Patient Case Similarity using AI with Black Fungus and Cancer Treatment

Abstract

This paper presents an Artificial Intelligence (AI)-based Patient Case Similarity System designed to assist doctors in analyzing and comparing cases of black fungus (mucormycosis) and cancer patients. The system uses machine learning algorithms to identify similar historical patient cases based on symptoms, laboratory reports, medical imaging features, comorbidities, and treatment responses. By applying similarity measures and classification techniques such as Random Forest and Cosine Similarity, the system helps clinicians make faster and evidence-based treatment decisions. The proposed model improves diagnostic support, enhances personalized treatment planning, and reduces clinical uncertainty, especially in high-risk conditions like mucormycosis and oncology cases.

Introduction

Healthcare systems are undergoing rapid digital transformation due to the increasing availability of Electronic Health Records (EHRs), imaging data, laboratory reports, and genomic information. This data explosion creates opportunities for Artificial Intelligence (AI) and Machine Learning (ML) to enhance diagnosis, prognosis, and personalized treatment planning, particularly for complex diseases such as cancer and mucormycosis (black fungus).

Mucormycosis is a rare but aggressive fungal infection that primarily affects immunocompromised individuals, including diabetic and cancer patients. During the COVID-19 pandemic, cases surged significantly, especially in India. Early detection is critical because delayed diagnosis leads to severe complications and high mortality.

Cancer, a leading global cause of death, requires highly personalized treatment strategies based on tumor stage, genetics, histopathology, comorbidities, and previous treatment responses. Manually analyzing large volumes of patient records to determine optimal therapy is time-consuming and limited by human capacity.

AI-driven Patient Case Similarity Models offer a promising solution. These systems represent each patient as a feature vector incorporating demographic, clinical, laboratory, imaging, and treatment data. Similarity metrics such as Cosine Similarity, Euclidean Distance, K-Nearest Neighbors (KNN), and advanced ML models (Random Forest, SVM, Deep Learning) are used to identify historical cases most similar to a new patient. This enables clinicians to learn from previous outcomes and support evidence-based, personalized treatment decisions.

The literature shows strong progress in AI applications for cancer diagnosis, imaging analysis (especially CNN-based models), and survival prediction. However, similarity-based retrieval systems are less explored, particularly for mucormycosis. Existing research often focuses on classification accuracy rather than multi-disease case comparison and retrieval. Additionally, integration of structured data, imaging features, and interpretable models remains limited.

To address these gaps, the proposed research introduces a multi-disease AI-driven Patient Case Similarity System for both cancer and mucormycosis. The system architecture includes:

1. Data acquisition from EHRs

2. Data preprocessing and feature engineering

3. Patient vector representation (embeddings)

4. Similarity computation engine

5. Treatment recommendation module

The methodology involves cleaning and encoding structured data, extracting imaging features using CNNs, generating patient embeddings, applying similarity metrics (primarily cosine similarity), retrieving top similar cases, and predicting treatment success rates based on historical outcomes.

By combining machine learning, deep learning embeddings, and similarity-based retrieval, the proposed system aims to reduce diagnostic delays, improve treatment selection, enhance clinical decision support, and ultimately improve patient survival outcomes—particularly for high-risk conditions like cancer and mucormycosis.

Conclusion

This study proposes an AI-driven Patient Case Similarity System to support treatment decision-making for mucormycosis and cancer. The system integrates structured clinical data, imaging features, and similarity-based retrieval mechanisms to enhance personalized medicine [1], [2]. By leveraging nearest neighbor approaches [16] and deep learning-based feature representation [20], the proposed framework improves case comparison accuracy beyond traditional classification systems. Cancer analytics research has demonstrated the effectiveness of AI in diagnosis and prognosis [7], [9], while fungal infection studies highlight the importance of early clinical intervention [11], [12]. The integration of these approaches into a unified similarity-based architecture contributes to intelligent clinical decision support systems [24]. Future work will focus on incorporating explainable AI models and large-scale validation using multi-institutional datasets. However, certain limitations remain, including data availability constraints, class imbalance issues, and the need for large-scale validated datasets. Future work will focus on incorporating deep learning-based patient embeddings, explainable AI mechanisms for improved interpretability, and real-time deployment within hospital Electronic Health Record systems.

