Research Output Journal of Public Health and Medicine 5(1):107-114, 2025

ROJPHM ISSN ONLINE: 1115-9715

ISSN PRINT: 1115-6147
https://rojournals.org/roj-public-health-and-medicine/ Page | 107

https://doi.org/10.59298/ROJPHM/2025/51107114

The Role of AI in Tailoring Treatment Plans for

Patients
Nyambura Achieng M.
School of Natural and Applied Sciences Kampala International University Uganda
ABSTRACT
Artificial Intelligence (AI) is transforming modern medicine by enabling the development of personalized
treatment plans tailored to individual patients’ genomic, physiological, and behavioral data. With
increasing access to large-scale biomedical datasets and advanced deep learning techniques, AI plays a
central role in precision medicine by enhancing drug discovery, optimizing treatment schedules, and
matching therapies to patient-specific profiles. This paper explores the historical evolution, current
applications, and future potential of AI in clinical decision-making, particularly in oncology and
neuroscience. Key AI technologies, such as machine learning, deep learning, and natural language
processing, are examined in the context of their roles in diagnostics, treatment planning, and patient
monitoring. Despite promising results, challenges such as algorithmic bias, data privacy, regulatory
barriers, and the interpretability of AI-driven decisions continue to limit their full integration into clinical
practice. The study highlights the importance of transparency, collaboration, and ethical oversight in
advancing AI-driven healthcare solutions that are inclusive, reliable, and equitable.
Keywords: Artificial Intelligence (AI), Personalized Medicine, Deep Learning, Machine Learning,
Clinical Decision Support, Precision Medicine.
INTRODUCTION
With substantial advances in recent years, Artificial Intelligence (AI) models for data interpretation
utilizing deep learning solutions have been assessed in numerous biomedically relevant applications.
These include medical imaging systems for disease screening and management, precision medicine (PM)
solutions for matching patients with targeted drugs tailored to their specific genomic/genetic alterations,
and gene and genetic variant prioritization pipelines. Encouraging initial demonstrations of success and
feasibility for all three types of AI solutions have been published. In the context of PM, recent efforts to
deploy AI solutions encompassing a broad range of several critical areas, including drug discovery and
development, targeted (anti-)cancer drug prediction, repurposing existing drugs, treatment scheduling
and dose optimization, and pairing patients with combination treatment plans, particularly chemotherapy
and immunotherapy have shown promise. However, as the AI-based treatment solutions reach
technological maturity, their clinical implementation may be impeded by challenges and concerns
associated with various technology-driven, methodological, and ethical issues. These challenges relate to
both primary AI and medical issues affecting day-to-day operations and usage effectiveness in a real-
world clinical setting. For their development, evaluation, and deployment, AI systems should ensure
transparency, accountability, and interpretability of data input sources and flow chart steps for user
feedback, scrutiny, and trust. A recent analysis of an AI-based radio-therapy regimen generator for
prostate cancer presented promising results for its clinical applicability. A typical patient treatment plan

