Did we personalize? Assessing personalization by an online reinforcement learning algorithm using resampling.

There is a growing interest in using reinforcement learning (RL) to personalize sequences of treatments in digital health to support users in adopting healthier behaviors. Such sequential decision-making problems involve decisions about when to treat and how to treat based on the user's context (e.g., prior activity level, location, etc.). Online RL is a promising datadriven approach for this problem as it learns based on each user's historical responses and uses that knowledge to personalize these decisions. However, to decide whether the RL algorithm should be included in an "optimized" intervention for real-world deployment, we must assess the data evidence indicating that the RL algorithm is actually personalizing the treatments to its users. Due to the stochasticity in the RL algorithm, one may get a false impression that it is learning in certain states and using this learning to provide specific treatments. We use a working definition of personalization and introduce a resampling-based methodology for investigating whether the personalization exhibited by the RL algorithm is an artifact of the RL algorithm stochasticity. We illustrate our methodology with a case study by analyzing the data from a physical activity clinical trial called HeartSteps, which included the use of an online RL algorithm. We demonstrate how our approach enhances data-driven truth-in-advertising of algorithm personalization both across all users as well as within specific users in the study.

Anthropometry of Nonarthritic Asian Knees: Is It Time for a Race-Specific Knee Implant?

BACKGROUND: This study was done to study the anthropometry of nonarthritic Asian knees; to determine the differences in morphology between knees of different ethnicities and to compare the knee anthropometry values with sizes of available knee implants. METHODS: Magnetic resonance imaging scans of 100 nonarthritic Indian knees were analyzed. Anteroposterior (AP) length, mediolateral (ML) length, and aspect ratio of the distal femur and proximal tibia, patellar length, and patellar tendon length were measured. These values were compared with values of other ethnicities from literature. The values were also compared with sizes of available knee implants and evaluated for mismatch. RESULTS: All the parameters of female knees were significantly smaller than those of male knees (p < 0.05). The distal femur of Indian knees resembled that of Chinese knees with similar AP and ML lengths and aspect ratio. The distal femur of Indian knees had a significantly smaller AP, ML, and aspect ratio than those of Hispanic knees did. In comparison to Caucasian distal femur, Indian knees had smaller AP and ML lengths and larger aspect ratio. In terms of the proximal tibia, the Indian knees were smaller than Chinese (only ML), Caucasian (AP and ML) and Hispanic (AP and ML) knees. On comparison with implant sizes, there was a mismatch between the distal femur morphology and the dimensions of all implants. For a given AP length, the ML dimensions of all implants were smaller than the measured ML length of the knee. However, the tibial components of all the studied implants correlated well with the tibial morphology. CONCLUSIONS: Distinct anthropometric differences exist between knees of different ethnicities. The knees of females were smaller than the knees of males. In Indian knees, the ML-AP aspect ratio of the distal femur was higher than that of the currently available femoral components. These results suggest the need for race-specific knee implants.