A. A. Skachkov

A. A. Skachkov

Auto-generated profile, based on publicly available information

A. A. Skachkov

LLC “United IT Space

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 June 2024 in Journal of radiology and nuclear medicine
Verified
Journal Article
Open access
PDF file
NNN. В. Nudnov
N. В. Nudnov
AKA. V. Korobov
A. V. Korobov
АSА. А. Skachkov
А. А. Skachkov
ТКТ. В. Кульнева
Т. В. Кульнева
VSV. V. Sherstoboev
V. V. Sherstoboev
MIMikhail Evgenevich Ivannikov
Mikhail Evgenevich Ivannikov
Objective: comparative evaluation of output data of a set of trained convolutional neural network (CNN) models and interpretation of pathological changes in lumbar spine by radiologists during magnetic resonance imaging.Material and methods. More than 12,000 anonymized archives were collected to generate training and test neural network datasets from patients aged over 18 years. Each archive consisted of a set of programs in two planes containing T2-TSE, T1-TSE and T2 sequences with fat suppression program. Subsequently, the selected studies were tagged in two steps, directly consisting of manual tagging and its validation by experts. CNN training was performed separately for normal analysis, qualitative detection of individual pathological changes, and quantitative analysis. The accuracy of the models was verified by comparing the protocols of five radiologists and the output of CNN models in two steps. The first, intermediate stage evaluated the accuracy of the neural networks in detecting disc bulges, protrusions and extrusions, spinal canal stenosis, lateral stenosis, foraminal stenosis, spondylolisthesis and facet joint arthrosis. In the final stage, in addition to the pathologies considered in the intermediate one, the accuracy of detecting degenerative changes of the occlusive plates, synovitis of intervertebral joints, intervertebral discs degeneration, osteophytes, transitional vertebrae, hypertrophy of yellow ligaments and Schmorl’s hernia was tested. The reference value for all pathological changes considered in this paper was determined by majority vote and, in case of disagreement, by an external radiologist. The radiologists’ interpretations were then compared with those of the trained model.Results. The artificial intelligence (AI) showed comparable sensitivity and specificity values compared to the reference result in a group of experienced radiologists for binary classification (presence/absence) of individual lumbosacral spine degenerative changes. The sensitivity and specificity of AI results were 0.88 and 0.97 for extrusions, 0.81 and 0.94 for protrusions, 0.87 and 0.98 for central stenosis, 0.83 and 0.85 for lateral stenosis, 0.92 and 0.84 for foraminal stenosis, 0.85 and 0.5 for osteoarthritis, 0.73 and 0.96 for occlusive plates degeneration, 0.85 and 0.84 for intervertebral joint synovitis, 0.91 and 0.88 for osteophytes, 0.93 and 0.72 for intervertebral disc degeneration, 1.0 and 1.0 for transitional vertebrae, 0.8 and 1.0 for spondylolisthesis, 0.67 and 0.99 for yellow ligament hypertrophy, and 0.75 and 1.0 for Schmorl’s hernia, respectively. The accuracy of quantitative size characterization of lumbosacral spine protrusions and extrusions showed unsatisfactory results, but improvements in the quality of determination of these parameters are planned in future work.Conclusion. AI models showed comparable performance to expert radiologists in detecting lumbosacral spine degenerative changes. Consistent improvement of CNN models based on comparative evaluation with radiologists improves the sensitivity and specificity of pathologic change detection.

20 January 2025 in Medical Visualization
Verified
Journal Article
Closed access
PDF file
ARA. S. Rusakov
A. S. Rusakov
VTV. V. Tumko
V. V. Tumko
RSR. S. Sarbaev
R. S. Sarbaev
NUN. A. Uspenskaya
N. A. Uspenskaya
NNN. V. Nudnov
N. V. Nudnov
MIMikhail Evgenevich Ivannikov
Mikhail Evgenevich Ivannikov
Objective: to develop a convolutional neural network (CNN) model to classify the presence and severity of central lumbar spinal stenosis (LSS) during MRI examination and to demonstrate its effectiveness as an accurate and consistent diagnostic tool. Methods. Morphological classifications of LSS and quantitative measurements of key anatomical structures were combined using two CNNs. To classify central stenosis, models were trained on 1635 labeled lumbar spine MRI studies consisting of T2-weighted sagittal and axial planes at the level of each vertebra. The accuracy of the model was assessed using an external validation set of 150 MRI studies graded by a panel of 7 radiologists as: no stenosis, mild, moderate or severe spinal canal stenosis. The reference value for all types of stenosis was determined by majority vote, and in the event of disagreement, a decision was made by an external radiologist. The radiologists' interpretations were then compared with those of the trained model. Results. The model demonstrated comparable performance to the average radiologist both in identifying the presence/absence of LSS and in classifying severity for all 3 types of stenosis. For central canal stenosis, the sensitivity and specificity of the CNN were (0.93; 0.85) for binary classification (presence/absence) compared to the average radiologist (0.86; 0.86). For lateral pocket stenosis, the sensitivity and specificity of CNN were (0.92; 0.80) compared to the radiologist's mean (0.83; 0.94). For foraminal stenosis, the sensitivity and specificity of CNN were (0.89; 0.86) compared to the radiologist's mean (0.81; 0.91). Multiclass classification of stenosis severity showed similar statistics. Conclusions. CNNs showed comparable performance to radiologists in detecting and classifying LSS. The integration of neural network models in pathology detection could provide higher accuracy, efficiency, systematicity, and the possibility of subsequent interpretation in diagnostic practice.