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
generated consisted of a median of 866 control points, and 89% of the plans generated were deemed
clinically acceptable [1, 2].
Understanding AI in Healthcare
IIn recent years, artificial intelligence (AI) has gained prominence in precision medicine (PM) due to its
ability to efficiently process large multi-omics data. PM approaches typically consider tissue, blood, and
imaging biomarkers, along with polygenic risk scores, to tailor treatment plans for patients. However,
successfully implementing these approaches involves extensive data engineering and integrating various Page | 108
biomarkers to predict treatment responses, all of which have become increasingly complex due to the rise
of multi-omics methodologies that analyze diverse biological data. This complexity has made
neuroscience PM approaches challenging to apply in practice. Though the foundational studies focus on
several key areas, the AI techniques explored usually reflect how these studies were conducted. AI is often
seen as an interdisciplinary umbrella encompassing various sub-disciplines, making it difficult to define
precisely. Many PM researchers, primarily consisting of physicians and biomedical engineers, often lack
in-depth AI knowledge, which limits their ability to leverage AI's potential. Over recent decades, there
has been growing interest in applying AI to healthcare-related challenges. As biomedical data volumes
surge, understanding AI's role as a crucial technology to manage this influx is vital. This review will
discuss both algorithmic and non-algorithmic AI and summarize current anticipated use cases in
healthcare with specific examples. It aims to provide a comprehensive overview for researchers interested
in using AI to address healthcare issues and serve as a reference for healthcare leaders seeking insights
into AI's contributions to their organizations [3, 4].
Definition of AI
Artificial intelligence (AI) encompasses systems that emulate human cognitive functions and behaviour to
resolve problems, amplify productivity, foster innovation, and facilitate novel solutions. The development
of AI has spurred global interest as well as concerns among governments, corporations, and citizens
regarding its efficacy and security. A subset of AI is machine learning (ML), wherein the system learns
from input data to yield predictions or estimates. ML entails various approaches, with deep learning (DL)
being a high-performing implementation of ML. Devices developed using deep learning employ
multilayered networks of mathematical functions to glean autonomous features from inputs while
continuously refining their parameters for optimal prediction. It thus functions more like a human brain
than traditional computation models. AI’s growing use in enhancing productivity across diverse sectors
has propelled interest and investment in its medical applications, as it holds the promise of improving
disease understanding and management with far-reaching effects on patient treatment and remedy
accessibility. The clinical approach has entered the big data era, owing to the expanse of physiologic
measurements, health data, and genomic information with diverse temporal and geographical scales,
which have fueled the growth of biomedical data sciences. Machine learning (ML) and deep learning (DL),
AI technologies with improved scalability and versatility for data processing, notification, pattern
recognition, simulation, and decision assistance, can be trained on these data sources to improve health
understanding and management. However, the absence of appropriate platforms and clinical trial designs
is hindering the implementation of recent translational advances of AI to practice. Despite the rapid rise
of this brand new field, many basic questions are yet to be answered, such as whether the amalgamation of
datasets, platforms, methods, and knowledge from multiple domains to develop AI technologies can
benefit health care. And if so, what are the required preconditions, designs, and regulations? Will AI-
aided diagnostics and treatments be able to stand against or even enhance human understandings and
decision-making? [5, 6].
History of AI in Medicine
Artificial intelligence (AI) has been available in rudimentary forms for many decades. Simple expert
systems were first described in the 1970s, and clinical decision support tools are a form of AI that has also
been available for decades. However, AI methods had little practical impact on the practice of medicine for
many decades. Beginning around 2012, AI has, however, emerged as an increasingly important tool in
healthcare. This is evident from the sharp increase in publication numbers in this field, as well as the
recent approval of AI-based devices for clinical use. These devices are now capable of processing image
data, making diagnoses, predicting biomarkers for solid tumors, and more 7. However, the development
of AI in medicine is still in its infancy. While many AI-based devices have recently been introduced to the
market, most literature still derives from early prototypes or methods that have not found translation into

