Table 2.

Inputs and outputs of identified services (services are mobile or Web, depending on the running device, and their outputs are typically the inputs of the next service).

Service	Type	Description	Inputs	Outputs
Accelerometer data acquisition	Mobile	Responsible for accelerometer data gathering and storage at run time, when elderly people perform a specific gait and balance test. This also includes mobile communication between the smartphone and the accelerometer sensor.	Accelerometer signal	Accelerometer values in x,y,z axes
Accelerometer data processing	Mobile	Responsible for accelerometer data handling through data filtering and segmentation as well as calculation of accelerometry indicators.	Accelerometer values (x,y,z axes)	Accelerometry indicators (dispersion measures)
Patient record extraction	Web	Defines the mechanisms to obtain frailty risk factors from the patient record. The use of clinical standards could be necessary.	Patient record, accelerometry indicators	Frailty risk factors
Frailty study procedure	Web	Responsible for performing a comparison between frailty risk factors from the elderly patient studied and each of the patients stored in the database (known as patient stack).	Frailty risk factors, patient stack	Frailty assessment
Setting up a built result	Web	Parse the comparison results in a formal language, easily readable by the mobile phone.	Frailty assessment	Frailty assessment formalized
Visualization of frailty assessment	Mobile	Defines the method for frailty result preparation and visualization on the smartphone screen, after receiving data from the server.	Frailty assessment formalized	Information, tips, and charts for the physician
Storage into Patient Stack	Web	Stores the new patient data in the patient stack structure, increasing the patient stack size and improving the accuracy of frailty assessments in the future.	Risk factors from a new patient	Patient stack with new patient