1.School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China
2.Department of Radiation Oncology, Guangdong Provincial People's Hospital (Guangdong Academy of Medical Sciences), Southern Medical University, Guangzhou 510080, China
3.Department of Gynecology, Sun Yat-sen University Cancer Center, South China State Key Laboratory of Oncology, Provincial-Ministry Collaborative Innovation Center for Medical Oncology, Guangzhou 510060, China
Objective To develop a TabMap image mapping and deep learning prediction framework that integrates medical prior knowledge to address the challenges of complex feature associations in tabular medical data and insufficient model interpretability in early ovarian cancer diagnosis. Methods Based on clinical semantics, 36 medical diagnostic features were partitioned into 4 spatially continuous regions, namely the routine blood test partition, biochemical indicators partition, tumor markers partition, and other (coagulation, inflammation, and other clinical variables) partition. The Gromov-Wasserstein optimal transport algorithm was then employed to solve the optimal coupling between feature space and pixel space, thus generating TabMap images that preserve the topological structures. A lightweight convolutional neural network incorporating SE attention mechanism and global average pooling was designed to handle sample imbalance using class-weighted loss and weighted sampling strategies. Finally, class activation mapping (CAM) technique was utilized to visualize the model's decision-making process for interpretability analysis. Results Experiments on a real ovarian cancer dataset demonstrated that the proposed method achieved a test accuracy of 91.82% with a precision of 89.96%, recall of 89.45%, F1-score of 0.8970 and balanced accuracy of 88.51%, representing accuracy improvements of 8.26%-14.93% over traditional machine learning methods and 2.28%-13.60% over standard deep learning models. Interpretability analysis showed that the model exhibits pronounced activation patterns in the tumor-marker region as well as in coagulation-, inflammation-, and age-related zones, consistent with established clinical understanding. The feature-importance ranking further underscored the pivotal roles of CA125, HE4, and other key biomarkers in guiding the model's decisions. Conclusion The proposed medical prior-guided TabMap method effectively integrates domain knowledge with data-driven learning, which enhances both its prediction performance and clinical interpretability. This strategy provides a novel approach for deep learning modeling of tabular medical data with good clinical potentials.
DexterJM, BrubakerLW, BitlerBG, et al. Ovarian cancer think tank: an overview of the current status of ovarian cancer screening and recommendations for future directions[J]. Gynecol Oncol Rep, 2024, 53: 101376. doi:10.1016/j.gore.2024.101376
[5]
Lawson-MichodKA, WattMH, GrieshoberL, et al. Pathways to ovarian cancer diagnosis: a qualitative study[J]. BMC Womens Health, 2022, 22(1): 430. doi:10.1186/s12905-022-02016-1
[6]
MatarasoSJ, EspinosaCA, SeongD, et al. A machine learning approach to leveraging electronic health records for enhanced omics analysis[J]. Nat Mach Intell, 2025, 7(2): 293-306. doi:10.1038/s42256-024-00974-9
[7]
KronesF, MarikkarU, ParsonsG, et al. Review of multimodal machine learning approaches in healthcare[J]. Inf Fusion, 2025, 114: 102690. doi:10.1016/j.inffus.2024.102690
[8]
HeX, BaiXH, ChenH, et al. Machine learning models in evaluating the malignancy risk of ovarian tumors: a comparative study[J]. J Ovarian Res, 2024, 17(1): 219. doi:10.1186/s13048-024-01544-8