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
mainstream clinical practice. Additionally, since 2022 there have been exponential technical
advancements in AI, with some AI programs now demonstrating human-level understanding of image
and text data. The idea of machines emulating human cognition dates back to antiquity. However, the
field of AI arose in earnest as an academic discipline in the 1950s when NASA coined the term. In the
1960s and 1970s, the first rudimentary AI systems, often based on rules or decision trees, were described.
Early successes led to overconfidence and unrealistic expectations so that the field faced reduced funding
and diminished interest during the so-called “AI Winter” in the late 20th century. There were successful Page | 109
ideas from AI such as expert systems found applications in healthcare or financial domains but with
limited impact. The introduction of deep learning and the availability of graphics processing units (GPUs)
for demanding computations led to massive advancements in AI since 2012. This newly found power was
quickly adopted by the tech industry, and today the majority of newly available AI methods are designed
for and applied in private sector domains such as technology, insurance, finance, or logistics [8, 9].
Personalized Medicine
Personalized medicine originates from specialized medicine and is a dynamic process involving intrinsic
knowledge about a patient’s treatment response, considering pharmacokinetic and pharmacodynamic
parameters. Treatment plans are continuously modulated based on individual responses, prior treatment
complexity, and defined criteria. Cancer therapies rarely succeed uniformly across all patients,
necessitating treatment modulation when heterogeneous responses are detected. This shift from a narrow
view of disease classification to a universal carcinogenic perspective allows for a more individualized
treatment approach, leading to hyper-personalized interventions. The prescription of medications can be
enhanced with artificial intelligence (AI) using individual patient data. AI aids in formulating hypotheses,
identifying risk factors, and assisting in trial designs through simulation and drug screening workflows
focused on single cases rather than population data. AI clinical trials consist of calibration and efficacy-
driven phases. Calibration trials aim to establish dose modulation methods supported by predictive
models, engineering patient-specific clinical trials while minimizing population learning, ensuring
attention remains on the individual patient [10, 11].
AI Technologies In Treatment Planning
Few recent advancements show more promise in the oncological field than artificial intelligence. The
ability of machines to work through the vast amounts of medical documents to find new insights goes
well beyond human capabilities. New treatment techniques that were developed from guidelines or
literature research can be integrated to AI systems or planning software, and all known cognitive
algorithms can be used to generate a treatment plan. However, the additional workload of using review
systems, and the knowledge of the treatment planning process still needed for statistical treatment plan
generation, are major barriers for facilities to start using machine learning systems in clinical practice,
similar to most machine learning systems. In most countries, reimbursement and hospital budgets are
tight. Plans generated by AI can provide the treating physician with additional knowledge and expertise
on top of the plans generated by manual planning. Additionally, AI can be used as a first reviewer to
create real-time benefit-risk ratios and plausibly optimal treatment plans on which the treatment planner
can base the final product. Most of the countries have restrictions on using third-party software: they are
often illegible for reimbursement and have limited guided workflows and development options. A vendor-
neutral solution required the creation of an extensive responsive user interface and plan review interfaces.
Each review step lacked guidelines like the one that guided user insight into treatment plans in the first
visit. The need for a foundation review plan that provides the reviewers with a review that goes beyond
just safety aspects cautions against any button being found sufficient. But careful fielding of these plans to
the clinical users should greatly complement the plans already reviewed by consistent and explanatory
comments and provide further insights into planning objectives and which modalities could prospectively
treat better [12, 13].
Data Sources for AI Algorithms
The development, validation, and implementation of AI algorithms occur within a cyclical process that
considers a range of issues, beginning with a business problem and concluding with the deployment of an
algorithm into the healthcare system. AI algorithms operate as an identification tool that interprets the
inputs seen during training. The vast majority of imaging data employed during development may take
the form of 2D images and volumetric (3D) data, with a wide range of formats. To deploy deep learning
algorithms into a health information system requires interoperability with a resource-constrained,

