Clinical validation is the process that evaluates the accuracy of a medical device or a software and determines if it meets all the required clinical standards. This essentially means whether the given product is able to safely identify, measure or predict a specific clinical, physical or biological state.  

For example, clinical validation of a blood pressure monitor can be done by comparing its measurements against a trained specialist using a validated mercury sphygmomanometer and a stethoscope.

To ensure the correctness of clinical validation, it is necessary to develop a full scientific protocol, have it independently verified for accuracy, and have all the findings peer reviewed. However, the process of clinical validation for digital health technologies such as software as a medical device is not the same as the standard evaluative techniques of traditional therapeutics.

This process is basically conducted like standard scientific research. You have to define your study hypothesis and endpoints, select a suitable population by defining the inclusion and exclusion criteria, calculate its required size and choose adequate statistical and analytical methods. After gathering sufficient amount of data, you proceed with data analysis and interpretation. Based on the gathered clinical evidence you can demonstrate safety, effectiveness, and efficacy of your solution to relevant regulators. Legislatively, these regulations fall under the Food and Drug Administration (FDA) in the United States and under the Medical Device Regulation in European Union.

References:
Goldsack, Jennifer C., et al. "Verification, analytical validation, and clinical validation (V3): the foundation of determining fit-for-purpose for Biometric Monitoring Technologies (BioMeTs)." npj digital Medicine 3.1 (2020): 55.
Shah, Sachin S., and Andrew Gvozdanovic. "Digital health; what do we mean by clinical validation?." Expert Review of Medical Devices 18.sup1 (2021): 5-8.
Park, Seong Ho, Jaesoon Choi, and Jeong-Sik Byeon. "Key principles of clinical validation, device approval, and insurance coverage decisions of artificial intelligence." Korean Journal of Radiology 22.3 (2021): 442.