Enhancing brain tumor classification with a diffusion denoising model and a conditional deep convolutional neural network

Abstract

The limited availability of medical imaging datasets and concerns over patient privacy pose significant challenges in artificial intelligence-driven disease diagnosis. To overcome these limitations, this study introduces the use of the denoising diffusion model (DDM) for generating synthetic datasets, marking a significant advancement over traditional generative adversarial networks (GANs). This research pioneers the integration of DDM with conditional deep convolutional neural networks (CDCNN) for brain tumor classification, focusing on four categories: Glioma, meningioma, pituitary tumors, and healthy tissue. The proposed CDCNN model, developed from existing convolutional neural network architectures, effectively processed both DDM-generated synthetic datasets and original datasets sourced from the Kaggle repository. The results demonstrate the remarkable efficacy of the DDM-based augmentation framework, with the CDCNN model achieving an accuracy of 96.2%, significantly outperforming traditional GAN-based models, such as Pix2Pix. A comparative analysis against established architectures, including ResNet50, Visual Geometry Group (VGG)16, VGG19, and InceptionV3, further highlights the superior sensitivity, specificity, and F1 score of the proposed framework. These findings underscore the transformative potential of diffusion models in enhancing dataset diversity, improving classification performance, and addressing data scarcity issues in medical imaging. The proposed framework offers a scalable, robust solution for brain tumor diagnosis, paving the way for improved disease prediction and treatment planning in clinical practice.

Keywords

Brain tumors / Conditional deep convolutional neural network / Denoising diffusion model / Synthetic data augmentation / Medical image classification / Generative adversarial networks

Funding

None.

Conflict of interest

The authors declare that they have no conflicts of interest.

References

[1] Khalighi S, Reddy K, Midya A, Pandav KB, Madabhushi A, Abedalthagafi M. Artificial intelligence in neuro-oncology: Advances and challenges in brain tumor diagnosis, prognosis, and precision treatment. NPJ Precis Oncol. 2024; 8(1):80. doi: 10.1038/s41698-024-00575-0

[2] International Agency for Research on Cancer. Cancer Today. Available from: https://gco.iarc.fr/today [Last accessed on 2024 Mar 21].

[3] D’Cruz CE, Shirodkar RK, Pathak Y, Kumar L. Nanoparticles for Brain Tumor Imaging and Therapy. Berlin: Springer; 2024. p. 345-372. doi: 10.1007/978-981-97-0308-1_14

[4] Nadeem MW, Al Ghamdi MA, Hussain M, et al. Brain tumor analysis empowered with deep learning: A review, taxonomy, and future challenges. Brain Sci. 2020; 10(2):118. doi: 10.3390/brainsci10020118

[5] Luzzi S, Lucifero G, Rabski A, Kadri PAS, Al-Mefty O. The party wall: redefining the indications of transcranial approaches for giant pituitary adenomas in endoscopic era. Cancers (Basel). 2023; 15(8):2235. doi: 10.3390/cancers15082235

[6] Ghandour F, Squassina A, Karaky R, et al. Presenting psychiatric and neurological symptoms and signs of brain tumors before diagnosis: A systematic review. Brain Sci. 2021; 11(3):301. doi: 10.3390/brainsci11030301

[7] Fuemmeler BF, Elkin TD, Mullins LL. Survivors of childhood brain tumors: Behavioral, emotional, and social adjustment. Clin Psychol Rev. 2002; 22(4):547-585. doi: 10.1016/S0272-7358(01)00120-9

[8] Onakpojeruo EP, Mustapha MT, Ozsahin DU, Ozsahin I. A comparative analysis of the novel conditional deep convolutional neural network model, using conditional deep convolutional generative adversarial network-generated synthetic and augmented brain tumor datasets for image classification. Brain Sci. 2024; 14(6):559. doi: 10.3390/brainsci14060559