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
heterogeneous environment, thus necessitating pre-processing of the input data. Both domain knowledge
and appropriate engineering resources are critical to the success of the deployment phase. Improvements
in the detection capabilities of AI algorithms typically rely on the effort put into the preprocessing phase.
As a direct takeaway, state-of-the-art AI algorithms rely on a lot of development effort on a wide range of
pre-processing techniques that, when not present in a production setting, can result in a significant
performance drop. Moreover, the production-ready algorithms considered in this paper are based on
commercial AI packages that allow for the implementation of the models by end-users with limited Page | 110
programming experience. Since these commercial approaches are closed, there are obvious concerns about
technological obsolescence and product life cycle. Therefore, companies and institutions investing in this
technology should consider the risks involved in the sustainability of the solutions developed as well as
the requisite activity needed to transition from the currently accessible development packages into self-
managed deployment-ready solutions. However, the majority of AI algorithms targeting long-term
utilization are either in the multiinstitution evaluation phase or remain as clinical proof-of-concept
studies. Consequently, although less connection-focused algorithms have been developed and are now
gaining traction within hospitals internationally, the majority of commercial efforts focus on algorithms
combining both 3D and 2D information. These algorithms must be trained on larger and more various
data repositories, thus raising the need for ongoing research cooperation and investment [14, 15].
AI-Driven Decision Support Systems
The advent of internet-enabled devices has facilitated the emergence of vast amounts of health-related
data, ranging from deep unstructured data from electrocardiogram or histopathology images to semi-
structured data, such as voice and text transcripts of doctor-patient conversations and free-text reports
on patients' conditions. This influx of information streams has the ability to revolutionise patient care and
enhance the quality of healthcare delivery. However, the increasing volume serves as a looming threat to
healthcare professionals, who confront a tidal wave of unmanageable data streams daily, leading to missed
opportunities and oversights in patient care. Advanced technologies that manage health-related content
already exist, but providing these technologies with the necessary healthcare knowledge and clinical
reasoning required to solve clinical problems is challenging. This gap between vast amounts of useful
data and tangible knowledge has rendered already available tools ineffective, necessitating the
development of more sophisticated technologies capable of bridging the gap. This creates an opportunity
for intelligent means possessing reasoning capabilities. AI has the potential to address this challenge by
using a variety of technologies, including ones capable of using both qualitative and quantitative
knowledge. Identification and intelligent extraction of the patient's picture of the world/psychological
state are covered; and, most importantly, intelligent decision-making procedures for personalized
adaptive treatments to care. All these AI technologies act in synergy with one another. For instance,
using sentiment analysis on recorded doctor-patient conversations helps extract the state of distress,
laden with qualitative uncertainty. If a high degree of patient dissatisfaction arises from the sentiment
analysis, a logical reasoning approach to diagnosis could be triggered to infer the cause of that
dissatisfaction. This approach explains the rationale behind the generated treatment plan. Reasoning
capabilities grounded on ontologies could be fed into fuzzy AI systems to create healthcare outcomes
expressed in natural language that could be easily understood by patients [16, 17].
Challenges In Implementing AI
Despite the emerging role of AI in identifying suitable treatment plans, challenges remain in
implementing them in clinical practice. The conventional approach to drug development and repurposing
requires applying for new drug applications (NDA) after conducting safety and efficacy studies based on
clinical trial phases I, II, III, and IV. The lack of certainty in AI drug repurposing further complicates the
enactment of drug repurposing texts developed by AI. Technically, AI devices and techniques are still in
their infancy in drug repurposing. Most drug repurposing methods are not novel and only use traditional
approaches. Moreover, most AI techniques relevant for drug repurposing do not lend themselves to easily
interpretable decision insights or explanations. It is difficult for specialized professionals to comprehend,
evaluate, and validate drug repurposing decisions rendered by such black-box AI tools. Another critical
challenge is related to biased and data privacy concerns. Only a few companies make their datasets
publicly available for research. Moreover, most proprietary datasets do not contain information on
ethnicities, which is vital for health disparities research. Access to such data is often limited to projects
draped in confidentiality clauses, resulting in biased datasets that fail to reflect the full spectrum of the

