Buscarini Leonardo, Romano Paola, Cocco Elena Sofia, Damiani Carlo, Pournajaf Sanaz, Franceschini Marco, Infarinato Francesco
Rehabilitation Bioengineering Laboratory, IRCCS San Raffaele Roma, 00166, Rome, Italy.
Neurorehabilitation and Robotic Rehabilitation, Department of Neurological and Rehabilitative Sciences, IRCCS San Raffaele Roma, 00166, Rome, Italy.
J Neuroeng Rehabil. 2025 May 26;22(1):117. doi: 10.1186/s12984-025-01654-4.
In recent years, the fusion of the medical and computer science domains has gained significant traction in the scientific research landscape. Progress in both fields has enabled the generation of a vast amount of data used for making predictions and identifying interesting clusters and pathways. The Machine Learning (ML) model's application in the medical domain is one of the most compelling and challenging topics to explore, bridging the gap between Artificial Intelligence (AI) and healthcare. The combination of AI and medical information offers the possibility to create tools that can benefit both healthcare providers and physicians. This enables the enhancement of rehabilitation therapy and patient care. In the rehabilitation context, this work provides an alternative perspective: prediction of patients' home discharge upon completing the rehabilitation protocol. Demographic and clinical data were collected on 7282 inpatients from electronic Medical Record, each record was categorized into Neurological Patients (NP, N = 3222) or Orthopedic Patients (OP, N = 4060). To identify the most suitable machine learning model, an extensive data preprocessing phase was conducted. This process involved variables recoding, scaling, and the evaluation of different dataset balancing methods to optimize model performance. Following a thorough review and comparison of algorithms commonly employed in the clinical-rehabilitative field, the Random Over Sampling (ROS) technique, in combination with the Random Forest (RF) machine learning model, was selected. Subsequently, a comprehensive hyperparameter tuning phase was performed using a grid search approach. The optimized model achieved an average accuracy of 98% for OP and 96% for NP, based on 10-fold cross-validation applied to the balanced training set (unrealistic scenario). When tested on the unbalanced dataset (real-world condition), the RF model maintained strong generalization performance, achieving 90% accuracy for OP and 83% for NP. This work points out the increasing importance of AI in medicine, especially in the realm of personalized rehabilitation. The use of such approaches could signify a transformative shift in healthcare. The integration of machine learning not only enhances the precision of treatment but also opens new possibilities for patient-centered care, improving outcomes and quality of care for individuals undergoing rehabilitation.
近年来，医学领域与计算机科学领域的融合在科研领域获得了显著的关注。两个领域的进展使得能够生成大量用于预测以及识别有趣的聚类和通路的数据。机器学习（ML）模型在医学领域的应用是最引人关注且具有挑战性的探索主题之一，它弥合了人工智能（AI）与医疗保健之间的差距。AI与医学信息的结合提供了创建对医疗保健提供者和医生都有益的工具的可能性。这有助于加强康复治疗和患者护理。在康复背景下，这项工作提供了一个不同的视角：预测患者完成康复方案后的出院情况。从电子病历中收集了7282名住院患者的人口统计学和临床数据，每条记录被分类为神经科患者（NP，N = 3222）或骨科患者（OP，N = 4060）。为了确定最合适的机器学习模型，进行了广泛的数据预处理阶段。这个过程包括变量重新编码、缩放以及评估不同的数据集平衡方法以优化模型性能。在对临床康复领域常用的算法进行全面审查和比较之后，选择了随机过采样（ROS）技术与随机森林（RF）机器学习模型相结合的方法。随后，使用网格搜索方法进行了全面的超参数调整阶段。基于对平衡训练集应用10折交叉验证（不现实的情况），优化后的模型对于OP的平均准确率达到98%，对于NP为96%。在不平衡数据集（现实情况）上进行测试时，RF模型保持了很强的泛化性能，对于OP的准确率达到90%，对于NP为83%。这项工作指出了AI在医学中日益增长的重要性，特别是在个性化康复领域。使用此类方法可能意味着医疗保健领域的变革性转变。机器学习的整合不仅提高了治疗的精度，还为以患者为中心的护理开辟了新的可能性，改善了接受康复治疗的个体的治疗效果和护理质量。
