The availability of smartphone and wearable sensor technology is leading to a rapid accumulation of human subject data, and machine learning is emerging as a technique to map those data into clinical predictions. As machine learning algorithms are increasingly used to support clinical decision making, it is vital to reliably quantify their prediction accuracy. Cross-validation (CV) is the standard approach where the accuracy of such algorithms is evaluated on part of the data the algorithm has not seen during training. However, for this procedure to be meaningful, the relationship between the training and the validation set should mimic the relationship between the training set and the dataset expected for the clinical use. Here we compared two popular CV methods: record-wise and subject-wise. While the subject-wise method mirrors the clinically relevant use-case scenario of diagnosis in newly recruited subjects, the record-wise strategy has no such interpretation. Using both a publicly available dataset and a simulation, we found that record-wise CV often massively overestimates the prediction accuracy of the algorithms. We also conducted a systematic review of the relevant literature, and found that this overly optimistic method was used by almost half of the retrieved studies that used accelerometers, wearable sensors, or smartphones to predict clinical outcomes. As we move towards an era of machine learning-based diagnosis and treatment, using proper methods to evaluate their accuracy is crucial, as inaccurate results can mislead both clinicians and data scientists.

AIH elicits sustained increases in volitional somatic motor output in persons with chronic SCI. Thus, AIH has promise as a therapeutic tool to induce plasticity and enhance motor function in SCI patients.

Objectives:To test the hypothesis that daily acute intermittent hypoxia (dAIH) and dAIH combined with overground walking improve walking speed and endurance in persons with chronic incomplete spinal cord injury (iSCI).Methods:Nineteen subjects completed the randomized, double-blind, placebo-controlled, crossover study. Participants received 15, 90-second hypoxic exposures (dAIH, fraction of inspired oxygen [Fio2] = 0.09) or daily normoxia (dSHAM, Fio2 = 0.21) at 60-second normoxic intervals on 5 consecutive days; dAIH was given alone or combined with 30 minutes of overground walking 1 hour later. Walking speed and endurance were quantified using 10-Meter and 6-Minute Walk Tests. The trial is registered at ClinicalTrials.gov (NCT01272349).Results:dAIH improved walking speed and endurance. Ten-Meter Walk time improved with dAIH vs dSHAM after 1 day (mean difference [MD] 3.8 seconds, 95% confidence interval [CI] 1.1–6.5 seconds, p = 0.006) and 2 weeks (MD 3.8 seconds, 95% CI 0.9–6.7 seconds, p = 0.010). Six-Minute Walk distance increased with combined dAIH + walking vs dSHAM + walking after 5 days (MD 94.4 m, 95% CI 17.5–171.3 m, p = 0.017) and 1-week follow-up (MD 97.0 m, 95% CI 20.1–173.9 m, p = 0.014). dAIH + walking increased walking distance more than dAIH after 1 day (MD 67.7 m, 95% CI 1.3–134.1 m, p = 0.046), 5 days (MD 107.0 m, 95% CI 40.6–173.4 m, p = 0.002), and 1-week follow-up (MD 136.0 m, 95% CI 65.3–206.6 m, p < 0.001).Conclusions:dAIH ± walking improved walking speed and distance in persons with chronic iSCI. The impact of dAIH is enhanced by combination with walking, demonstrating that combinatorial therapies may promote greater functional benefits in persons with iSCI.Classification of evidence:This study provides Class I evidence that transient hypoxia (through measured breathing treatments), along with overground walking training, improves walking speed and endurance after iSCI.

Machine learning algorithms that use data streams captured from soft wearable sensors have the potential to automatically detect PD symptoms and inform clinicians about the progression of disease. However, these algorithms must be trained with annotated data from clinical experts who can recognize symptoms, and collecting such data are costly. Understanding how many sensors and how much labeled data are required is key to successfully deploying these models outside of the clinic. Here we recorded movement data using 6 flexible wearable sensors in 20 individuals with PD over the course of multiple clinical assessments conducted on 1 day and repeated 2 weeks later. Participants performed 13 common tasks, such as walking or typing, and a clinician rated the severity of symptoms (bradykinesia and tremor). We then trained convolutional neural networks and statistical ensembles to detect whether a segment of movement showed signs of bradykinesia or tremor based on data from tasks performed by other individuals. Our results show that a single wearable sensor on the back of the hand is sufficient for detecting bradykinesia and tremor in the upper extremities, whereas using sensors on both sides does not improve performance. Increasing the amount of training data by adding other individuals can lead to improved performance, but repeating assessments with the same individuals—even at different medication states—does not substantially improve detection across days. Our results suggest that PD symptoms can be detected during a variety of activities and are best modeled by a dataset incorporating many individuals.