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
target population. Furthermore, most protected health information (PHI) must be de-identified to protect
privacy, further complicating dataset accessibility. The black-box nature of some AI algorithms invites
claims about lack of trust in AI systems and difficulties regarding liability and accountability if something
goes wrong. The lack of transparency, reliability, and faithfulness in the insights generated by AI systems
fuels distrust, while questions about the sovereignty and accountability of AI systems arise. These
challenges must be adequately addressed to facilitate the broader adoption of AI [18, 19].
Case Studies Page | 111
Clinical medicine is an area closely related to traditional medicine that is benefitting greatly from big
data, machine learning, and AI. From 1990-2019, publications pertaining to AI in clinical medicine
increased from less than 50 to over 1,350. The majority of these papers focus on the analysis of clinical
imaging and pathology data to assist the diagnostic processes. Similar to the genomics field, much of the
AI work in clinical medicine is centered around refining outputs generated by existing methods. Among
dozens of identified AI approaches for repositioning anti-cancer compounds, the most advanced models
are ensemble methods combining multiple approaches. In a blinded collaboration with a pharmaceutical
partner, the methods successfully identified a follow-up cancer drug with human activities within the top
four compounds predicted to be active, out of over 4,000 tested. Chemoinformatics approaches dissected
structural information from a library of drugs, while machine learning models assisted in the selection of
prospective compounds across major target families. Experimental testing confirmed the predicted re-
purposing of a number of anti-cancer compounds. As evidence accumulates regarding the associations
between drug side-effects and other biological properties for drug candidates, these data can also assist in
the discovery of new uses for existing drugs. An EU-funded project focusing on repositioning compounds
for repurposing as a therapy for Alzheimer’s disease developed an approach utilizing chemical and
biological data combined with association information from the literature. The objective of the approach
was to generate and prioritize compound candidates screened from commercial ready-to-test libraries.
The approach successfully identified compounds that impinge on different biological targets. Here,
association information from multiple information sources including literature-based, gene-related, side-
effect-related were utilized [20, 21].
Future Directions
AI-driven models have shown promising applicability in precision medicine but must address numerous
challenges before they surpass conventional approaches. There is a pressing need for additional rigorous
validation studies on the detection and interpretation of molecular biomarkers based on AI approaches.
Application of resources on the development of tools for open-access comparison among state-of-the-art
models and consequently selection of the optimal model to be utilized in a given setting is warranted. It is
highly desirable to establish libraries of AI-generated drugs to ensure novelties in potential treatment
options, which may also facilitate drug repurposing. The rapid evolution and proliferation of AI tools and
pipelines to aid in preclinical drug research has spawned a paradigm shift for research and clinical
communities. However, despite their enormous potential, potential users should remain cautious against
excessive marketing claims of AI tool developers and vendors, and striking a balance between hot-spot
innovations and established, conventional techniques should be needed to maximize the utility of
resources. Most importantly, regulation and legislation should keep pace with the vast changes
introduced by the adoption of AI. Granting certain hallmarks of citizenships could foster responsible
development and deployment of AI, while the setting up of global standards for tools, numerical data, and
interpretability protocols would facilitate comprehensive reviewing of algorithms [22, 23].
Ethical Considerations
The role of AI in treatment planning should feature minimal automation, with doctors leading decision-
making. AI can assist but should not take over. Using non-small cell lung cancer patients as an example,
it highlights the trade-offs between forecasting treatment outcomes with AI versus traditional methods.
The goal is to align new medical decisions with existing decision theories, resembling feature extraction
in supervised learning rather than policy transfer in reinforcement learning, as forecasting relies on
pattern similarity. Heuristics often govern decision-making; non-experts may struggle with decision
theory, while experts might only approximate correctness. Intuitive forecasting techniques like scenario
planning can aid policy-making despite uncertain logic. Doctors should blend their insights with AI
predictions, evaluating personal medical evidence and interpreting AI forecasts through their knowledge
and ethics. This would combine empirical reasoning with fairness principles, enhancing AI transparency