[9] Ozsahin DU, Onakpojeruo EP, Uzun B, Ozsahin I. Selection methods for the treatment of spinal cord tumors using analytical evaluation models. In: Advances in Science and Engineering Technology International Conferences (ASET); 2023. doi: 10.1109/ASET56582.2023.10180782

[10] Uzun Ozsahin D, Onakpojeruo EP, Uzun B. Hydrogel-based drug delivery nanoparticles with conventional treatment approaches for cancer tumors; a comparative study using MCDM technique. In: Advances in Science and Engineering Technology International Conferences (ASET); 2023. doi: 10.1109/ASET56582.2023.10180659

[11] Hussain S, Mubeen I, Ullah N, et al. Modern diagnostic imaging technique applications and risk factors in the medical field: A review. Biomed Res Int. 2022;2022:5164970. doi: 10.1155/2022/5164970

[12] Piorkowski A, Obuchowicz R, Najjar R. Redefining radiology: A review of artificial intelligence integration in medical imaging. Diagnostics (Basel). 2023; 13(17):2760. doi: 10.3390/diagnostics13172760

[13] Uzun Ozsahin D, Onakpojeruo EP, Uzun B, Mustapha MT, Ozsahin I. Mathematical assessment of machine learning models used for brain tumor diagnosis. Diagnostics (Basel). 2023; 13(4):618. doi: 10.3390/diagnostics13040618

[14] Abdusalomov AB, Mukhiddinov M, Whangbo TK. Brain tumor detection based on deep learning approaches and magnetic resonance imaging. Cancers (Basel). 2023; 15(16):4172. doi: 10.3390/cancers15164172

[15] Neamah K, Mohamed F, Adnan MM, et al. Brain tumor classification and detection based DL models: A systematic review. IEEE Access. 2024; 12:251-2542. doi: 10.1109/ACCESS.2023.3347545

[16] Lotan E, Tschider C, Sodickson DK, et al. Medical imaging and privacy in the era of artificial intelligence: Myth, fallacy, and the future. J Am Coll Radiol. 2020; 17(9):1159. doi: 10.1016/j.jacr.2020.04.007

[17] Paul M, Maglaras L, Ferrag MA, Almomani I. Digitization of healthcare sector: A study on privacy and security concerns. ICT Express. 2023; 9(4):571-588. doi: 10.1016/j.icte.2023.05.004

[18] Goodfellow IJ, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets. In: Advances in Neural Information Processing Systems. Proceedings of the 27th International Conference on Neural Information Processing Systems:Canada; 2014. p. 2672-2680.

[19] Muhammad A, Salman Z, Lee K, Han D. Harnessing the power of diffusion models for plant disease image augmentation. Front Plant Sci. 2023;14:1280496. doi: 10.3389/fpls.2023.1280496

[20] Ahmed MS, Hossain S, Haque MN, et al. MRI based automated detection of brain tumor using DWT, GLCM, PCA, ensemble of SVM and PNN in sequence. In: Lecture Notes in Computer Sciences, LNICST. Vol. 490. Berlin: Springer; 2023. p. 267-279. doi: 10.1007/978-3-031-34619-4_22

[21] Kang J, Ullah Z, Gwak J. MRI-based brain tumor classification using ensemble of deep features and machine learning classifiers. Sensors (Basel). 2021; 21(6):2222. doi: 10.3390/s21062222

[22] Ramakrishnan T, Sankaragomathi B. A professional estimate on the computed tomography brain tumor images using SVM-SMO for classification and MRG-GWO for segmentation. Pattern Recognit Lett. 2017; 94:163-171. doi: 10.1016/j.patrec.2017.03.026

[23] Eze MC, Vafaei LE, Eze CT, Tursoy T, Ozsahin DU, Mustapha MT. Development of a novel multi-modal contextual fusion model for early detection of Varicella Zoster Virus skin lesions in human subjects. Processes. 2023; 11(8):2268. doi: 10.3390/pr11082268

[24] Cheng J. Brain Tumor Dataset. Figshare; 2017. doi: 10.6084/m9.figshare.1512427

