Leveraging persistent homology for liver tumour classification Conference Paper
| Authors: | Ali, D. A.; Peoples, J. J.; Mojtahedi, R.; Kobayashi, K. S.; Jarnagin, W. R.; Do, R. K. G.; Simpson, A. L. |
| Title: | Leveraging persistent homology for liver tumour classification |
| Conference Title: | Medical Imaging 2025: Computer-Aided Diagnosis |
| Abstract: | Distinguishing between intrahepatic cholangiocarcinoma (ICC) and hepatocellular carcinoma (HCC) in imaging is a difficult task for a radiologist. We endeavoured to develop reliable models to automatically classify these tumour types. In this study, we propose to use persistent homology (PH), from the field of topological data analysis (TDA) to build topological shapes from computed tomography (CT) scans of the liver. PH is used to extract topological and geometrical summaries such as the number of persistent connected components and loops from CT scans in the form of persistent barcodes. Topological and geometrical features encapsulated by barcodes are stable to small perturbations to the input data such as variations in scan protocols. Extracted topological features are used as input to various classifiers achieving 97.56% F1-score with 97.5% accuracy. Similar experiment is performed with radiomics features achieving comparable metrics with TDA being marginally higher. Furthermore, pre-trained convolutional neural networks (CNNs) are also explored for benchmarking where comparable results are achieved. Our results suggest that TDA is an effective feature engineering approach for CT scans because, unlike traditional approaches, it uses persistent homology to capture the spatial distribution of texture and pixels in CT scans. Although CNNs achieve comparable results, TDA is preferable due to its interpretability and efficiency. Furthermore, TDA and radiomics features can complement each other whereby a miss-classified scan by radiomics is correctly classified by TDA and vice versa. © 2025 SPIE. |
| Keywords: | discriminant analysis; liver cancer; cholangiocarcinoma; tumor classification; image segmentation; liver tumors; computed tomography scan; convolutional neural network; liver cancers; taxonomies; classifieds; labeled data; topological data analysis; persistent homology; topological features |
| Journal Title | Progress in Biomedical Optics and Imaging - Proceedings of SPIE |
| Volume: | 13407 |
| Conference Dates: | 2025 Feb 17-20 |
| Conference Location: | San Diego, CA |
| ISBN: | 1605-7422 |
| Publisher: | SPIE |
| Date Published: | 2025-01-01 |
| Start Page: | 134071Q |
| Language: | English |
| DOI: | 10.1117/12.3045640 |
| PROVIDER: | scopus |
| DOI/URL: | |
| Notes: | Conference paper (ISBN: 9781510685925) -- Source: Scopus |
Altmetric
Citation Impact
BMJ Impact Analytics
Related MSK Work