American Journal of Multidisciplinary Research & Development (AJMRD)

Volume 07, Issue 02 (February - 2025), PP 73-79

ISSN: 2360-821X
www.ajmrd.com

Research Paper Open Access

The Role of AI in Predicting Cancer Recurrence and Patient

Survival Rates
1
Faroque Md Mohsin, 2Maherun Nesa, 3Shekh Mohammad Mostafa,
4
Aditta Das, 5Dr Zahidul Mostafa, 6Dr. Rahat Noor, 7Zeenat Un Nisa,
8
Sibat Asrafy, 9Sinigdha islam, 10Mahmudul Mannan
1
MBBS, MPH, MSS Medical officer Directorate General of Health
https://orcid.org/0000-0003-3432-4559
2
Registrar MBBS, MS National institute of Burn & Plastic Surgery
https://orcid.org/0000-0002-3299-7939
3
UHFPO, DGHS Cell-01711209243
https://orcid.org/0000-0002-9089-315
4
Hebei Medical University Shijiazhuang, CN MBBS
Orchid id -0009-0007-5650-2703
5
Assistant professor Cardiology Cox's Bazar medical college
https://orcid.org/0000-0002-8318-0929
6
Consultant radiology & Imaging Chevron Clinical Laboratory -Cox's Bazar,
7
Clin. Res. Project Coord.-RI NEPHROLOGY, The Hospital for Sick Children, Toronto, ON
https://orcid.org/0000-0003-1308
8
MBBS Holy Family Red Crescent Medical College and Hospital, Dhaka, Bangladesh Telemedicine Medical
Officer
9
MBBS, MPH
10
MSc Candidate, Institute of Medical Science (IMS), Temerty Faculty of Medicine, University of Toronto
https://orcid.org/0000-0002-9932-8772

ABSTRACT:- The paper analyzes machine learning algorithms, deep learning models, as well as AI based
genomic and imaging analysis, to talk over how AI systems predict cancer recurrence and survival estimates.
Three main features which make AI of valuable application to oncology practice includes early warning
technology, custom treatment building and real time healthcare record linkage for medical decision. Primary
issues in cancer recurrence and patient survival can be predicted by oncologist, where it has a significant impact
on treatment strategies and long term outcome. Such prediction methods utilized with statistical means and
doctor expertise in cancer care are, however, limited in terms of precision and flexibility. Artificial Intelligence
or AI is changing the healthcare operations by freeing up the capability to apply machine learning or deep
learning technology to create more precise cancer prognosis predictions. Vanguard prognostic cancer advances
will be based on recent developments in explainable AI systems, federated learning, and AI drug discovery
technologies. In order to properly set up the ethical criteria for AI implementation as well as augment its
positive effects for cancer treatment it will be necessary to initiate the research cooperation with oncologists and
policymakers. The research shows that AI technology is transforming oncology and improving results of
patients through better prediction systems.

Keywords: Artificial Intelligence (AI), Cancer Recurrence Prediction, Machine Learning in Oncology, Patient
Survival Rates, Medical Data Analytics

I. INTRODUCTION
Cancer ranks among the major contributors to worldwide mortality and disease burden since doctors
diagnose millions of new cancer cases each year. The primary issue in oncology centers on forecasting cancer
recurrence and survival outcomes of patients because these predictions guide treatment selection and ongoing

Multidisciplinary Journal www.ajmrd.com Page | 73

 The Role of AI in Predicting Cancer Recurrence and Patient Survival Rates

care approaches. The correct recognition of high-risk patients remains essential because cancer recurrence
appears after several months up to multiple years following initial therapy. The accuracy of existing prediction
strategies dependent on statistical methods, clinical evaluations, and biomarkers experiences limitations when
managing different aspects found in individual cancer patient cases. Healthcare professionals now benefit from
artificial intelligence (AI) as a strong analytical tool that optimizes studying extensive patient databases to better
forecast outcomes. AI implementation in oncology practice enables medical staff to base their choices on richer
information and design better treatments while monitoring patients continuously. The paper investigates how
artificial intelligence predicts cancer recurrence survival rates and its methods, benefits, and difficulties. The
paper examines both current and projected applications of AI technology in cancer care and shows their ability
to transform medical settings for the future.