[25] Ayadi W, Elhamzi W, Charfi I, Atri M. A hybrid feature extraction approach for brain MRI classification based on bag-of-words. Biomed Signal Process Control. 2019; 48:144-152. doi: 10.1016/j.bspc.2018.10.010

[26] Rehman A, Naz S, Razzak MI, Akram F, Imran M. A deep learning-based framework for automatic brain tumors classification using transfer learning. Circuits Syst Signal Process. 2020; 39(2):757-775. doi: 10.1007/s00034-019-01246-3

[27] Ullah N, Khan JA, Khan MS, et al. An effective approach to detect and identify brain tumors using transfer learning. Appl Sci. 2022; 12(11):5645. doi: 10.3390/app12115645

[28] Priyadarshini P, Khairuzzaman AKM, Kanungo P. Brain tumor detection and classification from MRI images using cascaded deep neural networks. In: Lecture Notes in Electrical Engineering. Vol. 976. Singapore:Springer Nature Singapore; 2023. p. 301-311. doi: 10.1007/978-981-99-0412-9_26

[29] Ahmmed S, Podder P, Mondal MRH, et al. Enhancing brain tumor classification with transfer learning across multiple classes: An in-depth analysis. BiomedInformatics. 2023; 3(4):1124-1144. doi: 10.3390/biomedinformatics3040068

[30] Jain J, Kubadia M, Mangla M, Tawde P. Comparison of transfer learning techniques to classify brain tumours using MRI images. Eng Proc. 2024; 59(1):144. doi: 10.3390/engproc2023059144

[31] Veni N, Manjula J. VGG-16 architecture for MRI brain tumor image classification. In: Lecture Notes in Electrical Engineering. Vol. 966. Singapore:Springer Nature Singapore; 2023. p. 319-328. doi: 10.1007/978-981-19-8338-2_26

[32] Srinivas C, Nandini NP, Zakariah M, et al. Deep transfer learning approaches in performance analysis of brain tumor classification using MRI images. J Healthc Eng. 2022;2022:3264367. doi: 10.1155/2022/3264367

[33] Suryawanshi S, Patil SB. Brain tumor detection using YOLOv5 and Faster R-CNN. Int J Intell Syst Appl Eng. 2023; 12(4s):335-342. doi: 10.1109/ViTECoN58111.2023.10157773

[34] Kesana A, Nallola J, Bootapally RT, Amaraneni S, Reddy GS. Brain tumor detection using YOLOv5 and Faster R-CNN. In: 2023 2nd International Conference on Vision Towards Emerging Trends in Communication and Networking Technologies (ViTECoN). IEEE; 2023. p. 1-6. doi: 10.1109/ViTECoN58111.2023.10157773

[35] Chlap P, Min H, Vandenberg N, et al. A review of medical image data augmentation techniques for deep learning applications. J Med Imaging Radiat Oncol. 2021; 65(5):545-563. doi: 10.1111/1754-9485.13261

[36] Mangaokar N, Pu J, Bhattacharya P, Reddy CK, Viswanath B. Jekyll: Attacking medical image diagnostics using deep generative models. In: Proceedings of EuroS&P 2020. Genoa, Italy; 2020 p. 39-157. doi: 10.1109/EuroSP48549.2020.00017

[37] Salvi M, Branciforti F, Molinari F, Meiburger KM. Generative models for color normalization in digital pathology and dermatology: Advancing the learning paradigm. Expert Syst Appl. 2024;245:123105. doi: 10.1016/j.eswa.2023.123105

[38] Kaur J, Singh AK, Jindal N. Systematic survey on generative adversarial networks for brain tumor segmentation and classification. Concurrency Comput Pract Exper. 2023; 35(27):e7850. doi: 10.1002/cpe.7850

[39] Myronenko A. 3D MRI brain tumor segmentation using autoencoder regularization. In: Lecture Notes in Computer Science. Vol. 11384. Cham, Switzerland: Springer Nature; 2019. p. 311-320. doi: 10.1007/978-3-030-11726-9_28

