The recent advancements in the field of neurorehabilitation and patient monitoring have significantly leveraged the capabilities of wearable technology and deep learning models. A notable trend is the shift towards object-centric approaches for recognizing Activities of Daily Living (ADL) in real-world rehabilitation settings, which provide clinically interpretable information about functional object use, robust to patient-specific movement variations. Another key development is the use of foundation deep learning models fine-tuned for specific neurodegenerative diseases, such as Huntington's disease, to accurately detect gait bouts in daily living environments, overcoming challenges posed by involuntary movements. Additionally, there is a growing emphasis on the identification of individual motion characteristics from upper-limb trajectories, which can offer diagnostic insights and track personal rehabilitation progress. Lastly, AI-driven platforms for continuous patient monitoring in hospital settings are emerging, providing real-time insights into patient behavior and interactions, which can enhance patient safety and care. Notably, the integration of pre-trained models with fine-tuning on specific datasets and the public availability of these datasets are fostering innovation and reproducibility in the field.