Spinal BT Taramasında Multipl Miyelom ve Osteolitik Kemik Metastazının Ayırt Edilmesi: Konvolüsyonel Sinir Ağı Kullanan Kapsamlı Bir Çalışma

Year 2025, Volume: 22 Issue: 1, 1 - 7

Muhammet Kürşat Şimşek , Yusuf Kenan Çetinoğlu , Resul Bircan , Ali Balcı

https://doi.org/10.35440/hutfd.1563046

Abstract

Amaç: Spinal multipl miyelom (MM) ve osteolitik metastatik kemik tümörünün (OMKT) doğru bir şekilde ayırt edilmesi zorlayıcı olabilir, genellikle doğru tanı için görüntüleme yöntemleri, laboratuvar testleri ve biyopsinin kombinasyonu uygulanır. Bu çalışmada, MM ve OMKT hastalarından elde edilen BT görüntülerini CNN modelleri kullanarak ayırt etmeyi amaçladık.
Materyal ve Metod: Ocak 2015 ile Ocak 2023 arasında elde edilen 91 hastanın (1886 OMKT görüntüsü ve 1821 MM görüntüsü; 46 erkek ve 45 kadın; ortalama yaş: 61,2 yıl) 3707 BT görüntüsü incelendi. Eğitim seti için 2667, doğrulama seti için 740, test seti için 300 görüntü rastgele seçildi. DenseNet121, DenseNet169, EfficientNetB0, MobileNet, MobileNetV2, VGG16 ve Xception CNN mimarilerine dayanan bir transfer öğrenimi yaklaşımı kullanıldı. Modellerin performansı değerlendirildi.
Bulgular: Modellerin MM ve OMKT ayrımındaki duyarlılık, özgüllük, pozitif prediktif değer, negatif prediktif değer, doğruluk, F1 skoru ve kappa ölçümleri değerlendirildiğinde, en başarılı modeller sırasıyla %88, %86,33 ve %86 doğruluk oranları ile MobileNetV2, MobileNet ve VGG16 olmuştur.
Sonuç: Çalışmamızda CNN tabanlı yapay zekâ modellerinin BT görüntülerinde MM ve OMKT'yi ayırt edebileceğini gösterdik.

Keywords

Yapay Zekâ, Multipl Miyelom, Spinal Metastaz

Differentiating Multiple Myeloma and Osteolytic Bone Metastasis on Spinal CT Scan: A Comprehensive Study Using Convolutional Neural Network

Abstract

Background: Accurate differentiation of spinal multiple myeloma (MM) and osteolytic metastatic bone tumor (OMBT) can be challenging. Usually, imaging methods, laboratory tests, and biopsy are performed for the correct diagnosis. In this study, we aimed to differentiate CT images from patients with MM and OMBT using CNN models.
Materials and Methods: 3707 CT images of 91 patients (1886 OMBT images and 1821 MM images; 46 males and 45 females; mean age: 61.2 years) obtained between January 2015 and January 2023 were reviewed. 2667 images were randomly selected for the training set, 740 for the validation set, and 300 for the test set. A transfer learning approach was used based on DenseNet121, DenseNet169, EfficientNetB0, MobileNet, MobileNetV2, VGG16, and Xception CNN architectures. The performance of the models was evaluated.
Results: When the sensitivity, specificity, positive predictive value, negative predictive value, accuracy, F1 score, and kappa measurements of the models in the MM and OMBT differentiation are evaluated, the most successful ones are MobileNetV2, MobileNet, and VGG16, with accuracy of 88%, 86.33%, and 86%, respectively.
Conclusions: Our study showed that CNN-based artificial intelligence models can differentiate MM and OMBT on CT images.

