Classifying individuals using finite mixture markov modelling and test trials with accounting for item position
Methods, systems, and apparatus, including medium-encoded computer program products, for analyzing data include: receiving data including responses, and lack thereof, for items of a cognitive test including multiple item-recall trials; processing the data using a stochastic model of a cognitive process, in which a conditional probability distribution of future states of the cognitive process depend upon a present state; and encoding a result of the processing on a non-transitory computer-readable medium for use in an assessment related to cognition; wherein the processing using the stochastic model includes representing recall or recognition of an item in the multiple item-recall trials using distinct cognitive states; and wherein the processing using the stochastic model includes adjusting separate memory storage and retrieval parameters for each of the distinct cognitive states in the modeled cognitive process to account for position of the items in each respective trial of the multiple item-recall trials.
1. A computer-implemented method comprising:
receiving data comprising responses, and lack thereof, for items of a cognitive test, wherein the cognitive test comprises test trials;
processing the data using a Finite Mixture Markov (FMM) model of a cognitive process, wherein the FMM model of the cognitive process
(i) represents recall, identification or recognition of an item in at least one of the test trials using distinct cognitive states, and
(ii) adjusts separate memory storage and retrieval parameters for each of the distinct cognitive states to account for position of the items in each respective trial of the test trials; and
classifying individuals into one of two groups that differ by a given biomarker using results of the processing the data using the FMM model of the cognitive process.
2. The computer-implemented method of claim 1 , wherein the distinct cognitive states comprise three cognitive states being an unlearned state, an intermediate state, and a learned state.
3. The computer-implemented method of claim 2 , wherein each of the memory storage and retrieval parameters have an assigned subscript corresponding to an item's position in a trial.
4. The computer-implemented method of claim 3 , wherein the test trials are at least three trials, at least two items of the at least three trials are free to be placed in different list positions in separate administrations of individual trials of the at least three trials, and respective ones of the subscripts correspond to an item's position in respective ones of the at least three trials.
5. The computer-implemented method of claim 2 , wherein processing the data using the FMM model comprises computing a probability of a given item's response pattern using a set of all possible cognitive state sequences except for cognitive state sequences that transition from the learned state to the intermediate state, from the learned state to the unlearned state, or from the intermediate state to the unlearned state.
6. The computer-implemented method of claim 2 , wherein the given biomarker is cerebrospinal fluid.
7. The computer-implemented method of claim 2 , wherein the given biomarker is amyloid.
8. The computer-implemented method of claim 2 , wherein classifying the individuals comprises measuring disease progression.
9. The computer-implemented method of claim 2 , wherein classifying the individuals comprises measuring treatment effects.
10. A system comprising:
a user device; and
one or more computers operable to interact with the user device, and the one or more computers being configured to (i) receive data comprising responses, and lack thereof, for items of a cognitive test, wherein the cognitive test comprises test trials, (ii) process the data using a Finite Mixture Markov (FMM) model of a cognitive process, wherein the FMM model of the cognitive process (a) represents recall, identification or recognition of an item in at least one of the test trials using distinct cognitive states, and (b) adjusts separate memory storage and retrieval parameters for each of the distinct cognitive states to account for position of the items in each respective trial of the test trials, and (iii) classify individuals into one of two groups that differ by a given biomarker using results of the processing the data using the FMM model of the cognitive process.
11. The system of claim 10 , wherein the distinct cognitive states comprise three cognitive states being an unlearned state, an intermediate state, and a learned state.
12. The system of claim 11 , wherein each of the memory storage and retrieval parameters have an assigned subscript corresponding to an item's position in a trial.
13. The system of claim 12 , wherein the test trials are at least three trials, at least two items of the at least three trials are free to be placed in different list positions in separate administrations of individual trials of the at least three trials, and respective ones of the subscripts correspond to an item's position in respective ones of the at least three trials.
14. The system of claim 11 , wherein the one or more computers are configured to compute a probability of a given item's response pattern using a set of all possible cognitive state sequences except for cognitive state sequences that transition from the learned state to the intermediate state, from the learned state to the unlearned state, or from the intermediate state to the unlearned state.
15. The system of claim 11 , wherein the given biomarker is cerebrospinal fluid.
16. The system of claim 11 , wherein the given biomarker is amyloid.
17. The system of claim 11 , wherein the one or more computers are configured to measure disease progression.
18. The system of claim 11 , wherein the one or more computers are configured to measure treatment effects.
19. A non-transitory computer-readable medium encoding a computer program product operable to cause data processing apparatus to perform operations comprising:
receiving data comprising responses, and lack thereof, for items of a cognitive test, wherein the cognitive test comprises test trials;
processing the data using a Finite Mixture Markov (FMM) model of a cognitive process, wherein the FMM model of the cognitive process
(i) represents recall, identification or recognition of an item in at least one of the test trials using distinct cognitive states, and
(ii) adjusts separate memory storage and retrieval parameters for each of the distinct cognitive states to account for position of the items in each respective trial of the test trials; and
classifying individuals into one of two groups that differ by a given biomarker using results of the processing the data using the FMM model of the cognitive process.
20. The non-transitory computer-readable medium of claim 19 , wherein the distinct cognitive states comprise three cognitive states being an unlearned state, an intermediate state, and a learned state.
21. The non-transitory computer-readable medium of claim 20 , wherein each of the memory storage and retrieval parameters have an assigned subscript corresponding to an item's position in a trial.
22. The non-transitory computer-readable medium of claim 21 , wherein the test trials are at least three trials, at least two items of the at least three trials are free to be placed in different list positions in separate administrations of individual trials of the at least three trials, and respective ones of the subscripts correspond to an item's position in respective ones of the at least three trials.
23. The non-transitory computer-readable medium of claim 20 , wherein processing the data using the FMM model comprises computing a probability of a given item's response pattern using a set of all possible cognitive state sequences except for cognitive state sequences that transition from the learned state to the intermediate state, from the learned state to the unlearned state, or from the intermediate state to the unlearned state.
24. The non-transitory computer-readable medium of claim 20 , wherein the given biomarker is cerebrospinal fluid.
25. The non-transitory computer-readable medium of claim 20 , wherein the given biomarker is amyloid.
26. The non-transitory computer-readable medium of claim 20 , wherein classifying the individuals comprises measuring disease progression.
27. The non-transitory computer-readable medium of claim 20 , wherein classifying the individuals comprises measuring treatment effects.