II. OVERVIEW OF CANCER RECURRENCE AND SURVIVAL RATE CHALLENGES

Oncology professionals face the recurrence of cancer as an ongoing issue since it impacts both patient
survival potential and duration of survival directly. Cancer recurrence appears when disease cells reappear after
remission ends through therapy and may happen within months or years from the beginning of the treatment
period. Cancer recurrence takes three primary forms: local recurrence develops where the first cancer originated,
and regional recurrence affects the nearest lymph nodes or tissues. However, distant recurrence means the
cancer has metastasized to distant body locations. It is essential to predict future incidence of cancer recurrence
and survival patterns in order to develop efficient treatment plans and thus, reduce the chances of fatalities and
help the patients sustain a better life. However, due to the different biological aspects of the tumor, and different
individual patient variations, predicting future cancer recurrence is made difficult. The fact that different patient
and tumor characteristics (e.g., the malignant tissue type, the genetic mutation, treatment sensitivity, and the
general patient condition) make verified prognostic models difficult to establish. As their main drawback, the
existing methods for survival predictions have a tremendous limit. While these predictive models do not reflect
the current developments in treatment approaches, targeted therapies, and immunotherapy, almost none depend
on data from past times. However, the conventional survival analysis models do not take into account changes in
the patient's wellness that occur by virtue of such practices as a change of lifestyle and secondary medical
treatments (Susič et al., 2023; Noman et al., 2025).

Multidisciplinary Journal www.ajmrd.com Page | 74

 The Role of AI in Predicting Cancer Recurrence and Patient Survival Rates

Furthermore, the availability and quality of data pose significant hurdles. In order to develop precise
predictions, datasets containing sufficient patient data with diverse demographics, full treatment response, along
with long enough monitoring periods are critical. Among the problems of predictive model development in the
healthcare sector is that various providers deal with medical data separately, there are problems with privacy,
and the healthcare data between providers is reported inconsistently (Mazaki et al., 2021; Shimada et al., 2022).

III. UNDERSTANDING CANCER RECURRENCE AND SURVIVAL RATE

Definition of Cancer Recurrence and Its Types When cancer patients finish their initial treatment and
reach remission, their disease may regrow at the same site or another part of their body. Treatment
advancements in oncology do not eliminate the significant challenge of recurrence because such events typically
indicate worse disease progression. There are three distinguishable types of cancer recurrence. Treated cancer
cells that disappear during remission return to cause disease recurrence when treatment leaves cells behind
undetectable to the eye. The cancer spreads to neighboring lymph nodes and surrounding body tissues in the
same anatomic region where the tumor originated. The high-risk nature of expanded tumor growth is indicated
through this recurrence style. Worldwide studies, including Mazo et al. (2022) and Noman et al. (2025), have
proved that distant recurrence along with poor survival rates become associated with the most severe type of
cancer return (Mazo et al., 2022; Noman et al., 2025). Doctors require knowledge about the recurrence type
when selecting proper treatments and evaluating patient survival forecasts. B. Factors Affecting Cancer
Recurrence and Survival Rates Cancer recurrence risk and a patient's survival duration prove challenging to
estimate due to multiple influencing factors. These factors include the following: The risk of tumor recurrence
directly depends upon three factors: tumor size, tumor stage, and histological subtype during diagnosis. The risk
of tumor recurrence grows higher when both tumor grade and stage of diagnosis are advanced (Susič et al.,
2023).
The risk of tumor recurrence decreases when patients receive either entire tumor removal surgery or
receive responsive chemotherapy, or targeted therapy treatments. The chances of recurrence become higher
when patients develop treatment resistance or when treatment outcome is only partial (Shimada et al., 2022).
The risk of recurrence increases significantly when patients experience smoking together with obesity and
chronic inflammation (Wei et al., 2024). The worst recurrence risks arise approximately five years following the
initial diagnosis. Breast and prostate cancer, along with selected other cancers, maintain prolonged recurrence
potential beyond ten years of survivorship, according to Dam and Wieder (2024).

