Youth with severe motor and communication impairments (SMCI) encounter challenges in expressing emotions, often relying on familiar interpreters for assistance. This may introduce bias in self-reports of emotion. The study contributes to the intersection of assistive technology and emotional communication, offering a promising avenue to improve the quality of life for youth with SMCI. The algorithmic approach not only addresses interpreter limitations but also introduces possibilities for enhancing emotion communication and understanding experiences. While investigating the use of physiologic signal data in the detection of emotions experienced by those with SMCI, multiple challenges were identified by the research team. The current work details the challenges that arose during recruitment of participants, data collection, and analysis of physiological signals in youth with SMCI. This work can inform future inclusive data collection and algorithmic approaches. Integrating the proposed recommendations in this paper aims to assist in the development of AI systems that are more inclusive of persons with disabilities. The recommendations address various phases of the research process, including tailored recruitment strategies, improvement of data collection methodologies, and identification of analysis tools that account for differences in SMCI participants relative to the general population.