IMPACT OF PREPROCESSING AND COMPARISON OF NEURAL NETWORK ENSEMBLE METHODS FOR SEGMENTATION OF THE THORACIC SPINE IN X-RAY IMAGES

Authors

  • V. D. Koniukhov A. Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine, Kharkiv, Ukraine
  • O. M. Morgun “Laboratory of X-ray Medical Equipment” LTD, Kharkiv, Ukraine
  • K. E. Nemchenko V. N. Karazin Kharkiv National University, Kharkiv, Ukraine

DOI:

https://doi.org/10.15588/1607-3274-2024-4-10

Keywords:

machine learning; image recognition; neural network; image segmentation, computer vision

Abstract

Context. Automatic segmentation of medical images plays an important role in the process of automating the detection of various diseases in the spine and the use of radiography is the most accessible means of predicting diseases. Over the years many studies have been conducted on the topic of image segmentation. One of the many methods for improving image segmentation is the use of neural network ensembles.

Objective. The aims of this study were to investigate the impact of preprocessing and compare the main methods of neural network ensembles and their effect on the segmentation of the thoracic region, in this study the area was considered which consists of the vertebrae: Th8, Th9, Th10, Th11.

Method. To begin with, the influence of preprocessing of X-ray images was considered, which included the following methods: histogram equalization for contrast enhancement, contrast-limited adaptive histogram equalization, logarithmic transform method, median filter, Gaussian filter, and bilateral filter. To study the influence of neural network ensemble on segmentation quality, several methods were used. Averaging method – a simple half-averaging method. Weighted averaging method – an improved version of the averaging method which uses weights for each network, the higher the network weight, the greater its influence on averaging. Method of cumulative averaging – a modified averaging method in which each ensemble receives an averaged image, after which all the results of the ensembles are averaged. Bagging – method of averaging networks trained on different data, n networks are used, the training sample is divided into n parts, and each neural network is trained on its own subset of data, as a result, the averaging method is used for predictions. Averaging method for a large number of networks – in this method, 100 neural networks were trained, after which the averaging method was used. Method of averaging mask shapes – this method uses a distance transform to average multiple masks into one shape average.

Results. It was investigated that the use of different methods of image preprocessing does not guarantee an improvement in the quality of segmentation of the spine region on X-ray images, but even on the contrary worsens the quality of segmentation. Different methods of combining predictions of neural network ensembles were considered, which made it possible to find out the pros and cons of specific methods for the task of segmentation of X-ray images.

Conclusions. The experiments conducted allowed us to conclude that the use of any preprocessing methods should not be used for segmentation of X-ray images. Also, due to a large number of architectures and methods for combining predictions, the behavior of ensemble methods was studied, which will allow us to further determine the necessary approach for segmentation of X-ray images. Further study of the weighted averaging method and the mask shape averaging method will make it possible to improve the obtained result and achieve even greater success in segmentation.

Author Biographies

V. D. Koniukhov, A. Pidhornyi Institute of Power Machines and Systems of NAS of Ukraine, Kharkiv

Postgraduate student

O. M. Morgun, “Laboratory of X-ray Medical Equipment” LTD, Kharkiv

PhD, Director

K. E. Nemchenko, V. N. Karazin Kharkiv National University, Kharkiv

Dr. Sc., Head of the Department

References

Wu H., Wu X., Wu T., Miao X., Zheng S., Huang G., Cheng X. Detection Ewingella americana from a patient with Andersson lesion in ankylosing spondylitis by metagenomic next-generation sequencing test: a case report, BMC Musculoskelet Disord, 2024, Vol. 25, P. 568. DOI: 10.1186/s12891-024-07680-y

Zhou Y., Huang X., Liu Y., Zhou X., Liu Q. Destructive Cryptococcal Osteomyelitis Mimicking Tuberculous Spondylitis, American Journal of Case Reports, 2024, Vol. 25, P. e944291. DOI: 10.12659/AJCR.944291

Smolle M. A., Maier A., Lindenmann J., Porubsky C., Leithner J., Smolle-Juettner F. M. Esophageal perforation with near fatal mediastinitis secondary to Th3 fracture, Wiener klinische Wochenschrift, 2024. DOI: 10.1007/s00508024-02397-3

Sønderby A. H., Thomsen H., Skals R. G., Storm S., Leutscher P.D.C., Simony A. Thoracic spine X-ray examination of patients with back pain using different breathing technique and exposure times – A diagnostic study, Radiography, 2024, Vol. 30, pp 582–288. DOI: 10.1016/j.radi.2024.01.011

Kjelle E., Chilanga C. The assessment of image quality and diagnostic value in X-ray images: a survey on radiographers’ reasons for rejecting images, Insights into imaging, 2022, Vol. 13(1), № 36. DOI: 10.1186/s13244-022-01169-9

Ullman G. Quantifying image quality in diagnostic radiology using simulation of the imaging system and model observers. Linköping, Sweden, 2008, 85 p.