Multidisciplinary Journal www.ajmrd.com Page | 75

 The Role of AI in Predicting Cancer Recurrence and Patient Survival Rates

IV. AI TECHNIQUES FOR PREDICTING CANCER RECURRENCE AND SURVIVAL

Advantages of AI over statistical models and clinical prediction methods that predict probabilities of
cancer recurrence and survival are obtained. Deep learning (CNNs, RNNs, and Transformer Models) has
achieved great success as in oncology applications processing complex unorganized data sets (e.g., medical
images and time base clinical information). Using algorithms, deep learning, convolutional neural networks
(CNNs), recurrent neural networks (RNNs), and transformer models, modern prediction techniques make cancer
recurrence, and survival success predictions. CNNs are the main method for image processing, especially so in
medical imaging use cases. The analysis of CT scans, along with MRIs and biopsies through CNN,s enables the
detection of small tumor appearance or dimension modifications that signal cancer recurrence. The spatial
pattern identification abilities of CNNs help medical staff identify early signs of tumor growth and metastasis
development (Shimada et al., 2022). Recurrent Neural Networks (RNNs) show exceptional performance when
operating on sequential data because they were made specifically for time-dependent outcome prediction. The
application of RNNs in cancer prognosis consists of their ability to process patient records containing long-term
information, including treatment reactions and health condition transformations. The temporal patterns of
patient health, which manifests in disease evolution, become predictable by RNNs due to their capability (Yang
et al., 2022). New applications of Transformer models emerged in oncology for analyzing clinical text, genomic
sequences, and various types of structured healthcare data. Transformer models stand out from other approaches
because they handle enormous datasets including EHRs and genomic information and deliver more dependable
survival prognosis results through joint analysis of complex variables (Kaushik et al., 2024).
Tumor Detection benefits from AI-based image analysis through CNNs to achieve top performance
when screening cancerous lesions in medical | images, starting with mammograms up to CT MRIs, and PET
scans. The algorithms detect tiny tissue structure alterations to identify cancers at their earliest stage,s which go
unnoticed by human vision (Akram et al., 2023). Medical detection at an early stage remains vital because it
raises survival outcomes and prevents disease return. Medical staff use this method to monitor tumor
developments across periods for making appropriate treatment decisions (Feng et al., 2022). Artificial
intelligence models review sequential medical images to identify changes in tumor anatomy, enabling them to
project the probability of disease relapse. Artificial Intelligence tracks the changes in patient medical scans after
therapy to identify precursors of tumor growth or distant spread so healthcare providers can implement prompt
therapeutic procedures (Shimada et al., 2022). The process of radiomics allows research teams to extract
massive quantitative data from medical images by analyzing their texture features and shape characteristics
along with the image intensities. The predictive value of recurrence risk and patient survival can be computed
by AI models through deep learning algorithms while analyzing radiomic features (Wei et al., 2024).
Personalized cancer treatment planning receives its power from radiomics technology because this method
offers biological knowledge about tumors that go past visual identification.

