17.1.1. Definition and overview of AI

AI is the development of computer systems that can perform tasks that would require human intelligence, such as learning, reasoning and self-correction. Examples found in the health industry include machine learning (ML), natural language processing (NLP) and computer vision applied to interpret complicated medical records. Such systems aid in the diagnosis of diseases, in predicting prognoses and in prescribing individualized treatment (Sonar 2025). For example, AI ML algorithms could scour vast electronic health records (EHRs) for patterns that improve diagnostic accuracy or predict disease progression. AI is also being widely used in drug discovery, analysis of medical images and in tracking patients remotely, offering an opportunity to enhance patient care and cut out inefficiencies from operations.

17.1.2. Evolution of AI in medicine

The history of AI in medicine demonstrates many decades of technological progress. Early work started in the 1950s with the idea of ML algorithms that could work on simple computational problems (Singh 2025). During the 1970s, expert systems such as MYCIN were created to guide doctors in diagnosing bacterial infections and suggesting subsequent treatments. AI was introduced in EHRs during the 1980s and 1990s, making data-driven clinical system decisions possible.

Things changed significantly once deep learning (DL) models were introduced in the early 2000s, which could make sense of complex data, like medical images used to detect diseases, where utility has been demonstrated in starch AI for predicting robotic-assisted surgeries and real-time patient monitoring. Figure 17.1 shows major milestones in the evolution of AI in healthcare. It all began with the rise of AI technology in the 1950s, and continued to the time of AI expert systems like MYCIN in the 1970s. The use of AI in EHRs in 1980–1990s was a milestone in evidence-based health practice. It was not until the early 2000s that more powerful models such as DL ones came into existence, capable of interpreting complex datasets (such as the medical images used to detect diseases). Not to mention, advisable use-cases now include sophisticated tech such as predictive analytics and robot-assisted surgery. They are transformational in terms of how care is being delivered in healthcare. Figure 17.1 presents a short history of AI alongside a timeline of its increasing relevance to changing healthcare practice. From diagnoses that are more accurate to individualized treatment plans, AI is revolutionizing the way the medical industry is using and looking at patient care (Chatzivasileiou et al. 2025).

17.2.1. Current trends in AI applications in healthcare

AI has been applied to healthcare on a large scale in these few years, with breakthroughs in diagnosis, personalized medicine and patient care (Patel and Dave 2023).

Key trends include:

• AI-powered diagnosis: AI tools are increasingly being leveraged to analyze medical imaging data to detect diseases at an early stage. For instance, AI systems can now detect breast cancer from mammograms with more than 94% accuracy, as effectively as expert radiologists.
  
• Predictive analytics: AI models forecast patient outcomes based on EHRs, facilitating early action on diseases such as sepsis and heart failure.
  
• Virtual health and public information: AI-powered chatbot systems can be driven by generative AI for symptom triage and patient engagement/education to offload the system and enhance access.
  
• Neurorehabilitation robotics: AI-driven robotic devices are used to help patients regain their mobility, especially after stroke or surgery.
  
• Drug discovery: AI speeds the discovery of treatment targets and drug development, making timelines and budgets shorter.

In Figure 17.2, we summarize the significant developments in the application of AI to healthcare from 2019 to 2025. It displays AI’s presence in predictive analytics, robotics, virtual patient care and drug discovery. The most recent breakthrough, in 2025, highlights AI’s athletic diagnostic skills through state-of-the-art imaging.

17.2.2. Comparative study of AI methods in physiotherapy and oncology

AI methods have been developed to solve the specific problems of physiotherapy and oncology (Fernandes et al. 2023). Table 17.1 presents the different uses of AI in physiotherapy and oncology, stressing the specificity of the applications in terms of approach, technology and issues. AI is used in physiotherapy for motion analysis and personalized physiotherapy programs for mobility recovery and in oncology for imaging-based diagnostics and genomic sequencing for early cancer detection and personalized therapy. For physiotherapy, there is a relative scarcity of available training data for the models, while for oncology, a challenge is the heterogeneity of multi-mode data (e.g. genomics and imaging) into the model. Both fields offer significant future potential: physiotherapy with wearable sensors for remote care and oncology with predictive models for personalized medicine.

The following table presents a summary of a comparative study.

17.3.1. Research design and framework

The study is based on a mixed-methods methodology, comprising both empirical data analysis and a literature review to investigate AI-enabled healthcare innovations (Sigatapu et al. 2023). The framework incorporates:

• Exploratory analysis: investigating previous work on AI-based interventions in physiotherapy and oncology to acquire background knowledge.
  
