System and method for determining subject conditions in mobile health clinical trials

1. A system for predicting a state of a subject in a clinical trial using a predictive model, the system comprising:

a processor having a memory, the memory storing instructions executable by the processor to perform:

receiving data during time intervals from a plurality of mobile health sensors, including sensors within a single device or multiple devices;

receiving a diary that records subject state determinations;

integrating the data from the mobile health sensors and the subject state determinations from the diary;

cleaning the integrated data;

determining which variables optimally predict the subject state;

partitioning the integrated and cleaned data using a combination of ranges of values for the variables that optimally predict the subject state;

generating a training set of data comprising a portion of the integrated, cleaned, and partitioned data;

training the predictive model using the training set and a machine learning algorithm; and

generating, using the processor, the predictive model of the state of the subject of the clinical trial based on the training set; and

using the predictive model to determine, using the processor, subject state for data without the diary that records said subject state determinations, wherein the subject state comprises a digital bio-marker for a disease condition, and wherein to determine the ranges of the values for the variables that optimally predict the subject state, the memory further comprises instructions executable by the processor to perform:

(a) for a first variable X 1 , using a range of data, X 1 >A, where A is a defined value, to classify the data according to a first state Y 1 and a second state Y 2 ;

(b) computing a ratio between Y 1 and Y 2 , or between Y 1 and Y 1 +Y 2 , for data where X 1 is not greater than A and for data where X 1 is greater than A, respectively; and

iteratively repeating (a) and (b) for selected values of A to maximize said ratio between Y 1 and Y 2 , or between Y 1 and Y 1 +Y 2 .

2. The system of claim 1 , wherein the memory further comprises instructions executable by the processor to perform:

classifying a combination of ranges of value for variables received from the mobile health sensors; and

testing a second portion of the integrated and cleaned data using the predictive model and comparing the predicted subject states against the subject state determinations from the diary that records said subject state determinations.

3. The system of claim 1 , wherein the memory further comprises instructions executable by the processor to perform:

integrating new data that are input;

cleaning the integrated new data; and

receiving the predictive model and the cleaned data and determining a subject state.

4. The system of claim 1 , wherein the plurality of mobile health sensors are within a single mobile health device.

5. The system of claim 1 , wherein the plurality of mobile health sensors are within multiple mobile health devices.

6. The system of claim 1 , wherein the memory further comprises instructions executable by the processor to perform receiving data from a user in addition to the plurality of sensors.

7. The system of claim 1 , wherein the memory further comprises instructions executable by the processor to perform calculating time intervals in which to collect an optimal amount of data to predict a state of a subject in a clinical trial.

8. The system of claim 1 , wherein the disease condition is an arrhythmia.

9. The system of claim 1 , wherein the memory further comprises instructions executable by the processor to perform:

(c) for a second variable X 2 , using a range of data, X 2 >F, where F is a defined value, to classify the data according to the first state Y 1 and the second state Y 2 ;

(d) computing a ratio between Y 1 and Y 2 , or between Y 1 and Y 1 +Y 2 , for data where X 2 is not greater than F and for data where X 2 is greater than F, respectively; and

iteratively repeating (c) and (d) for selected values of F to maximize said ratio between Y 1 and Y 2 , or between Y 1 and Y 1 +Y 2 .

10. The system of claim 1 , wherein the predictive model represents subject state as a function of multiple variables X n , which represent clinical data, the variables X n including at least a plurality of: heart rate, electrocardiogram, body temperature, skin temperature, movement, and blood pressure.

11. The system of claim 2 , wherein to generate the training set of data comprising the portion of the integrated, cleaned, and partitioned data, the memory further comprises instructions executable by the processor to perform randomly separating the integrated, cleaned, and partitioned data into said portion and said second portion based on a predetermined percentage.