Multidisciplinary Journal www.ajmrd.com Page | 76

 The Role of AI in Predicting Cancer Recurrence and Patient Survival Rates

The novel genomic and biomarker analysis powered by artificial intelligence is meant to predict the
patient’s recurrence and survival from cancer .In addition, AI provides an additional approach to exploring and
validating biomarkers based on clinical, genomic, and proteomic data for identifying early detection and
recurrence risk biomarkers, such as liquid biopsy and proteomic analysis In this way, AI takes it to a
comprehensive level by integrating data from genomics, proteomics and transcriptomics (also known as
multiomics data) to get an integrated picture to bring more accuracy in predicting and prognosticating the
cancers based on the condition of the patient. Therefore, the treatment process can also include AI models that
will adapt to evolving data and deliver dynamic risk assessments along the way that could be used to detect
early signs of recurrence to prompt timely interventions. In addition, AI is fundamental to real-time patient
monitoring to utilize genetic tests, imaging, and clinical results in real time to risk of recurrence and use in the
area of precision medicine. In the end, as AI can crunch and understand massive quantities of detailed data, it
can provide the better conjunctured and customized predictions on the recurrence and survival in order to extend
the patient outcomes.

V. METHODOLOGY
Using an extensive method, this study evaluates the contribution of the artificial intelligence (AI) to
predict cancer recurrence and its survival rate. This research framework is integrated based on a systematic
literature review, AI technique analysis based on the data, and case studies on AI applications in oncology. The
purpose of that side of the paper is to examine AI hand versions of risk interpretation, analyse the development
of imaging, and analyse genomic data to predict cancer outcomes.
With regard to the peer review journal articles, conference proceedings and clinical studies pertaining
to AI driven cancer prediction models and Machine learning algorithms or their subset which is deep learning
techniques, a review has been done. This research synthesizes public information from relevant datasets
containing health records, medical imaging, and genomics for analysis to support the analysis. Case studies and
reports from oncological institutions that apply the prediction systems using AI are included. In addition, the
study compares traditional statistical models and AI-based techniques in cancer prognosis. the study compares
AI-powered models with standard statistical methods, such as Kaplan-Meier survival curves and TNM staging
systems, regarding accuracy, efficiency, and real-time adaptation. This evaluation proves AI can process
dynamic patient data and enhance predictive precision. Even so, the research focuses on addressing ethical
problems in automatic implementation, including data privacy issues in the context of AI models, bias in them,
and the demand for interpretability in clinical decision-making.

VI. OBJECTIVES OF THE STUDY

This study attempts to explain the role of artificial intelligence in forecasting cancer recovery and survival rates
of a patient via machine learning algorithms, deep learning models, and artificial intelligence based genomic and
imaging analysis. According to it, it is meant to assess the role of AI in bestowing advanced oncology services
using early warning systems, personalized treatment planning, and real-time healthcare data integration. The
study aims to establish a strong argument and achieves it by identifying and using the current and future AI
technologies to help enhance accuracy of prognosis, optimization of treatment decisions, and patient outcomes.
Sharing final comments underlines the relevance of the ethical perspective and the collaboration of oncologists,
data scientists, and policymakers for the responsible AI introduction in cancer care.

VII. RESULTS AND DISCUSSION

These findings show that artificial intelligence (AI) can predict a patient’s survival and recurrence
better than traditional statistical models. While lots of medical data such as clinical records, imaging scans and
genomic sequence are complex and unstructured, AI techniques, especially deep neural algorithms, have
presented an amazing capacity for processing such high dimensional data. Results demonstrate the development
of prognostic accuracy that far outperforms AI based models due to their capability for dynamic adjustments of
predictions made on real time patient data. The main outcome is the increased ability of convolutional neural
networks (CNNs) to detect changes in medical imaging with minor tumor pathology. When early tumor growth
is recognized in conventional radiology, it is typically too late or CNN-driven image recognition does a far
superior job helping prevent or welcomes timely intervention. It is consistent with a few previous studies like
Shimada et al. (2022), highlighting the importance of AI-enhanced radiomics to enhance the detection of early
stage lung cancer. Recurrent neural networks (RNNs) and transformer models too have strong capabilities of
analysing sequential patient data, tracking health status changes, and improving long term survival predictions
(Yang et al., 2022).
A comprehensive approach to recurrence risk assessment is based on the integration of multiomics data,
genomic, proteomic and transcriptomic data. The results presented by Wei et al. (2024) and Kaushik et al.
(2024) confirm the value of AI in kitting huge datasets to uncover recurrence biomarkers and how to better

