Thanks in great part to its explosive expansion, artificial intelligence (AI) is increasingly seen in healthcare, particularly in portable medical devices. Using AI models in these sorts of gadgets is rather difficult, however, largely due to technological concerns like memory, processing capability, and energy economy. This study investigates how to create compact and light enough artificial intelligence models fit for portable medical equipment. These models would permit real-time monitoring and health diagnosis without compromising speed or accuracy. Our major objective is to simplify and shrink present artificial intelligence models by means of model trimming, quantisation, information distillation, and hardware-specific optimisation techniques. These methods retain artificial intelligence systems’ capacity to forecast the future and increase their efficiency. For environments with limited resources, like portable devices, this makes them ideal. We also discuss the advantages and drawbacks of computation demands, accuracy, and model size. At last, we demonstrate the many medical environments personal health monitors, diagnostic instruments, point-of-care devices where these models find application. Since these systems often have to follow rigorous guidelines imposed by authorities, we also consider how difficult it is to maintain data secure and private when employing artificial intelligence models in medical equipment. We demonstrate how a lightweight artificial intelligence model may be used in a portable ECG monitor to detect cardiac illness using a case study. This indicates in real-life clinical environments that these models may provide quick, accurate, and consistent findings. Ultimately, this research reveals how lightweight artificial intelligence models might transform portable medical equipment, hence simplifying the usage of modern medical technologies in many clinical and distant environments.