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
and patient involvement. Medical decisions should remain rooted in doctors’ judgment over AI outputs. A
comprehensive response requires a true representation of reality, including facts and decision criteria. In
machine ethics, AI alignment with human values is crucial; misalignment would yield optimal decisions
for the AI itself rather than humanity. Value alignment, in tandem with uncertainty reasoning and model
fidelity, represents another key dimension of the first H2 question. AI safety is typically framed around
three main concerns: unintended misuse, undesired side effects, and control issues [24, 25].
Patient Perspectives Page | 112
Incorporating perspectives from patients with implantable cardioverter defibrillators (ICDs), this study
highlights concerns regarding future developments in the use of AI technology for risk stratification and
treatment planning in SCD (sudden cardiac death) prevention. Given the highly personal implications of
implanting an ICD, patients desire decision-making that takes into account human experiences and
emotions. AI-driven decisions may jeopardize this. In general, incorporating observational, real-world,
and unstructured data may lessen the need for and accessibility of specialist knowledge, altering the
dynamics of the patient-doctor relationship. Although AI has noteworthy advantages, purpose and design
will determine its ethical and societal acceptability. In the present context, addressing conditions in which
implementing the same ML algorithm has different effects is equally relevant for those involved in the
development of personalized medicine, such as AI-based disease risk stratification and treatment
planning. Whereas patients unanimously viewed data-driven protocolization of work processes with great
skepticism, in parts of this, a proper and careful implementation of the notion of developing AI-based
personalized medicine was welcomed. Given the intentions and interests involved, while favoring data-
driven systems offers an opportunity for further investigation, supporting determinism is just as risky. It
could inadvertently legitimize a concept that entails serious injustices span to large groups of patients
[26, 27].
CONCLUSION
AI has emerged as a transformative force in the landscape of personalized medicine, offering
unprecedented opportunities to tailor treatment plans based on individual biological and clinical
characteristics. From automating data interpretation and supporting diagnosis to designing patient-
specific therapeutic regimens, AI technologies have demonstrated considerable promise in enhancing
healthcare outcomes. However, realizing AI's full potential in clinical settings requires overcoming
significant hurdles, including issues of data quality, transparency, ethical use, and regulatory compliance.
Equally critical is the need for interdisciplinary collaboration among clinicians, data scientists, and
policymakers to develop frameworks that ensure the responsible use of AI while maintaining patient
trust. As the field progresses, integrating AI with existing medical infrastructure in a way that
emphasizes interpretability and inclusivity will be essential to fostering broader adoption and ensuring
equitable health benefits for all populations.
REFERENCES
1. Naskar S, Sharma S, Kuotsu K, Halder S, Pal G, Saha S, Mondal S, Biswas UK, Jana M,
Bhattacharjee S. The biomedical applications of artificial intelligence: an overview of decades of
research. Journal of Drug Targeting. 2024 Dec 31:1-32. researchgate.net
2. Doke-Bagade PE, Bagade O. Applications of AI-Based Models in the Field of Biomedicine.
InBiosystems, Biomedical & Drug Delivery Systems: Characterization, Restoration and
Optimization 2024 Jun 14 (pp. 259-286). Singapore: Springer Nature Singapore. [HTML]
3. Srivastava R. Applications of artificial intelligence multiomics in precision oncology. Journal of
cancer research and clinical oncology. 2023 Jan;149(1):503-10.
4. Agarwal N, Nupur, Paul PK, Mishra SK. Artificial Intelligence and Machine Learning for
Analysis of Multi-omics. InMulti-Omics Analysis of the Human Microbiome: From Technology
to Clinical Applications 2024 May 30 (pp. 339-354). Singapore: Springer Nature Singapore.
[HTML]
5. Abbasi N, Nizamullah FN, Zeb S. AI in healthcare: integrating advanced technologies with
traditional practices for enhanced patient care. BULLET: Jurnal Multidisiplin Ilmu. 2023 Jun
13;2(3):546-56. researchgate.net
6. Sherani AM, Khan M, Qayyum MU, Hussain HK. Synergizing AI and Healthcare: Pioneering
advances in cancer medicine for personalized treatment. International Journal of
Multidisciplinary Sciences and Arts. 2024 Feb 4;3(2):270-7. [HTML]