Multidisciplinary Journal www.ajmrd.com Page | 77

 The Role of AI in Predicting Cancer Recurrence and Patient Survival Rates

regulate treatment strategies. AI models allow for saving time and eliminating resources that are needed for the
study of data. As compared to traditional methods, the AI method based on AI is dynamic and personalized, in
contrast to the static datasets used in traditional methods (Dam & Wieder, 2024). AI's adaptability, however,
proves incredibly valuable in clinical decision making as they can predict relapse of a disease ahead of
conventional means and drto active treatment manipulation.

VIII. CONCLUSION
The study confirms the results of this study; more so, artificial intelligence (AI) significantly improves
accuracy and efficiency for cancer recurrence and survival prediction compared to traditional statistical models.
Convolutional neural networks (CNNs), recurrent neural networks (RNNs), transformers, and other complex
deep learning algorithms are well suited to analyze complex medical data via AI in medicine. Early tumor
growth is discovered to be resonant with these models, recurrence risk is predicted, and personalized treatment
strategies are outlined with the patient’s clinical and genomic information. AI analyzes images for cancer
detection early on, detecting soft tumor changes in medical scans that can be caught in time and improve patient
outcomes. Furthermore, the multi-omics integration (genomics, proteomics, and transcriptomics) upholds the
AI’s prediction of biomarkers for biomarker prediction, exceptionally personalized oncology. By adapting
forecasts in real-time, real-world patient data can be used to keep the risk assessments up to date and accurate.
It will be a while before we have AI come up with cancer prognosis, the scale is not easily achievable, there are
issues with data standardization, ethical concerns, and model interpretability. Nonetheless, with additional
research and technological progress in overcoming the limitations of AI in oncology, AI will contribute to
achieving the true potential of AI in oncology. Looking forward, the main objective is to develop reliable and
transparent predictive models that could fit seamlessly into the healthcare workflow, maintaining
trustworthiness and transparency for medical practitioners. Overall, AI can facilitate transformation in the field
of cancer anticipation, including increased precision of prognostic findings, reduced healthcare expenses, and
even facilitated discovery of precision medicine. In the coming years, as AI creates more and more
achievements that step closer to antiquity, AI will see more and more changes in oncology, improving cancer
treatment survival rates and quality.

REFERENCES:
[1] Adam, N., & Wieder, R. (2024). AI Survival Prediction Modeling: The Importance of Considering
Treatments and Changes in Health Status over Time.
[2] Mazo, C., Aura, C., Rahman, A., Gallagher, W. M., & others. (2022). Application of artificial
intelligence techniques to predict risk of breast cancer recurrence: a systematic review. Journal of
Personalized Medicine. Retrieved from mdpi.com
[3] Susič, D., Syed-Abdul, S., Dovgan, E., & others. (2023). Artificial intelligence-based personalized,
predictive survival among colorectal cancer patients. Computer Methods and Programs in Biomedicine.
Retrieved from sciencedirect.com
[4] Noman, S. M., Fadel, Y. M., Henedak, M. T., Attia, N. A., & others. (2025). Leveraging survival
analysis and machine learning for accurate prediction of breast cancer recurrence and metastasis.
Scientific Reports. Retrieved from nature.com
[5] Mazaki, J., Katsumata, K., Ohno, Y., Udo, R., & others. (2021). A novel prediction model for colon
cancer recurrence using auto-artificial intelligence. Anticancer Research. Retrieved from
iiarjournals.org
[6] Shimada, Y., Kudo, Y., Maehara, S., Amemiya, R., & others. (2022). Radiomics with artificial
intelligence for the prediction of early recurrence in patients with clinical stage IA lung cancer. Annals
of Surgical Oncology. Retrieved from Springer
[7] Delen, D., Walker, G., & Kadam, A. (2005). Predicting breast cancer survivability: a comparison of
three data mining methods. Artificial Intelligence in Medicine. Retrieved from Elsevier
[8] Akcay, M., Etiz, D., & Celik, O. (2020). Machine learning predicts survival and recurrence patterns in
gastric cancer cases undergoing radiation therapy and chemotherapy. Advances in Radiation Oncology.
Retrieved from ScienceDirect
[9] Javanmard, Z., Shahraki, S. Z., Safari, K., & others. (2025). Artificial intelligence in breast cancer
survival prediction: a comprehensive systematic review and meta-analysis. Frontiers in Oncology.
[10] Yang, Y., Xu, L., Sun, L., Zhang, P., & Farid, S. S. (2022). Machine learning application in
personalized lung cancer recurrence and survivability prediction. Computational and Structural
Biotechnology Journal. Retrieved from ScienceDirect
[11] Fernandez, G., Prastawa, M., Madduri, A. S., Scott, R., & others. (2022). Development and validation
of an AI-enabled digital breast cancer assay to predict early-stage breast cancer recurrence within 6
years. Breast Cancer Research. Retrieved from Springer

