The future of computer-aided diagnostics in chest computed tomography

Николаев, А. Е.; Chernina, Valeria; Blokhin, Ivan; Shapiev, A. N.; Gonchar, Anna P.; Gombolevskiy, Victor A.; Petraikin, A. V.; Silin, A. Yu.; Petrova, Galina D.; Morozov, S. P.

doi:10.17116/hirurgia201912191

Cited by 4 publications

(1 citation statement)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A E Nikolaev et al in 2019 discussed the possibilities of machine learning in the chest computed tomography. Even though the amount of information that a radiologist has to process is snowballing due to the use of computed tomography for lung screening, the automatic image analysis makes it possible for more studies to be interpreted (Nikolaev et al, 2019[ 28 ]). Nikolaev felt that most software developers operationalize radiology since it is a medical specialty with high digitalization.…”

Section: Reviewmentioning

confidence: 99%

Artificial intelligence and its future potential in lung cancer screening

Mathew

David

Mathew

2020

EXCLI Journal; 19:Doc1552; ISSN 1611-2156

View full text Add to dashboard Cite

Section: Reviewmentioning

confidence: 99%

Artificial intelligence and its future potential in lung cancer screening

Mathew

David

Mathew

2020

EXCLI Journal; 19:Doc1552; ISSN 1611-2156

View full text Add to dashboard Cite

Comparability of coronary risk assessment methods with chest ultra-LDCT and CT coronography with ECG synchronization

Николаев¹,

Korkunova²,

Khutoroi

et al. 2021

Medicinskaâ vizualizaciâ

View full text Add to dashboard Cite

Purpose. To assess the comparability of coronary calcium values measured on ultralow-dose computed tomography studies without ECG-synchronization versus a) non-contrast computed tomography with ECG synchronization, b) CT coronography with ECG synchronization.Materials and methods. The study comprised 283 studies: 68 patients who underwent contrast-free ultra-LDCT without ECG synchronization and contrast-free CT with ECG synchronization performed in a single visit, and 49 patients with contrast-free ultra-LDCT without ECG synchronization, non-contrast CT with ECG synchronization, and CT coronography with ECG synchronization and intravenous injection of contrast agent, also carried out in one visit, meeting all inclusion and exclusion criteria of the study.Quantitative coronary calcium values were calculated with the Agatston score and the CAC-DRS scale (score of calcification degree from 0 to 3 and the number of affected arteries from 0 to 4 points). The degree of coronary artery stenosis was analyzed with CAD-RADS scale (0-5).The above parameters were compared using visual/quantitative assessment of coronary calcium on ultra-LDCT without ECG synchronization and visual/quantitative assessment for CT with ECG synchronization, as well as the degree of stenosis on CT coronography in the same patients.Results. Based on the results of accuracy indices comparison, the possibility to use quantitative scale (Agatston score, CAC-DRS quantitative scale) to assess coronary calcification in the lung cancer screening in comparison with ECG-synchronized CT was determined during interpretation of ultra-LDCT without ECG synchronization. The correlation matrix to assess correlation between visual, quantitative scales of coronary artery changes and calcification at ultra-LDCT without ECG synchronization and quantitative scale at CT with ECG synchronization vs. CT coronography identifies very strong positive statistically significant correlations.Conclusion. Methods of coronary calcinosis assessment with chest ultra-LDCT and CT with ECG synchronization are comparable, therefore it is possible to assess coronary calcium in lung cancer screening by ultra-LDCT data at a reliable-high level using both quantitative and visual CAC-DRS scales.

show abstract

Automated analysis of lung lesions in COVID-19: comparison of standard and low-dose CT

Blokhin¹,

Solovev²,

Vladzymyrskyy³

et al. 2023

SJCEM

View full text Add to dashboard Cite

Introduction. Chest computed tomography (CT) plays a prominent role in determining the extent of pulmonary parenchymal lesions in COVID-19. At the same time, subjectivity of lung lesion volume assessment using 0-4 CT scale in COVID-19 and gradual introduction of low-dose CT (LDCT) requires an investigation of semi-automated lung segmentation accuracy in LDCT compared to CT.Study Objective. To compare the accuracy of affected lung tissue volume calculation between CT and LDCT in COVID-19 using a semi-automatic segmentation program.Material and Methods. The retrospective study was performed on data from the earlier prospective multicenter study registered at ClinicalTrials.gov, NCT04379531. CT and LDCT data were processed in 3D Slicer software with Lung CT Segmenter and Lung CT Analyzer extensions, and the volume of affected lung tissue and lung volume were determined by thresholding.Results. The sample size was 84 patients with signs of COVID-19-associated pneumonia. Mean age was 50.6 ± 13.3 years, and the median body mass index (BMI) was 28.15 [24.85; 31.31] kg/m2. The effective doses were 10.1 ± 3.26 mSv for the standard CT protocol and 2.64 mSv [1.99; 3.67] for the developed LDCT protocol. The analysis of absolute lung lesion volume in cubic centimeters with Wilcoxon Signed Ranks Test revealed a statistically significant difference between CT and LDCT (p-value < 0.001). No statistically significant differences were found in the relative values of lung tissue lesion volume (lesion volume/lung volume) between CT and LDCT using Wilcoxon Signed Ranks Test (p-value = 0.95).Conclusion. The reliability of developed LDCT protocol in COVID-19 for the semi-automated calculation of affected tissue percentage was comparable to the standard chest CT protocol when using 3D Slicer with Lung CT Segmenter and Lung CT Analyzer extensions.

show abstract

The future of computer-aided diagnostics in chest computed tomography

Cited by 4 publications

References 46 publications

Artificial intelligence and its future potential in lung cancer screening

Artificial intelligence and its future potential in lung cancer screening

Comparability of coronary risk assessment methods with chest ultra-LDCT and CT coronography with ECG synchronization

Automated analysis of lung lesions in COVID-19: comparison of standard and low-dose CT

Contact Info

Product

Resources

About