Department of Biomedical Engineering, Faculty of Electrical Engineering, K. N. Toosi University of Technology, Tehran, Iran.
Department of Mechatronics, Faculty of Electrical Engineering, K. N. Toosi University of Technology, Tehran, Iran.
Sci Rep. 2024 Apr 10;14(1):8424. doi: 10.1038/s41598-024-59179-0.
Using deep learning has demonstrated significant potential in making informed decisions based on clinical evidence. In this study, we deal with optimizing medication and quantitatively present the role of deep learning in predicting the medication dosage for patients with Parkinson's disease (PD). The proposed method is based on recurrent neural networks (RNNs) and tries to predict the dosage of five critical medication types for PD, including levodopa, dopamine agonists, monoamine oxidase-B inhibitors, catechol-O-methyltransferase inhibitors, and amantadine. Recurrent neural networks have memory blocks that retain crucial information from previous patient visits. This feature is helpful for patients with PD, as the neurologist can refer to the patient's previous state and the prescribed medication to make informed decisions. We employed data from the Parkinson's Progression Markers Initiative. The dataset included information on the Unified Parkinson's Disease Rating Scale, Activities of Daily Living, Hoehn and Yahr scale, demographic details, and medication use logs for each patient. We evaluated several models, such as multi-layer perceptron (MLP), Simple-RNN, long short-term memory (LSTM), and gated recurrent units (GRU). Our analysis found that recurrent neural networks (LSTM and GRU) performed the best. More specifically, when using LSTM, we were able to predict levodopa and dopamine agonist dosage with a mean squared error of 0.009 and 0.003, mean absolute error of 0.062 and 0.030, root mean square error of 0.099 and 0.053, and R-squared of 0.514 and 0.711, respectively.
深度学习在基于临床证据做出明智决策方面具有巨大潜力。在这项研究中，我们致力于优化药物治疗，并定量展示深度学习在预测帕金森病（PD）患者药物剂量方面的作用。所提出的方法基于递归神经网络（RNN），并尝试预测五种关键 PD 药物类型的剂量，包括左旋多巴、多巴胺激动剂、单胺氧化酶-B 抑制剂、儿茶酚-O-甲基转移酶抑制剂和金刚烷胺。递归神经网络具有记忆块，可以保留来自之前患者就诊的关键信息。这一特性对 PD 患者很有帮助，因为神经科医生可以参考患者之前的状态和所开药物来做出明智的决策。我们使用了帕金森进展标志物倡议的数据。该数据集包括每位患者的统一帕金森病评定量表、日常生活活动、Hoehn 和 Yahr 量表、人口统计学细节和药物使用日志信息。我们评估了几种模型，如多层感知器（MLP）、Simple-RNN、长短期记忆（LSTM）和门控循环单元（GRU）。我们的分析发现，递归神经网络（LSTM 和 GRU）表现最佳。具体来说，当使用 LSTM 时，我们能够以均方误差 0.009 和 0.003、平均绝对误差 0.062 和 0.030、均方根误差 0.099 和 0.053 以及 R 方 0.514 和 0.711 来预测左旋多巴和多巴胺激动剂的剂量。
