Impact of Wavelet Pre-processing on Deep Learning Architectures for Renal Cancer Diagnosis

Vinícius Meireles Pereira Santos; Ana Claudia Patrocinio; Pedro Moises de Sousa

doi:10.22456/2175-2745.150824

Authors

Vinícius Meireles Pereira Santos Universidade Federal de Viçosa (UFV) https://orcid.org/0009-0006-1067-342X
Ana Claudia Patrocinio Universidade Federal de Uberlândia (UFU) https://orcid.org/0000-0001-9376-7689
Pedro Moises de Sousa Universidade Federal de Viçosa (UFV) https://orcid.org/0000-0003-4563-0033

DOI:

https://doi.org/10.22456/2175-2745.150824

Keywords:

deep learning, renal cancer diagnosis, computed tomography, vision transformer, wavelet transform, explainable AI, computer-aided diagnosis

Abstract

Renal cancer is a lethal neoplasm where early detection is critical yet challenging. To define an optimal diagnostic pipeline, this study conducts a comparative analysis on a public tomography dataset (CT KIDNEY DATASET), evaluating the impact of a wavelet filter. Six deep learning architectures were tested, spanning classic (VGG16), modern (ResNet50), efficient (MobileNetV2), and attention-based (Vision Transformer - ViT) paradigms. Our results reveal a complex, architecture-dependent impact: wavelet pre-processing enables a ResNet50 to achieve a test accuracy of 100.00%. However, we demonstrate that a more efficient architecture, MobileNetV2, reaches the same state-of-the-art performance without this pre-processing step. These findings highlight that the optimal diagnostic pipeline is a balance between architectural efficiency and pre-processing complexity, positioning MobileNetV2 as the most pragmatic and effective solution.

Downloads

Download data is not yet available.

References

[1] INSTITUTO NACIONAL DE CÂNCER (INCA). Atlas de Mortalidade por Câncer. 2023. Available: ⟨https://www.inca.gov.br/MortalidadeWeb/pages/Modelo03/consultar.xhtml⟩. Accessed on: August 02 2025.

[2] ZOLIN, B. Câncer renal: doença é silenciosa, mas pode ser prevenida. Portal Drauzio Varella, 2023. Available: ⟨https://drauziovarella.uol.com.br/oncologia/cancer-renal-doenca-e-silenciosa-mas-pode-ser-prevenida/⟩. Accessed on: August 02 2025.

[3] INSTITUTO NACIONAL DE CÂNCER (INCA). Estimativa 2023: Incidência de Câncer no Brasil. 2023. Available: ⟨https://www.inca.gov.br/publicacoes/livros/estimativa-2023-incidencia-de-cancer-no-brasil⟩. Accessed on: August 02 2025.

[4] ALZU’BI, D. et al. Kidney tumor detection and classification based on deep learning approaches: a new dataset in CT scans. Journal of Healthcare Engineering, v. 2022, Art. ID 3861161, 2022.

[5] RAJKUMAR, K. et al. Kidney cancer detection using deep learning models. In: 7TH INTERNATIONAL CONFERENCE ON TRENDS IN ELECTRONICS AND INFORMATICS (ICOEI). [S.l.]: IEEE, 2023. p. 1197–1203.

[6] HADJIYSKI, N. Deep learning neural network based approach. In: INTERNATIONAL CONFERENCE ON E-HEALTH AND BIOENGINEERING (EHB). [S.l.]: IEEE, 2020. p. 1–4.

[7] UHM, K. H. et al. Deep learning for end-to-end kidney cancer diagnosis on multi-phase abdominal computed tomography. NPJ Precision Oncology, v. 5, n. 1, p. 54, 2021.

[8] SANTOS, V. M. P. et al. Renal cancer classification in tomographic images using convolutional neural networks to assist early diagnosis. Revista de Informática Teórica e Aplicada, v. 32, n. 1, p. 61–66, 2025.

[9] SINGH, A. R. et al. An enhanced denoising system for mammogram images using deep transformer model with fusion of local and global features. Scientific Reports, v. 15, n. 1, p. 6562, 2025.

[10] RASTOGI, D. et al. Deep learning-integrated MRI brain tumor analysis: feature extraction, segmentation, and survival prediction using replicator and volumetric networks. Scientific Reports, v. 15, n. 1, p. 1437, 2025.