[40] Huang P, Li D, Jiao Z, et al. Common feature learning for brain tumor MRI synthesis by context-aware generative adversarial network. Med Image Anal. 2022;79:102472. doi: 10.1016/j.media.2022.102472

[41] Cirillo MD, Abramian D, Eklund A. Vox2Vox: 3D-GAN for brain tumour segmentation. In: Lecture Notes in Computer Science. Vol. 12658. Cham, Switzerland: Springer Nature; 2021. p. 274-284. doi: 10.1007/978-3-030-72084-1_25

[42] Aljohani A, Alharbe N. Generating synthetic images for healthcare with novel deep Pix2Pix GAN. Electronics. 2022; 11(21):3470. doi: 10.3390/electronics11213470

[43] Mehmood M, Alshammari N, Alanazi SA, et al. Improved colorization and classification of intracranial tumor expanse in MRI images via hybrid scheme of Pix2Pix-cGANs and NASNet-large. J King Saud Univ Comp Inf Sci. 2022; 34(7):4358-4374. doi: 10.1016/j.jksuci.2022.05.015

[44] Narayanan SJ, Anil AS, Ashtikar C, Chunduri S, Saman S. Automated brain tumor segmentation using GAN augmentation and optimized U-Net. In: Lecture Notes in Networks and Systems. Vol. 519. Singapore:Springer Nature; 2023. p. 635-646. doi: 10.1007/978-981-19-5191-6_51

[45] Li Q, Yu Z, Wang Y, Zheng H. TumorGAN: A multi-modal data augmentation framework for brain tumor segmentation. Sensors (Basel). 2020; 20(15):4203. doi: 10.3390/s20154203

[46] Mukherjee D, Saha P, Kaplun D, Sinitca A, Sarkar R. Brain tumor image generation using an aggregation of GAN models with style transfer. Sci Rep. 2022;12:9380. doi: 10.1038/s41598-022-12646-y

[47] Azni HM, Afsharchi M, Allahverdi A. Improving brain tumor segmentation performance using CycleGAN-based feature extraction. Multimed Tools Appl. 2023; 82(12):18039-18058. doi: 10.1007/s11042-022-14174-3

[48] Kim S, Kim B, Park H. Synthesis of brain tumor MR images for learning data augmentation. Med Phys. 2020; 48(5):2185-2198. doi: 10.1002/mp.14701

[49] Saxena D, Cao J. Generative adversarial networks (GANs). ACM Comput Surv. 2021; 54(3):1-38. doi: 10.1145/3446374

[50] Akbar MU, Larsson M, Blystad I, Eklund A. Brain tumor segmentation using synthetic MR images-a comparison of GANs and diffusion models. Sci Data. 2024;11:159. doi: 10.1038/s41597-024-03073-x

[51] Dhariwal P, Nichol A. Diffusion models beat GANs on image synthesis. Adv Neural Inf Process Syst. 2021;34:8780-8794.

[52] Onakpojeruo EP, Mustapha MT, Ozsahin DU, Ozsahin I. Enhanced MRI-based brain tumour classification with a novel Pix2pix generative adversarial network augmentation framework. Brain Commun. 2024; 6(6):fcae372. doi: 10.1093/braincomms/fcae372

[53] Brain Tumor Classification (MRI) Dataset. Kaggle; 2024. Available from: https://www.kaggle.com/datasets/sartajbhuvaji/brain-tumor-classification-mri [Last accessed on 2024 Mar 13].

[54] Cheng J. Brain Tumor Dataset. Figshare; 2017. Available from: https://figshare.com/articles/dataset/brain_tumor_dataset/1512427/5 [Last accessed on 2024 Mar 13].

[55] Yu J, Oh H, Lee Y, Yang J. Denoising diffusion model with adversarial learning for unsupervised anomaly detection on brain MRI images. Pattern Recognit Lett. 2024; 186:229-235. doi: 10.1016/j.patrec.2024.10.007