Adaptive Threshold-Enhanced Deep Segmentation of Acute Intracranial Hemorrhage and its Subtypes in Brain CT Images

R. Suganthi; Pratibha C. Kaladeep Yalagi; Rini Chowdhury; Prashant Kumar; D. Sharmila; Kunchanapalli Rama  Krishna

doi:10.35882/jeeemi.v7i4.1048

R. Suganthi Department of Electronics and Communication Engineering, Panimalar Engineering College, Bangalore Trunk Road, Varadharajapuram, Poonamallee, Chennai, Tamil Nadu, India https://orcid.org/0000-0002-7045-5321
Pratibha C. Kaladeep Yalagi Department of Computer Science and Engineering, Walchand Institute of Technology, Solapur, Maharashtra, India https://orcid.org/0000-0002-3472-9671
Rini Chowdhury Department of Information Technology Project Circle, Bharat Sanchar Nigam Limited, Saltlake Telephone Exchange, Block DE, Lalkuthi, West Bengal, Kolkata, India https://orcid.org/0009-0009-6592-4216
Prashant Kumar Department of Information Technology Project Circle, Bharat Sanchar Nigam Limited, Saltlake Telephone Exchange, Block DE, Lalkuthi, West Bengal, Kolkata, India https://orcid.org/0009-0007-3378-615X
D. Sharmila Department of Electronics and Communication Engineering, Sri Krishna College of Engineering and Technology, BK Pudur, Kuniyamuthur, Tamil Nadu, India https://orcid.org/0000-0003-1381-636X
Kunchanapalli Rama Krishna Department of Computer Science & Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Guntur District, Andhra Pradesh, India https://orcid.org/0000-0001-9393-7713

DOI: https://doi.org/10.35882/jeeemi.v7i4.1048

Abstract

Accurate segmentation of acute intracranial haemorrhage (ICH) in brain computed tomography (CT) scans is crucial for timely diagnosis and effective treatment planning. While the RSNA Intracranial Hemorrhage Detection dataset provides a substantial amount of labeled CT data, most prior research has focused on slice-level classification rather than precise pixel-level segmentation. To address this limitation, a novel segmentation pipeline is proposed that combines a 2.5D U-Net architecture with a dynamic adaptive thresholding technique for enhanced delineation of hemorrhagic lesions and their subtypes. The 2.5D U-Net model leverages spatial continuity across adjacent slices to generate initial lesion probability maps, which are subsequently refined using an adaptive thresholding method that adjusts based on local pixel intensity histograms and edge gradients. Unlike fixed global thresholding approaches such as Otsu’s method, the proposed technique dynamically varies thresholds, enabling more accurate differentiation between hemorrhagic tissue and surrounding brain structures, especially in challenging cases with diffuse or overlapping boundaries. The model was evaluated on carefully selected subsets of the RSNA dataset, achieving a mean Dice similarity coefficient of 0.82 across all ICH subtypes. Compared to standard U-Net and DeepLabV3+ architectures, the hybrid approach demonstrated superior accuracy, boundary precision, and fewer false positives. Visual analysis confirmed more precise lesion delineation and better correspondence with manual annotations, particularly in low-contrast or complex anatomical regions. This integrated approach proves effective for robust segmentation in clinical environments. It holds promise for deployment in computer-aided diagnosis systems, providing radiologists and neurosurgeons with a reliable tool for comprehensive ICH assessment and enhanced decision-making during emergency care

Downloads

Download data is not yet available.

References

Tian, J., Yin, M., & Jiang, J. (2024). Fault self-healing: A biological immune heuristic reinforcement learning method with root cause reasoning in industrial manufacturing process. Engineering Applications of Artificial Intelligence, 133, 108553.

Danilov, G., Kotik, K., Negreeva, A., Tsukanova, T., Shifrin, M., Zakharova, N., ... & Potapov, A. (2020). Classification of intracranial hemorrhage subtypes using deep learning on CT scans. In The importance of health informatics in public health during a pandemic (pp. 370-373). IOS press.

Zimmerman, R. A., & Bilaniuk, L. T. (1980). Computed tomography of acute intratumoral hemorrhage. Radiology, 135(2), 355-359.

Huisman, T. A. (2005). Intracranial hemorrhage: ultrasound, CT and MRI findings. European radiology, 15(3), 434-440.

Yeo, M., Tahayori, B., Kok, H. K., Maingard, J., Kutaiba, N., Russell, J., ... & Asadi, H. (2021). Review of deep learning algorithms for the automatic detection of intracranial hemorrhages on computed tomography head imaging. Journal of neurointerventional surgery, 13(4), 369-378.

Wang, X., Shen, T., Yang, S., Lan, J., Xu, Y., Wang, M., ... & Han, X. (2021). A deep learning algorithm for automatic detection and classification of acute intracranial hemorrhages in head CT scans. NeuroImage: Clinical, 32, 102785.