[11] LI, J. et al. Transforming medical imaging with transformers? A comparative review of key properties, current progresses, and future perspectives. Medical Image Analysis, v. 85, p. 102762, 2023.

[12] OKOLO, G. I. et al. IEViT: an enhanced vision transformer architecture for chest x-ray image classification. Computer Methods and Programs in Biomedicine, v. 226, p. 107141, 2022.

[13] AWASTHI, V. et al. ViT-HHO: optimized vision transformer for diabetic retinopathy detection using harris hawk optimization. MethodsX, v. 13, p. 103018, 2024.

[14] SHEIKHY, A. et al. State of the art review of AI in renal imaging. Abdominal Radiology, 2025.

[15] ISLAM, M. N. et al. Vision transformer and explainable transfer learning models for auto detection of kidney cyst, stone and tumor from CT-radiography. Scientific Reports, v. 12, n. 1, Art. ID 11352, 2022.

[16] JAPKOWICZ, N.; STEPHEN, S. The class imbalance problem: a systematic study. Intelligent Data Analysis, v. 6, n. 5, p. 429–449, 2002.

[17] HE, K. et al. Deep residual learning for image recognition. In: IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR). Las Vegas: IEEE, 2016. p. 770–778.

[18] KRIZHEVSKY, A.; SUTSKEVER, I.; HINTON, G. E. Imagenet classification with deep convolutional neural networks. In: ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS (NIPS). 2012. p. 1097–1105.

[19] HOWARD, A. G. et al. Mobilenets: efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861, 2017.

[20] SIMONYAN, K.; ZISSERMAN, A. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014.

[21] TAN, M.; LE, Q. V. Efficientnet: rethinking model scaling for convolutional neural networks. In: 36TH INTERNATIONAL CONFERENCE ON MACHINE LEARNING (ICML). [S.l.]: ACM, 2019. p. 6105–6114.

[22] DOSOVITSKIY, A. et al. An image is worth 16x16 words: transformers for image recognition at scale. In: INTERNATIONAL CONFERENCE ON LEARNING REPRESENTATIONS (ICLR). 2021.

[23] SOKOLOVA, M.; LAPALME, G. A systematic analysis of performance measures for classification tasks. Information Processing & Management, v. 45, n. 4, p. 427–437, 2009.

[24] SELVARAJU, R. R. et al. Grad-cam: visual explanations from deep networks via gradient-based localization. In: IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV). [S.l.]: IEEE, 2017. p. 618–626.

[25] ABDELTAWAB, H. et al. A pyramidal deep learning pipeline for kidney whole-slide histology images classification. Scientific Reports, v. 11, n. 1, p. 20189, 2021.

[26] QADIR, A. M.; ABD, D. F. Kidney diseases classification using hybrid transfer-learning densenet201-based and random forest classifier. KJAR, v. 7, n. 2, p. 131–144, 2023.

[27] CHANCHAL, A. K. et al. A novel dataset and efficient deep learning framework for automated grading of renal cell carcinoma from kidney histopathology images. Scientific Reports, v. 13, n. 1, p. 5728, 2023.

[28] CHEFER, H.; GUR, S.; WOLF, L. Generic attention-model explainability for interpreting bi-modal and encoder-decoder transformers. In: IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV). Montreal, QC, Canada, 2021. p. 387–396.

[29] ISHWARAN, H.; O’BRIEN, R. Commentary: The problem of class imbalance in biomedical data. The Journal of Thoracic and Cardiovascular Surgery, v. 161, n. 6, p. 1940-1941, 2021.

[30] CROSS, J. L.; CHOMA, M. A.; ONOFREY, J. A. Bias in medical AI: Implications for clinical decision-making. PLOS Digital Health, v. 3, n. 11, p. e0000651, 2024.

[31] GONZALEZ, R. C.; WOODS, R. E. Digital Image Processing. 3rd ed. Upper Saddle River, NJ: Prentice-Hall, Inc., 2006.

[32] WANG, Z. et al. Image quality assessment: from error visibility to structural similarity. IEEE Transactions on Image Processing, v. 13, n. 4, p. 600-612, 2004.