Deep Learning-Based Functional Independence Measure Score Prediction After Stroke in Kaifukuki (Convalescent) Rehabilitation Ward Annexed to Acute Care Hospital

Abstract

IntroductionPrediction models of functional independent measure (FIM) score after kaifukuki (convalescent) rehabilitation ward (KRW) are needed to decide the treatment strategies and save medical resources. Statistical models were reported, but their accuracies were not satisfactory. We made such prediction models using the deep learning (DL) framework, Prediction One (Sony Network Communications Inc., Tokyo, Japan). MethodsOf the 559 consecutive stroke patients, 122 patients were transferred to our KRW. We divided our 122 patients' data randomly into halves of training and validation datasets. Prediction One made three prediction models from the training dataset using (1) variables at the acute care ward admission, (2) those at the KRW admission, and (3) those combined (1) and (2). The models' determination coefficients (R 2 ), correlation coefficients (rs), and residuals were calculated using the validation dataset. ResultsOf the 122 patients, the median age was 71, length of stay (LOS) in acute care ward 23 (17-30) days, LOS in KRW 53 days, total FIM scores at the admission of KRW 85, those at discharge 108. The mean FIM gain and FIM efficiency were 19 and 0.417. All patients were discharged home. Model (1), (2), and (3)'s R 2 were 0.794, 0.970, and 0.972. Their mean residuals between the predicted and actual total FIM scores were -1.56±24.6, -4.49±17.1, and -2.69±15.7. ConclusionOur FIM gain and efficiency were better than national averages of FIM gain 17.1 and FIM efficiency 0.187. We made DL-based total FIM score prediction models, and their accuracies were superior to those of previous statistically calculated ones. The DL-based FIM score prediction models would save medical costs and perform efficient stroke and rehabilitation medicine.

Can AI predict walking independence in patients with stroke upon admission to a recovery-phase rehabilitation ward?

“…Practical examples of the use of Prediction One in financial institutions presented an AUC of 85% or higher [16]. The prediction model in this study had an AUC of 91.7%, which we considered to be highly accurate, similar to previous studies employing Prediction One [13][14][15]. For the cutoff values of the endpoints used to determine walking independence in patients with stroke hospitalized in recovery wards, Kitaji [10] reported that for patients with a first stroke, the berg balance scale (BBS) had an AUC of 97.9% with a cutoff of 45.5 points, and the timed up and go test (TUG) had an AUC of 97.6% with a cutoff of 15.6 seconds in the maximum walking speed condition.…”

“…Although the process of building a prediction model is difficult to track because of the nature of machine learning, Prediction One shows the "prediction contribution," which indicates the degree to which each item contributes to the prediction results. In analyses using Prediction One, the coefficient of determination for the model predicting independence at discharge from a recovery unit on admission was 0.972 [13], that for independence at 6 months after surgery in patients with cerebral hemorrhage was 0.997 [14], and that for independence at 6 months after surgery for subarachnoid hemorrhage was 0.994 [15]. All of these models are reportedly highly accurate, and their use in predicting outcomes is highly anticipated.…”

“…The author considers that it is necessary to continue working on outcome predictions from various perspectives and methods. Various reports on the use of "Prediction One" (Sony Network Communications Corporation, Tokyo, Japan) in consequence prediction have been published [13][14][15]. Prediction One is an AI predictive analytics software that automatically adjusts and standardizes variables, building optimal predictive models from multiple onboard algorithms.…”

Machine learning predicts improvement of functional outcomes in traumatic brain injury patients after inpatient rehabilitation

“…The Functional Independence Measure (FIM) is a widely accepted, 18-item motor and cognitive function score that originated from the American Academy of Physical Medicine and Rehabilitation and the American Congress of Rehabilitation Medicine ( 27 , 28 ). The FIM is calculated as a composite score based on direct observation by a multidisciplinary rehabilitation team that documents the patient's disability and level of assistance required to perform activities of daily living ( 29 ). Each motor and cognitive item earns an ordinal numerical score ranging from 1 to 7.…”