[9]
GuidoR, FerrisiS, LofaroD, et al. An overview on the advancements of support vector machine models in healthcare applications: a review[J]. Information, 2024, 15(4): 235. doi:10.3390/info15040235
[10]
BarreñadaL, DhimanP, TimmermanD, et al. Understanding overfitting in random forest for probability estimation: a visualization and simulation study[J]. Diagn Progn Res, 2024, 8(1): 14. doi:10.1186/s41512-024-00177-1
[11]
MadakkatelI, LumsdenAL, MulugetaA, et al. Large-scale analysis to identify risk factors for ovarian cancer[J]. Int J Gynecol Cancer, 2025, 35(8): 101844. doi:10.1136/ijgc-2024-005424
[12]
ZhuYT, BrettinT, XiaFF, et al. Converting tabular data into images for deep learning with convolutional neural networks[J]. Sci Rep, 2021, 11(1): 11325. doi:10.1038/s41598-021-90923-y
[13]
SharmaA, VansE, ShigemizuD, et al. DeepInsight: a methodology to transform a non-image data to an image for convolution neural network architecture[J]. Sci Rep, 2019, 9(1): 11399. doi:10.1038/s41598-019-47765-6
[14]
BazgirO, ZhangRB, DhrubaSR, et al. Representation of features as images with neighborhood dependencies for compatibility with convolutional neural networks[J]. Nat Commun, 2020, 11(1): 4391. doi:10.1038/s41467-020-18197-y
[15]
YanR, IslamMT, XingL. Interpretable discovery of patterns in tabular data via spatially semantic topographic maps[J]. Nat Biomed Eng, 2025, 9(4): 471-82. doi:10.1038/s41551-024-01268-6
[16]
ZhangTS, PangAB, LyuJG, et al. Application of nonlinear models combined with conventional laboratory indicators for the diagnosis and differential diagnosis of ovarian cancer[J]. J Clin Med, 2023, 12(3): 844. doi:10.3390/jcm12030844
[17]
叶应妩. 全国临床检验操作规程[M]. 3版. 南京: 东南大学出版社, 2006.
[18]
BurtisCA, AshwoodER, BrunsDE. Tietz textbook of clinical chemistry and molecular diagnostics[M]. Elsevier, 2012. doi:10.1016/b978-1-4160-6164-9.00139-6
[19]
PeyréG, CuturiM, SolomonJM. Gromov-Wasserstein averaging of kernel and distance matrices[C]//International Conference on Machine Learning., 2016.
[20]
XiongLL, YiC, XiongQL, et al. SEA-NET: medical image segmentation network based on spiral squeeze-and-excitation and attention modules[J]. BMC Med Imaging, 2024, 24(1): 17. doi:10.1186/s12880-024-01194-8
[21]
SongZY, ShiZL, YanXM, et al. An improved weighted cross-entropy-based convolutional neural network for auxiliary diagnosis of pneumonia[J]. Electronics, 2024, 13(15): 2929. doi:10.3390/electronics13152929
ChenTQ, GuestrinC. XGBoost: a scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco California USA. ACM, 2016: 785-794. doi:10.1145/2939672.2939785
[24]
Abu-DoushI, AhmedB, AwadallahMA, et al. Enhancing multilayer perceptron neural network using archive-based Harris Hawks optimizer to predict gold prices[J]. J King Saud Univ Comput Inf Sci, 2023, 35(5): 101557. doi:10.1016/j.jksuci.2023.101557
HeKM, ZhangXY, RenSQ, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). June 27-30, 2016, Las Vegas, NV, USA. IEEE, 2016: 770-8. doi:10.1109/cvpr.2016.90
[27]
MomenimovahedZ, MazidimoradiA, AllahqoliL, et al. The role of CA-125 in the management of ovarian cancer: a systematic review[J]. Cancer Rep, 2025, 8(3): e70142. doi:10.1002/cnr2.70142
[28]
FarzanehF, SalimnezhadM, HosseiniMS, et al. D-dimer, fibrinogen and tumor marker levels in patients with benign and malignant ovarian tumors[J]. Asian Pac J Cancer Prev, 2023, 24(12): 4263-8. doi:10.31557/apjcp.2023.24.12.4263
[29]
ZhangLY, GuoLL, WangHY, et al. Knowledge graph and bibliometric analysis of inflammatory indicators in ovarian cancer[J]. Front Oncol, 2025, 15: 1533537. doi:10.3389/fonc.2025.1533537