Analysis of vertebrae without fracture on spine MRI to assess bone fragility: A Comparison of Traditional Machine Learning and Deep Learning

Author(s):Show less -

Ramos, Jonathan S. ; de Aguiar, Erikson J. ; Belizario, Ivar, V ; Costa, Marcus V. L. ; Maciel, Jamilly G. ; Cazzolato, Mirela T. ; Traina, Caetano, Jr. ; Nogueira-Barbosa, Marcello H. ; Traina, Agma J. M. ; Shen, L ; Gonzalez, AR ; Santosh, KC ; Lai, Z ; Sicilia, R ; Almeida, JR ; Kane, B

• Document type: Journal article
• Source: 2022 IEEE 35TH INTERNATIONAL SYMPOSIUM ON COMPUTER-BASED MEDICAL SYSTEMS (CBMS); v. N/A, p. 6-pg., 2022-01-01.

Abstract

Bone mineral density (BMD) is the international standard for evaluating osteoporosis/osteopenia. The success rate of BMD alone in estimating the risk of vertebral fragility fracture (VFF) is approximately 50%, making BMD far from ideal in predicting VFF. In addition, whether or not a patient has been diagnosed with osteoporosis or osteopenia, he or she may suffer a VFF. For this reason, we conducted an extensive empirical study to assess VFFs in postmenopausal women. We considered a representative dataset of 94 T1- and T2-weighted routine spine MRI (with osteopenia or osteoporosis), split into 2,400 samples (slices). Comparing the classification results of machine learning and deep learning (DL) techniques showed that DL generally achieved better results at the cost of higher computational power and hard explainability. ResNet achieved the best results in discriminating patients from groups with and without VFFs with 83% accuracy and 90% AUC (with a confidence interval of 99%). Our results represent a significant step toward prospective and longitudinal studies investigating methods to achieve higher accuracy in predicting VFFs based on spine MRI features of vertebrae without fracture. (AU)

• FAPESP's process: 21/02412-0 - Development of a platform for similarity query execution on complex data using multimetric indexing structures
• Grantee: Jonathan da Silva Ramos
• Support Opportunities: Scholarships in Brazil - Technical Training Program - Technical Training
• FAPESP's process: 20/11258-2 - Interoperability and similarity queries on medical databases
• Grantee: Mirela Teixeira Cazzolato
• Support Opportunities: Scholarships in Brazil - Post-Doctoral
• FAPESP's process: 16/17078-0 - Mining, indexing and visualizing Big Data in clinical decision support systems (MIVisBD)
• Grantee: Agma Juci Machado Traina
• Support Opportunities: Research Projects - Thematic Grants
• FAPESP's process: 21/11403-5 - Mining multimodal records: explainable patterns and anomalies discovery
• Grantee: Mirela Teixeira Cazzolato
• Support Opportunities: Scholarships abroad - Research Internship - Post-doctor
• FAPESP's process: 21/00360-3 - Development of software to support similarity queries in healthcare databases
• Grantee: Ivar Vargas Belizario
• Support Opportunities: Scholarships in Brazil - Technical Training Program - Technical Training
• FAPESP's process: 21/08982-3 - Security and privacy in machine learning models to medical images against adversarial attacks
• Grantee: Erikson Júlio de Aguiar
• Support Opportunities: Scholarships in Brazil - Doctorate
• FAPESP's process: 18/04266-9 - Multiparametric analysis of lumbar vertebrae texture on magnetic resonance imaging and correlation with fragility fractures.
• Grantee: Marcello Henrique Nogueira Barbosa
• Support Opportunities: Regular Research Grants