Keywords

Artificial Intelligence, Multiple Myeloma, Spinal Metastasis

References

• 1. van de Donk NWCJ, Pawlyn C, Yong KL. Multiple myeloma. Lancet. 2021;397(10272):410-427. doi: 10.1016/S0140-6736(21)00135-5.
• 2. Fornetti J, Welm AL, Stewart SA. Understanding the Bone in Cancer Metastasis. J Bone Miner Res. 2018;33(12):2099-2113. doi: 10.1002/jbmr.3618.
• 3. Sahgal A, Myrehaug SD, Siva S, Masucci GL, Maralani PJ, Brundage M, et all. Stereotactic body radiotherapy versus conventional external beam radiotherapy in patients with painful spinal metastases: an open-label, multicentre, ran-domised, controlled, phase 2/3 trial. Lancet Oncol. 2021;22(7):1023-1033. doi: 10.1016/S1470-2045(21)00196-0.
• 4. Hillengass J, Usmani S, Rajkumar SV, Durie BGM, Mateos MV, Lonial S, et all. International myeloma working group consensus recommendations on imaging in monoclonal plasma cell disorders. Lancet Oncol. 2019 Jun;20(6):e302-e312. doi: 10.1016/S1470-2045(19)30309-2. Erratum in: Lancet Oncol. 2019;20(7):e346. doi: 10.1016/S1470-2045(19)30423-1.
• 5. Gao R, Zhao S, Aishanjiang K, Cai H, Wei T, Zhang Y, et all. Deep learning for differential diagnosis of malignant hepatic tumors based on multi-phase contrast-enhanced CT and clini-cal data. J Hematol Oncol. 2021;14(1):154. doi: 10.1186/s13045-021-01167-2.
• 6. Chea P, Mandell JC. Current applications and future directions of deep learning in musculoskeletal radiology. Skeletal Radiol. 2020;49(2):183-197. doi: 10.1007/s00256-019-03284-z.
• 7. Chen K, Cao J, Zhang X, Wang X, Zhao X, Li Q, et all. Differen-tiation between spinal multiple myeloma and metastases originated from lung using multi-view attention-guided net-work. Front Oncol. 2022;12:981769. doi: 10.3389/fonc.2022.981769.
• 8. D'Angelo T, Caudo D, Blandino A, Albrecht MH, Vogl TJ, Grue-newald LD, et all. Artificial intelligence, machine learning and deep learning in musculoskeletal imaging: Current applica-tions. J Clin Ultrasound. 2022;50(9):1414-1431. doi: 10.1002/jcu.23321.
• 9. Y. Fu. Image classification via fine-tuning with EfficientNet, 2020. [accessed June 25, 2023, updated July 10, 2023]. 2023 [cited 2023 Jun 25]. Available from: https://keras.io/examples/vision/image_classification_efficientnet_fine_tuning/
• 10. Vollmer A, Saravi B, Vollmer M, Lang GM, Straub A, Brands RC, et all. Artificial Intelligence-Based Prediction of Oroantral Communication after Tooth Extraction Utilizing Preoperative Panoramic Radiography. Diagnostics (Basel). 2022;12(6):1406. doi: 10.3390/diagnostics12061406.
• 11. Cetinoglu YK, Koska IO, Uluc ME, Gelal MF. Detection and vascular territorial classification of stroke on diffusion-weighted MRI by deep learning. Eur J Radiol. 2021;145:110050. doi: 10.1016/j.ejrad.2021.110050.
• 12. Lang N, Zhang Y, Zhang E, Zhang J, Chow D, Chang P, et all. Differentiation of spinal metastases originated from lung and other cancers using radiomics and deep learning based on DCE-MRI. Magn Reson Imaging. 2019;64:4-12. doi: 10.1016/j.mri.2019.02.013.
• 13. Xiong X, Wang J, Hu S, Dai Y, Zhang Y, Hu C. Differentiating Between Multiple Myeloma and Metastasis Subtypes of Lumbar Vertebra Lesions Using Machine Learning-Based Ra-diomics. Front Oncol. 2021;11:601699. doi: 10.3389/fonc.2021.601699.
• 14. Baykara M, Yildirim M. Differentiation of multiple myeloma and metastases with apparent diffusion coefficient map his-togram analysis. North Clin Istanb. 2022 Jul 5;9(3):256-260. doi: 10.14744/nci.2021.59376.
• 15. O'Connor SD, Yao J, Summers RM. Lytic metastases in thora-columbar spine: computer-aided detection at CT--preliminary study. Radiology. 2007 Mar;242(3):811-6. doi: 10.1148/radiol.2423060260. Erratum in: Radiology. 2007;244(1):320.
• 16. Wels M, Kelm BM, Tsymbal A, Hammon M, Soza G, Sühling M, et al. Multi-stage osteolytic spinal bone lesion detection from CT data with internal sensitivity control. SPIE Proceed-ings. 2012;8315:831513. doi:10.1117/12.911169
• 17. Mutlu U, Balci A, Özsan GH, Özkal S, Şeyhanli A, Özgül HA. Computed tomography characteristics of multiple myeloma and other osteolytic metastatic bone lesions. Acta Radiol. 2021;62(12):1639-1647.