Burduja, M., Ionescu, R. T., & Verga, N. (2020). Accurate and efficient intracranial hemorrhage detection and subtype classification in 3D CT scans with convolutional and long short-term memory neural networks. Sensors, 20(19), 5611.

Ye, H., Gao, F., Yin, Y., Guo, D., Zhao, P., Lu, Y., ... & Xia, J. (2019). Precise diagnosis of intracranial hemorrhage and subtypes using a three-dimensional joint convolutional and recurrent neural network. European radiology, 29, 6191-6201.

Phaphuangwittayakul, A., Guo, Y., Ying, F., Dawod, A. Y., Angkurawaranon, S., & Angkurawaranon, C. (2022). An optimal deep learning framework for multi-type hemorrhagic lesions detection and quantification in head CT images for traumatic brain injury. Applied Intelligence, 1-19.

Hssayeni, M., Croock, M., Salman, A., Al-khafaji, H., Yahya, Z., & Ghoraani, B. (2020). Computed tomography images for intracranial hemorrhage detection and segmentation. Intracranial hemorrhage segmentation using a deep convolutional model. Data, 5(1), 14.

Lee, J. Y., Kim, J. S., Kim, T. Y., & Kim, Y. S. (2020). Detection and classification of intracranial haemorrhage on CT images using a novel deep-learning algorithm. Scientific reports, 10(1), 20546.

Wu, Y., Supanich, M. P., & Deng, J. (2021). Ensembled deep neural network for intracranial hemorrhage detection and subtype classification on noncontrast CT images. Journal of Artificial Intelligence for Medical Sciences, 2(1), 12-20.

Inkeaw, P., Angkurawaranon, S., Khumrin, P., Inmutto, N., Traisathit, P., Chaijaruwanich, J., ... & Chitapanarux, I. (2022). Automatic hemorrhage segmentation on head CT scan for traumatic brain injury using 3D deep learning model. Computers in Biology and Medicine, 146, 105530.

Kumaravel, P., Mohan, S., Arivudaiyanambi, J., Shajil, N., & Venkatakrishnan, H. N. (2021). A simplified framework for the detection of intracranial hemorrhage in CT brain images using deep learning. Current Medical Imaging Reviews, 17(10), 1226-1236.

Kuo, W., Hӓne, C., Mukherjee, P., Malik, J., & Yuh, E. L. (2019). Expert-level detection of acute intracranial hemorrhage on head computed tomography using deep learning. Proceedings of the National Academy of Sciences, 116(45), 22737-22745.

Chen, Y. R., Chen, C. C., Kuo, C. F., & Lin, C. H. (2024). An efficient deep neural network for automatic classification of acute intracranial hemorrhages in brain CT scans. Computers in Biology and Medicine, 176, 108587.

Lee, H., Yune, S., Mansouri, M., Kim, M., Tajmir, S. H., Guerrier, C. E., ... & Do, S. (2019). An explainable deep-learning algorithm for the detection of acute intracranial haemorrhage from small datasets. Nature biomedical engineering, 3(3), 173-182.

Suganyadevi, S., Pershiya, A. S., Balasamy, K., et al. “Deep learning based alzheimer disease diagnosis: A comprehensive review”. SN Computer Science, Vol.5 no.4, pp.391, 2024, doi:10.1007/s42979-024-02743-2

Balasamy, K., Krishnaraj, N., & Vijayalakshmi, K. “An adaptive neuro-fuzzy based region selection and authenticating medical image through watermarking for secure communication”, Wireless Personal Communications, Vol.122, no.3, pp. 2817–2837, 2021, doi:10.1007/s11277-021-09031-9 .

Suganyadevi, S., & Seethalakshmi, V. “CVD-HNet: Classifying Pneumonia and COVID-19 in Chest X-ray Images Using Deep Network”. Wireless Personal Communications, Vol.126, no. 4, pp.3279–3303, 2022, doi: 10.1007/s11277-022-09864-y

Balasamy, K., & Suganyadevi, S. “Multi-dimensional fuzzy based diabetic retinopathy detection in retinal images through deep CNN method”. Multimedia Tools and Applications, Vol 83, no. 5, pp.1–23. 2024, doi: 10.1007/s11042-024-19798-1

Shamia, D., Balasamy, K., and Suganyadevi, S. “A secure framework for medical image by integrating watermarking and encryption through fuzzy based roi selection”, Journal of Intelligent & Fuzzy systems, 2023, Vol. 44, no.5, pp.7449-7457, doi: 10.3233/JIFS-222618.

Balasamy, K., Seethalakshmi, V. & Suganyadevi, S. Medical Image Analysis Through Deep Learning Techniques: A Comprehensive Survey. Wireless Pers Commun 137, 1685–1714 (2024). https://doi.org/10.1007/s11277-024-11428-1.