Multidisciplinary Journal www.ajmrd.com Page | 78

 The Role of AI in Predicting Cancer Recurrence and Patient Survival Rates

[12] Kumar, Y., Gupta, S., Singla, R., & Hu, Y. C. (2022). A systematic review of artificial intelligence
techniques in cancer prediction and diagnosis. Archives of Computational Methods in Engineering.
Retrieved from Springer
[13] Cruz, J. A., & Wishart, D. S. (2006). Applications of machine learning in cancer prediction and
prognosis. Cancer Informatics. Retrieved from SAGE Journals
[14] Kaushik, A., Madhuranath, B., Rao, D., & others. (2024). Interpreting breast cancer recurrence
prediction models: Exploring feature importance with explainable AI. IEEE Xplore. Retrieved from
IEEE
[15] Wei, G. X., Zhou, Y. W., Li, Z. P., & Qiu, M. (2024). Application of artificial intelligence in the
diagnosis, treatment, and recurrence prediction of peritoneal carcinomatosis. Heliyon. Retrieved from
Cell Press
[16] Jerez-Aragonés, J. M., Gómez-Ruiz, J. A., & others. (2003). A combined neural network and decision
trees model for prognosis of breast cancer relapse. Artificial Intelligence in Medicine. Retrieved from
Elsevier
[17] Feng, Y., Wang, Z., Cui, R., Xiao, M., Gao, H., Bai, H., & others. (2022). Clinical analysis and
artificial intelligence survival prediction of serous ovarian cancer based on preoperative circulating
leukocytes. Journal of Ovarian Research. Retrieved from Springer
[18] Janik, A., Torrente, M., Costabello, L., Calvo, V., & others. (2023). Machine learning–assisted
recurrence prediction for patients with early-stage non–non-small-cell lung cancer. JCO Clinical
Cancer Informatics. Retrieved from ASCO
[19] Palapattu, G. S., Cote, R. J., Sagalowsky, A. I., & others. (2008). Multiple biomarkers improve
prediction of bladder cancer recurrence and mortality in patients undergoing cystectomy. Cancer.
Retrieved from Wiley
[20] Zuo, D., Yang, L., Jin, Y., Qi, H., Liu, Y., Ren, L. (2023). Machine learning-based models for the
prediction of breast cancer recurrence risk. BMC Medical Informatics and Decision Making. Retrieved
from Springer
[21] Akram, F., Wolf, J. L., Trandafir, T. E., Dingemans, A. M. C., & others. (2023). Artificial intelligence-
based recurrence prediction outperforms classical histopathological methods in pulmonary
adenocarcinoma biopsies. Lung Cancer. Retrieved from ScienceDirect

1
Faroque Md Mohsin,
1
MBBS, MPH, MSS Medical officer Directorate General of Health
https://orcid.org/0000-0003-3432-4559

Multidisciplinary Journal www.ajmrd.com Page | 79