Dietterich T. G. Ensemble Methods in Machine Learning, Multiple Classifier Systems. MCS 2000. Springer. Berlin,Heidelberg, 2000, pp. 1–15. (Lecture Notes in Computer Science, Vol. 1857). DOI: 10.1007/3-540-45014-9_1

Mohammed A., Kora R. A comprehensive review on ensemble deep learning: Opportunities and challenges, Journal of King Saud University – Computer and Information Sciences, 2023, Vol. 35(2), pp. 757–774. DOI: 10.1016/j.jksuci.2023.01.014

Maclin R., Opitz D. W. Popular Ensemble Methods: An Empirical Study, Journal of Artificial Intelligence Research, 1999, Vol. 11, pp. 169–198.

Lu H., Li M., Zhang Y., Yu L. Lumbar spine segmentation method based on deep learning, Journal of applied clinical medical physics, 2023, Vol. 24, Iss. 6, P. e13996. DOI: 10.1002/acm2.13996

Xiong X. Graves S. A., Gross B. A., Buatti J. M., Beichel R. R. Lumbar and Thoracic Vertebrae Segmentation in CT Scans Using a 3D Multi-Object Localization and Segmentation CNN, Tomography, 2024, Vol. 10(5), pp. 738– 760. DOI: 10.3390/tomography10050057

Li H., Luo H., Huan W., Shi Z., Yan C., Wang L., Mu Y., Liu Y. Automatic lumbar spinal MRI image segmentation with a multi-scale attention network, Neural computing & applications, 2021, Vol. 33(18), pp. 11589–11602. DOI: 10.1007/s00521-021-05856-4

Khandelwal P., Collins L. D., Siddiqi K. Spine and Individual Vertebrae Segmentation in Computed Tomography Images Using Geometric Flows and Shape Priors, Frontiers in Computer Science, 2021, Vol. 3. DOI: 10.3389/fcomp.2021.592296

Mushtaq M., Akram M. U., Alghamdi N. S., Fatima J., Masood R. F. Localization and Edge-Based Segmentation of Lumbar Spine Vertebrae to Identify the Deformities Using Deep Learning Models, Sensors, 2022, Vol. 22(4), P. 1547. DOI: 10.3390/s22041547

Liang Y., Fang Y. T., Lin T. C., Yang C. R., Chang C. C., Chang H. K., Ko C. C., Tu T. H., Fay L. Y., Wu J. C., Huang W. C., Hu H. W., Chen Y. Y., Kuo C. H. The Quantitative Evaluation of Automatic Segmentation in Lumbar Magnetic Resonance Images, Neurospine, 2024, Vol. 21(2). pp. 665–675. DOI: 10.14245/ns.2448060.030

Zhou Z., Wang S., Zhang S., Pan X., Yang H., Zhuang Y., Lu Z. Deep learning-based spinal canal segmentation of computed tomography image for disease diagnosis: A proposed system for spinal stenosis diagnosis, Medicine, 2024, Vol. 103(18), P. e37943. DOI: 10.1097/MD.0000000000037943

Vindr.ai Datasets: SpineXR. [Electronic resource]. Access mode: https://vindr.ai/datasets/spinexr

Downloads

Published

2024-12-26

How to Cite

Koniukhov, V. D., Morgun, O. M., & Nemchenko, K. E. (2024). IMPACT OF PREPROCESSING AND COMPARISON OF NEURAL NETWORK ENSEMBLE METHODS FOR SEGMENTATION OF THE THORACIC SPINE IN X-RAY IMAGES. Radio Electronics, Computer Science, Control, (4), 102–112. https://doi.org/10.15588/1607-3274-2024-4-10

Issue

Section

Neuroinformatics and intelligent systems