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
 7. Kather JN. Artificial intelligence in oncology: chances and pitfalls. Journal of Cancer Research
and Clinical Oncology. 2023 Aug;149(10):7995-6.
8. Krishnamoorthy S, Dua A, Gupta S. Role of emerging technologies in future IoT-driven
Healthcare 4.0 technologies: A survey, current challenges and future directions. Journal of
Ambient Intelligence and Humanized Computing. 2023 Jan;14(1):361-407. researchgate.net
9. Junaid SB, Imam AA, Balogun AO, De Silva LC, Surakat YA, Kumar G, Abdulkarim M, Shuaibu
AN, Garba A, Sahalu Y, Mohammed A. Recent advancements in emerging technologies for Page | 113
healthcare management systems: a survey. InHealthcare 2022 Oct 3 (Vol. 10, No. 10, p. 1940).
MDPI. mdpi.com
10. Lotter W, Hassett MJ, Schultz N, Kehl KL, Van Allen EM, Cerami E. Artificial intelligence in
oncology: current landscape, challenges, and future directions. Cancer discovery. 2024 May
1;14(5):711-26.
11. Derraz B, Breda G, Kaempf C, Baenke F, Cotte F, Reiche K, Köhl U, Kather JN, Eskenazy D,
Gilbert S. New regulatory thinking is needed for AI-based personalised drug and cell therapies
in precision oncology. NPJ Precision Oncology. 2024 Jan 30;8(1):23.
12. Kolluri S, Lin J, Liu R, Zhang Y, Zhang W. Machine learning and artificial intelligence in
pharmaceutical research and development: a review. The AAPS journal. 2022 Feb;24:1-0.
13. Minz NK, Sha A, Yadav M. Adapting to Change: Strategic Evolution of Human Resource
Management Practices in SMEs Amidst Shifting Market Dynamics. InInnovative Human
Resource Management for SMEs 2024 (pp. 148-164). IGI Global Scientific Publishing.
14. Masoudi S, Harmon SA, Mehralivand S, Walker SM, Raviprakash H, Bagci U, Choyke PL,
Turkbey B. Quick guide on radiology image pre-processing for deep learning applications in
prostate cancer research. Journal of Medical Imaging. 2021 Jan 1;8(1):010901-.
spiedigitallibrary.org
15. Tunkiel AT, Sui D, Wiktorski T. Impact of data pre-processing techniques on recurrent neural
network performance in context of real-time drilling logs in an automated prediction framework.
Journal of Petroleum Science and Engineering. 2022 Jan 1;208:109760. sciencedirect.com
16. Johnson KB, Wei WQ, Weeraratne D, Frisse ME, Misulis K, Rhee K, Zhao J, Snowdon JL.
Precision medicine, AI, and the future of personalized health care. Clinical and translational
science. 2021 Jan;14(1):86-93. wiley.com
17. Ghanem M, Ghaith AK, Bydon M. Artificial intelligence and personalized medicine:
transforming patient care. InThe New Era of Precision Medicine 2024 Jan 1 (pp. 131-142).
Academic Press.
18. Kumar S, Roy V. Repurposing drugs: An empowering approach to drug discovery and
development. Drug Research. 2023 Nov;73(09):481-90.
19. Niazi SK, Mariam Z. Artificial intelligence in drug development: reshaping the therapeutic
landscape. Therapeutic Advances in Drug Safety. 2025 Feb;16:20420986251321704.
20. Adams D, Tournev IL, Taylor MS, Coelho T, Planté-Bordeneuve V, Berk JL, González-Duarte
A, Gillmore JD, Low SC, Sekijima Y, Obici L. Efficacy and safety of vutrisiran for patients with
hereditary transthyretin-mediated amyloidosis with polyneuropathy: a randomized clinical trial.
Amyloid. 2023 Jan 2;30(1):18-26. tandfonline.com
21. Ktena I, Wiles O, Albuquerque I, Rebuffi SA, Tanno R, Roy AG, Azizi S, Belgrave D, Kohli P,
Cemgil T, Karthikesalingam A. Generative models improve fairness of medical classifiers under
distribution shifts. Nature Medicine. 2024 Apr;30(4):1166-73. nature.com
22. Kalusivalingam AK, Sharma A, Patel N, Singh V. Leveraging Deep Learning and Random
Forest Algorithms for AI-Driven Genomics in Personalized Medicine. International Journal of
AI and ML. 2021 Feb 15;2(3). cognitivecomputingjournal.com
23. Boniolo F, Dorigatti E, Ohnmacht AJ, Saur D, Schubert B, Menden MP. Artificial intelligence in
early drug discovery enabling precision medicine. Expert Opinion on Drug Discovery. 2021 Sep
2;16(9):991-1007. tandfonline.com
24. Conroy L, Winter J, Khalifa A, Tsui G, Berlin A, Purdie TG. Artificial Intelligence for Radiation
Treatment Planning: Bridging Gaps From Retrospective Promise to Clinical Reality. Clinical
Oncology. 2025 Jan 1;37:103630. clinicaloncologyonline.net

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
 25. Satheeskumar R. AI-driven diagnostics and personalized treatment planning in oral oncology:
Innovations and future directions. Oral Oncology Reports. 2025 Mar 1;13:100704.
26. Varzakas T, Antoniadou M. A holistic approach for ethics and sustainability in the food chain:
the gateway to oral and systemic health. Foods. 2024 Apr 17;13(8):1224.
27. Elhaddad M, Hamam S. AI-driven clinical decision support systems: an ongoing pursuit of
potential. Cureus. 2024 Apr 6;16(4).
Page | 114
CITE AS: Nyambura Achieng M. (2025). The Role of AI in Tailoring Treatment Plans
for Patients. Research Output Journal of Public Health and Medicine 5(1):107-114.
https://doi.org/10.59298/ROJPHM/2025/51107114

This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.