creative Commons
Asefa Adimasu Taddese, Binyam Tilahun,

Tadesse Awoke

et al.

Frontiers in Oncology, Journal Year: 2024, Volume and Issue: 13

Published: Jan. 11, 2024

Gynecological cancers pose a significant threat to women worldwide, especially those in resource-limited settings. Human analysis of images remains the primary method diagnosis, but it can be inconsistent and inaccurate. Deep learning (DL) potentially enhance image-based diagnosis by providing objective accurate results. This systematic review meta-analysis aimed summarize recent advances deep techniques for gynecological cancer using various explore their future implications.

Language: Английский

Citations

7
Machine learning analyses can differentiate meningioma grade by features on magnetic resonance imaging DOI Open Access
Andrew T. Hale, David P. Stonko, Li Wang

et al.

Neurosurgical FOCUS, Journal Year: 2018, Volume and Issue: 45(5), P. E4 - E4

Published: Nov. 1, 2018

OBJECTIVEPrognostication and surgical planning for WHO grade I versus II meningioma requires thoughtful decision-making based on radiographic evidence, among other factors. Although conventional statistical models such as logistic regression are useful, machine learning (ML) algorithms often more predictive, have higher discriminative ability, can learn from new data. The authors used an array of ML to predict atypical radiologist-interpreted preoperative MRI findings. goal this study was compare the performance standard methods when predicting grade.METHODSThe cohort included patients aged 18-65 years with (n = 94) 34) in whom obtained between 1998 2010. A board-certified neuroradiologist, blinded histological grade, interpreted all MR images tumor volume, degree peritumoral edema, presence necrosis, location, a draining vein, patient sex. trained validated several binary classifiers: k-nearest neighbors models, support vector machines, naïve Bayes classifiers, artificial neural networks well grade. area under curve-receiver operating characteristic curve comparison across within model classes. All analyses were performed MATLAB using MacBook Pro.RESULTSThe 6 imaging demographic variables: sex, vein construct models. outperformed true-positive false-positive (receiver characteristic) space (area 0.8895).CONCLUSIONSML powerful computational tools that great accuracy.

Language: Английский

Citations

57