Effect of a computer-aided diagnosis system on radiologists' performance in grading gliomas with MRI

Abstract

The effects of a computer-aided diagnosis (CAD) system based on quantitative intensity features with magnetic resonance (MR) imaging (MRI) were evaluated by examining radiologists' performance in grading gliomas. The acquired MRI database included 71 lower-grade gliomas and 34 glioblastomas. Quantitative image features were extracted from the tumor area and combined in a CAD system to generate a prediction model. The effect of the CAD system was evaluated in a two-stage procedure. First, a radiologist performed a conventional reading. A sequential second reading was determined with a malignancy estimation by the CAD system. Each MR image was regularly read by one radiologist out of a group of three radiologists. The CAD system achieved an accuracy of 87% (91/105), a sensitivity of 79% (27/34), a specificity of 90% (64/71), and an area under the receiver operating characteristic curve (Az) of 0.89. In the evaluation, the radiologists’ Az values significantly improved from 0.81, 0.87, and 0.84 to 0.90, 0.90, and 0.88 with p = 0.0011, 0.0076, and 0.0167, respectively. Based on the MR image features, the proposed CAD system not only performed well in distinguishing glioblastomas from lower-grade gliomas but also provided suggestions about glioma grading to reinforce radiologists’ confidence rating.

Radiomics Nomogram Building From Multiparametric MRI to Predict Grade in Patients With Glioma: A Cohort Study

“…The nomogram was developed based on radiomics features from three MRI sequences and clinical factors with high classification accuracy achieved in both the training cohort and validation cohort. The performance of nomogram using the proposed method is higher than other studies, like Hsieh et al, 19 who used only CET 1 WI sequences to classify gliomas with an accuracy of 0.87 obtained. This is lower than the accuracy (0.929) of the radiomic signature in our study.…”

“…14,15 Radiomics has been widely used in tumor diagnosis, prognosis prediction, treatment selection, and so on. [15][16][17][18] Hsieh et al 19 used global and local intensity features extracted from only a contrast-enhanced T 1 -weighted (CET 1 WI) sequence to generate a computer-aided diagnosis (CAD) for distinguishing glioblastomas from lower-grade glioma patients. Their results showed that MRI texture assessments were useful for accurate tumor grading.…”

A machine learning texture model for classifying lung cancer subtypes using preliminary bronchoscopic findings

“…Nevertheless, this is a preliminary experiment regarding the use of quantitative image features and machine learning techniques to classify different lung cancer subtypes via bronchoscopic images. In the future, a larger prospective study comparing physicians and the CAD system should be conducted . More cases should be added in future studies to enhance the generalization of the proposed method.…”

“…In the future, a larger prospective study comparing physicians and the CAD system should be conducted. 25 More cases should be added in future studies to enhance the generalization of the proposed method. Simultaneously, correlations between image findings and pathology findings can also be explored.…”

Computer-Aided Diagnoses (CAD) System: An Artificial Neural Network Approach to MRI Analysis and Diagnosis of Alzheimer’s Disease (AD)

“…In the diagnosis of gliomas with MRI images the CAD system achieved an accuracy of 87 percent, a sensitivity of 79 percent and a specificity of 90 percent. The radiologist results were lower with an 81 percent for accuracy, 87 percent for sensitivity, and 84 percent for specificity [33].…”