Evaluation of Diagnostic Value of Mediastinum for Differentiation of Drug Sensitive, Multi and Extensively Drug Resistant Tuberculosis Using Chest X-Rays

Tulo, Sukanta Kumar; Ramu, Palaniappan; Swaminathan, Ramakrishnan

doi:10.1016/j.irbm.2022.02.004

Cited by 7 publications

(16 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When classifying between DS-TB and DR-TB, we can conclude that the suggested method is 30.01% more accurate than the alternative methods. Comparing the classification results of DR-TB and MDR-TB between the proposed approaches of Jaeger et al [ 24 ] and Tulo et al [ 12 ], utilizing all KPIs, the proposed method provides a superior answer by 2.52–46.27% (using the information provided in Table 9 ). The classification results of DR-TB and XDR-TB utilizing the data presented in Table 10 demonstrated that the proposed approach is 1.71–5.52% superior to the method of Tulo et al [ 12 ].…”

Section: Computational Resultsmentioning

confidence: 99%

“…Comparing the classification results of DR-TB and MDR-TB between the proposed approaches of Jaeger et al [ 24 ] and Tulo et al [ 12 ], utilizing all KPIs, the proposed method provides a superior answer by 2.52–46.27% (using the information provided in Table 9 ). The classification results of DR-TB and XDR-TB utilizing the data presented in Table 10 demonstrated that the proposed approach is 1.71–5.52% superior to the method of Tulo et al [ 12 ]. Finally, Table 11 reveals that the proposed approaches produce solutions that are 2.43% to 6.67% more accurate than the method of Tulo et al…”

Section: Computational Resultsmentioning

confidence: 99%

“…Table 1 indicates that the total numbers of DS-TB, DR-TB, MDR-TB, and XDR-TB have 1607, 468, 2087, and 857 datasets, respectively; or 32.02%, 9.32%, 41.58%, and 17.08%, respectively. These data will be used to train and test for the accuracy of the proposed methods, compared with the methodology proposed by Ureta and Shrestha [ 76 ], Tulo et al [ 77 ], Jaeger et al [ 24 ], Kovalev et al [ 78 ], Tulo et al [ 12 ], and Karki et al [ 11 ]. The details of the compared method are presented in Table 2 .…”

Section: Methodsmentioning

confidence: 99%

“…The Original CXR is then cropped, and the lungs are segmented inside the bounding box that has been diminished. Tulo et al [ 12 ] used machine learning (ML) to examine the lungs and mediastinum, in order to classify the TB patients’ drug response. Despite the use of binary categorization [ 19 ] in this study, other kinds of drug-resistant tuberculosis, including DS, MDR, and XDR, are predicted.…”

Section: Related Workmentioning

confidence: 99%

“…Machine learning was ultimately used by Tulo et al [ 12 ] to categorize four distinct types of drug-resistant TB. This includes DS-TB, DR-TB, MDR, and XDR, which some refer to forms of tuberculosis that have developed resistance to at least some of the available treatments.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Embedded AMIS-Deep Learning with Dialog-Based Object Query System for Multi-Class Tuberculosis Drug Response Classification

et al. 2022

View full text Add to dashboard Cite

A person infected with drug-resistant tuberculosis (DR-TB) is the one who does not respond to typical TB treatment. DR-TB necessitates a longer treatment period and a more difficult treatment protocol. In addition, it can spread and infect individuals in the same manner as regular TB, despite the fact that early detection of DR-TB could reduce the cost and length of TB treatment. This study provided a fast and effective classification scheme for the four subtypes of TB: Drug-sensitive tuberculosis (DS-TB), drug-resistant tuberculosis (DR-TB), multidrug-resistant tuberculosis (MDR-TB), and extensively drug-resistant tuberculosis (XDR-TB). The drug response classification system (DRCS) has been developed as a classification tool for DR-TB subtypes. As a classification method, ensemble deep learning (EDL) with two types of image preprocessing methods, four convolutional neural network (CNN) architectures, and three decision fusion methods have been created. Later, the model developed by EDL will be included in the dialog-based object query system (DBOQS), in order to enable the use of DRCS as the classification tool for DR-TB in assisting medical professionals with diagnosing DR-TB. EDL yields an improvement of 1.17–43.43% over the existing methods for classifying DR-TB, while compared with classic deep learning, it generates 31.25% more accuracy. DRCS was able to increase accuracy to 95.8% and user trust to 95.1%, and after thae trial period, 99.70% of users were interested in continuing the utilization of the system as a supportive diagnostic tool.

show abstract

Section: Computational Resultsmentioning

confidence: 99%

Section: Computational Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Embedded AMIS-Deep Learning with Dialog-Based Object Query System for Multi-Class Tuberculosis Drug Response Classification

et al. 2022

View full text Add to dashboard Cite

show abstract

Evaluation of geometric differences between right and left lungs in bacterial pneumonia chest radiographs

Tulo

Govindarajan

Ramu

et al. 2023

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

Geometric Analysis of Structural Changes in Microscopic Nuclei Images for Drug-Induced Cytotoxic Assessment

Murali

Govindarajan

Swaminathan

2023

J. Mech. Med. Biol.

View full text Add to dashboard Cite

Investigation of drug-induced structural changes in cell lines at different concentrations using microscopic images is essential to understand their cytotoxic effects. In this study, geometric shape descriptors to evaluate the toxicity effects of a particular drug in cell images are formulated. For this, fluorescence microscopic images of drug-untreated and drug-treated mouse cardiac muscle HL1 cells are considered. Ratiometric index of cellular to non-cellular area and, Zernike moment measures are calculated for three different thresholds at different drug concentrations namely 0.6, 1.2, 2.5, 5, and 10[Formula: see text][Formula: see text]M. Statistical analysis is performed to find the significant features. Classification is performed using Support Vector Machine (SVM) to differentiate drug untreated with treated cells at different concentrations. Results demonstrate that the proposed features are able to characterize the shape variations in cell images at different concentrations, and validates the efficacy of segmentation. Mean cellular area ratio is found to decrease from drug-untreated to drug-treated at various concentrations. Significant shape alterations in cellular structures are also obtained using Zernike moment measures for these cases. The machine learning approach using SVM provides better performance in classifying the drug untreated with progressively increasing drug concentrations. Hence, the proposed pipeline of methods could be clinically used to determine the maximum permissible drug tolerance levels during the development of new drugs.

show abstract

Evaluation of Diagnostic Value of Mediastinum for Differentiation of Drug Sensitive, Multi and Extensively Drug Resistant Tuberculosis Using Chest X-Rays

Cited by 7 publications

References 30 publications

Embedded AMIS-Deep Learning with Dialog-Based Object Query System for Multi-Class Tuberculosis Drug Response Classification

Embedded AMIS-Deep Learning with Dialog-Based Object Query System for Multi-Class Tuberculosis Drug Response Classification

Evaluation of geometric differences between right and left lungs in bacterial pneumonia chest radiographs

Geometric Analysis of Structural Changes in Microscopic Nuclei Images for Drug-Induced Cytotoxic Assessment

Contact Info

Product

Resources

About