References

[1] E. J. Topol, “High-performance medicine: the convergence of human and artificial intelligence,” Nature Medicine, vol. 25, no. 1, pp. 44–56, 2019. [2] R. Miotto, F. Wang, S. Wang, X. Jiang, and J. T. Dudley, “Deep learning for healthcare: review, opportunities and challenges,” Briefings in Bioinformatics, vol. 19, no. 6, pp. 1236–1246, 2018. [3] A. Rajkomar, E. Oren, K. Chen et al., “Scalable and accurate deep learning for electronic health records,” npj Digital Medicine, vol. 1, no. 1, pp. 1–10, 2018. [4] P. Weng, Y. Chen, and C. Li, “Patient similarity networks for precision medicine,” Journal of Biomedical Informatics, vol. 85, pp. 67–77, 2018. [5] H. Avati, K. Jung, and N. Shah, “Improving palliative care with deep learning,” BMC Medical Informatics and Decision Making, vol. 18, no. 4, pp. 1–12, 2018. [6] S. Sohn, J. Kim, and L. Li, “Comparison of distance metrics for clinical similarity search,” IEEE Journal of Biomedical and Health Informatics, vol. 22, no. 4, pp. 1285–1293, 2018. [7] A. Esteva et al., “Dermatologist-level classification of skin cancer with deep neural networks,” Nature, vol. 542, pp. 115–118, 2017. [8] J. Litjens et al., “A survey on deep learning in medical image analysis,” Medical Image Analysis, vol. 42, pp. 60–88, 2017. [9] T. Albahri et al., “Systematic review of artificial intelligence techniques in cancer detection and diagnosis,” Computers in Biology and Medicine, vol. 122, 2020. [10] S. J. Pan and Q. Yang, “A survey on transfer learning,” IEEE Transactions on Knowledge and Data Engineering, vol. 22, no. 10, pp. 1345–1359, 2010. [11] R. Garg et al., “Coronavirus disease-associated mucormycosis: a systematic review,” Mycoses, vol. 64, no. 12, pp. 1452–1459, 2021 [12] A. Skiada et al., “Epidemiology and diagnosis of mucormycosis: an update,” Journal of Fungi, vol. 6, no. 4, 2020. [13] N. Petrikkos et al., “Epidemiology and clinical manifestations of mucormycosis,” Clinical Infectious Diseases, vol. 54, pp. S23–S34, 2012. [14] D. Chicco and G. Jurman, “The advantages of the Matthews correlation coefficient over F1 score,” BMC Genomics, vol. 21, no. 6, 2020. [15] L. Breiman, “Random forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, 2001. [16] T. Cover and P. Hart, “Nearest neighbor pattern classification,” IEEE Transactions on Information Theory, vol. 13, no. 1, pp. 21–27, 1967. [17] D. Dua and C. Graff, “UCI Machine Learning Repository,” University of California, Irvine, 2019. [18] O. Chapelle, B. Schölkopf, and A. Zien, Semi-Supervised Learning. MIT Press, 2006. [19] F. Wang et al., “Clinical information retrieval using patient similarity,” AMIA Annual Symposium Proceedings, pp. 1926–1935, 2015. [20] M. Shickel, P. Tighe, A. Bihorac, and P. Rashidi, “Deep EHR: a survey of recent advances,” IEEE Journal of Biomedical and Health Informatics, vol. 22, no. 5, pp. 1589–1604, 2018. [21] C. Shortliffe and J. Cimino, Biomedical Informatics: Computer Applications in Health Care and Biomedicine. Springer, 2014. [22] World Health Organization, “Cancer fact sheet,” WHO, 2023. [23] American Cancer Society, “Cancer treatment and survivorship statistics,” 2022. [24] R. K. Singh et al., “AI-based clinical decision support systems in healthcare,” IEEE Access, vol. 8, pp. 184–205, 2020 [25] J. Pearl, Causality: Models, Reasoning, and Inference, 2nd ed. Cambridge University Press, 2009.

Copyright

Copyright © 2026 Anand Pal, Alok Singh, Abhinav Pandey, Ms. Shekhar Srivastav. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.