Suganyadevi, S., Seethalakshmi, V. Deep recurrent learning based qualified sequence segment analytical model (QS2AM) for infectious disease detection using CT images. Evolving Systems 15, 505–521 (2024). https://doi.org/10.1007/s12530-023-09554-5

Mahadzir, N. (2021). Sentiment analysis of code-mixed text: A review. Turkish Journal of Computer and Mathematics Education (TURCOMAT), 12, 2469–2478. https://doi.org/ 10.17762/turcomat.v12i3.1239, 04/1.

Ma, J., Chen, J., Ng, M., Huang, R., Li, Y., Li, C., Yang, X., Martel, A.L., 2021. Loss odyssey in medical image segmentation. Med. Image Anal. 71, 102035.

Ma, Y., Ren, F., Li, W., Yu, N., Zhang, D., Li, Y., Ke, M.J.B.S.P.C., 2023. IHA-Net: An automatic segmentation framework for computer-tomography of tiny intracerebral hemorrhage based on improved attention U-net. Biomed. Signal Process. Control 80, 104320.

Ramakrishnan, S., Gopalakrishnan, T., and Balasamy, K. (2011). A wavelet based hybrid SVD algorithm for digital image watermarking. Signal & Image Processing An International Journal 2(3):157–174.

Melinosky, C., Kincaid, H., Claassen, J., Parikh, G., Badjatia, N., Morris, N.A., 2021. The modified fisher scale lacks interrater reliability. Neurocrit. Care 35, 72–78.

Milletari, F., Navab, N., Ahmadi, S.A., 2016. V-Net: fully convolutional neural networks for volumetric medical image segmentation. In: Proceedings of the 2016 Fourth International Conference on 3d Vision (3dv), pp. 565–571.

Neifert, S.N., Chapman, E.K., Martini, M.L., Shuman, W.H., Schupper, A.J., Oermann, E. K., Mocco, J., Macdonald, R.L., 2021. Aneurysmal subarachnoid hemorrhage: the last decade. Transl. Stroke Res. 12, 428–446.

Neves, G., Warman, P.I., Warman, A., Warman, R., Bueso, T., Vadhan, J.D., Windisch, T., 2023. External validation of an artificial intelligence device for intracranial hemorrhage detection. World Neurosurg. 173, e800–e807.

Nijiati, M., Tuersun, A., Zhang, Y., Yuan, Q., Gong, P., Abulizi, A., Tuoheti, A., Abulaiti, A., Zou, X., 2022. A symmetric prior knowledge based deep learning model for intracerebral hemorrhage lesion segmentation. Front. Physiol. 13, 977427.

Nishi, T., Yamashiro, S., Okumura, S., Takei, M., Tachibana, A., Akahori, S., Kaji, M., Uekawa, K., Amadatsu, T., 2021. Artificial intelligence trained by deep learning can improve computed tomography diagnosis of nontraumatic subarachnoid hemorrhage by nonspecialists. Neurol. Med. Chir. 61, 652–660.

Petridis, A.K., Kamp, M.A., Cornelius, J.F., Beez, T., Beseoglu, K., Turowski, B., Steiger, H.J., 2017. Aneurysmal subarachnoid hemorrhage. Dtsch. Arztebl. Int. 114, 226–236.

Platz, J., Güresir, E., Wagner, M., Seifert, V., Konczalla, J., 2017. Increased risk of delayed cerebral ischemia in subarachnoid hemorrhage patients with additional intracerebral hematoma. J. Neurosurg. 126, 504–510.

Rava, R.A., Seymour, S.E., LaQue, M.E., Peterson, B.A., Snyder, K.V., Mokin, M., Waqas, M., Hoi, Y., Davies, J.M., Levy, E.I., Siddiqui, A.H., Ionita, C.N., 2021. Assessment of an artificial intelligence algorithm for detection of intracranial hemorrhage. World Neurosurg. 150, e209–e217.

Ronneberger, O., Fischer, P., Brox, T., 2015. U-Net: Convolutional Networks for Biomedical Image Segmentation. Springer International Publishing, Cham, pp. 234–241.

Krizhevsky, A., Sutskever, I., Hinton, G.E., 2012. Imagenet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems 25, 1097–1105.

Gopalakrishnan, T., Ramakrishnan, S., Balasamy, K., and Murugavel A. S. M.(2011). Semi fragile watermarking using Gaussian mixture model for malicious image attacks. In: 2011 World Congress on Information and Communication Technologies, pp. 120–125.

Labovitz, D.L., Halim, A., Boden-Albala, B., Hauser, W., Sacco, R., 2005. The incidence of deep and lobar intracerebral hemorrhage in whites, blacks, and hispanics. Neurology 65, 518–522.