• AI model identification: assessment of AI-based models (DL, NLP, predictive analytics, etc.) in the context of healthcare.
  
• Validation methods: strap cross-validation method, craziness matrix and statistical indices to validate AI models for diagnostic, personalized treatment and rehabilitation.
  
• Comparative study: rivalry of AI-enabled versus conventional medical devices in physiotherapy and oncology that demonstrates superior efficiency and precision.
  
• Ethical concerns: dealing with data privacy, biases in AI algorithms and regulation in healthcare applications with a focus on oncology and rehabilitation.

The structure of AI-enabled applications in healthcare is shown in Figure 17.3.

Figure 17.3 represents various healthcare fields where AI is implemented, for example, diagnosis, the treatment process and rehabilitation. AI-based imaging methods and predictive analytics make diagnosis more accurate and AI-powered drug discovery and personalized medicine make treatment smarter. There are also rehabilitation gains with robotic-assisted therapy and telemedicine solutions, leading to better patient recovery in both physiotherapy and oncology.

17.3.2. Data collection and analysis techniques

The study adopts the methodology of systematic data collection and analysis, which is a robust and valid approach to understanding the transformative contributions of AI in physiotherapy, oncology and general healthcare (Singh et al. 2023).

• Dataset selection: the data are collected from open-source EHRs, medical imaging databases (e.g. magnetic resonance imaging (MRI), computed tomography (CT) scans), genomic databases and by using real-world case studies in physiotherapy and cancer treatment (Jimma 2023).
  
• Preprocessing techniques: AI models need comprehensive preprocessing, such as augmented images, normalization, characterizing features, and removing unwanted data in databases that can be used for DL.
  
• ML pipelines: for developing various algorithms such as CNNs (for medical image analysis), recurrent neural networks (RNNs) (for patient data processing) and reinforcement learning (for robotic rehabilitation) with various combinations.
  
• NLP: for processing unstructured medical notes and patient feedback to influence decision-making within cancer and physiotherapy practice.
  
• Performance measures: AI model performance is evaluated using accuracy, recall, precision, F1 score and area under the curve (AUC). Interpretation and explainability methods, such as SHAP (Shapley Additive explanations), are also explained to ensure transparency in AI predictions.
  
• Long-term analysis: for examining the sustainability and effects of AI-assisted diagnostics and treatment models and their influence on the patients, especially on oncology rehabilitation.
  
• Comparative benchmarking: AI-enabled healthcare models can be compared to clinical techniques based on diagnostic accuracy, effectiveness of treatment and the patient’s recovery process.

17.4.1. Improving the discriminative performance

AI greatly enhances the diagnostic accuracy with medical image analysis and early disease identification. Techniques have enhanced radiology by detecting the defaults in MRI, CT scans and X-rays, when compared with the difference in accuracy with state-of-the-art methods, such as deep neural network (DNN), convolutional neural networks (CNNs) and generative adversarial networks (GANs). AI-driven symptom checkers and diagnostic models help minimize false positives and false negatives, enabling earlier intervention for conditions such as Alzheimer’s disease, Parkinson’s disease, and various cancers. In addition, AI-powered pathological analysis enhances the detection of tumors and genetic (molecular) mutations, supporting more accurate and effective therapeutic decision-making (Ling et al. 2023).

17.4.2. Customized treatment strategies

AI also powers precision medicine, treating patients based on their genetic profile, medical history and real-time health data. ML algorithms sift through terabytes of data to identify the best therapies, minimize side effects and maximize treatment effectiveness (Musat et al. 2023). In regard to AI models, NLP-based AI models that analyze patient history as well as medical literature can help clinicians give evidence-based treatment. AI is also improving drug discovery, increasing the speed at which potentially effective compounds are found for complex diseases such as cancer and neurodegenerative diseases.

17.4.3. Early disease prevention and predictive analytics

AI-based predictive analytics can analyze extensive health datasets, searching for patterns that indicate the early onset of disease (Seoni et al. 2022

Related posts:

The Role of Artificial Intelligence in Personalized Physiotherapy and Cancer Treatment Model Validation Techniques for AI in Cancer Research Based on Physiotherapy and Oncology Machine Learning in Oncology and Physiotherapy: A New Era of Precision Medicine Automating Cancer Detection and Rehabilitation: The AI Revolution AI and Wearable Technology in Physiotherapy for Oncology Patients Smart Robotics in Physiotherapy and Oncology: Redefining Patient Outcomes

Stay updated, free articles. Join our Telegram channel

Join