Evaluating psychological anxiety in patients receiving radiation therapy using smartwatch.

PURPOSE: Patients undergoing radiation therapy (RT) often experience psychological anxiety that manifests as muscle contraction. Our study explored psychological anxiety in these patients by using biological signals recorded using a smartwatch. MATERIALS AND METHODS: Informed consent was obtained from participating patients prior to the initiation of RT. The patients wore a smartwatch from the waiting room until the conclusion of the treatment. The smartwatch acquired data related to heart rate features (average, minimum, and maximum) and stress score features (average, minimum, and maximum). On the first day of treatment, we analyzed the participants' heart rates and stress scores before and during the treatment. The acquired data were categorized according to sex and age. For patients with more than three days of data, we observed trends in heart rate during treatment relative to heart rate before treatment (HRtb) over the course of treatment. Statistical analyses were performed using the Wilcoxon signed-rank test and paired t-test. RESULTS: Twenty-nine individuals participated in the study, of which 17 had more than 3 days of data. During treatment, all patients exhibited elevated heart rates and stress scores, particularly those in the younger groups. The HRtb levels decreased as treatment progresses. CONCLUSION: Patients undergoing RT experience notable psychological anxiety, which tends to diminish as the treatment progresses. Early stage interventions are crucial to alleviate patient anxiety during RT.

The Prediction of Stress in Radiation Therapy: Integrating Artificial Intelligence with Biological Signals.

This study aimed to predict stress in patients using artificial intelligence (AI) from biological signals and verify the effect of stress on respiratory irregularity. We measured 123 cases in 41 patients and calculated stress scores with seven stress-related features derived from heart-rate variability. The distribution and trends of stress scores across the treatment period were analyzed. Before-treatment information was used to predict the stress features during treatment. AI models included both non-pretrained (decision tree, random forest, support vector machine, long short-term memory (LSTM), and transformer) and pretrained (ChatGPT) models. Performance was evaluated using 10-fold cross-validation, exact match ratio, accuracy, recall, precision, and F1 score. Respiratory irregularities were calculated in phase and amplitude and analyzed for correlation with stress score. Over 90% of the patients experienced stress during radiation therapy. LSTM and prompt engineering GPT4.0 had the highest accuracy (feature classification, LSTM: 0.703, GPT4.0: 0.659; stress classification, LSTM: 0.846, GPT4.0: 0.769). A 10% increase in stress score was associated with a 0.286 higher phase irregularity (p < 0.025). Our research pioneers the use of AI and biological signals for stress prediction in patients undergoing radiation therapy, potentially identifying those needing psychological support and suggesting methods to improve radiotherapy effectiveness through stress management.

Clinical applicability of deep learning-based respiratory signal prediction models for four-dimensional radiation therapy.

For accurate respiration gated radiation therapy, compensation for the beam latency of the beam control system is necessary. Therefore, we evaluate deep learning models for predicting patient respiration signals and investigate their clinical feasibility. Herein, long short-term memory (LSTM), bidirectional LSTM (Bi-LSTM), and the Transformer are evaluated. Among the 540 respiration signals, 60 signals are used as test data. Each of the remaining 480 signals was spilt into training and validation data in a 7:3 ratio. A total of 1000 ms of the signal sequence (Ts) is entered to the models, and the signal at 500 ms afterward (Pt) is predicted (standard training condition). The accuracy measures are: (1) root mean square error (RMSE) and Pearson correlation coefficient (CC), (2) accuracy dependency on Ts and Pt, (3) respiratory pattern dependency, and (4) error for 30% and 70% of the respiration gating for a 5 mm tumor motion for latencies of 300, 500, and 700 ms. Under standard conditions, the Transformer model exhibits the highest accuracy with an RMSE and CC of 0.1554 and 0.9768, respectively. An increase in Ts improves accuracy, whereas an increase in Pt decreases accuracy. An evaluation of the regularity of the respiratory signals reveals that the lowest predictive accuracy is achieved with irregular amplitude patterns. For 30% and 70% of the phases, the average error of the three models is <1.4 mm for a latency of 500 ms and >2.0 mm for a latency of 700 ms. The prediction accuracy of the Transformer is superior to LSTM and Bi-LSTM. Thus, the three models have clinically applicable accuracies for a latency <500 ms for 10 mm of regular tumor motion. The clinical acceptability of the deep learning models depends on the inherent latency and the strategy for reducing the irregularity of respiration.

Neutrophil-to-lymphocyte ratio and carbohydrate antigen 19-9 as prognostic markers for advanced pancreatic cancer patients receiving first-line chemotherapy.

BACKGROUND: A decline in serum carbohydrate antigen 19-9 (CA19-9) levels during systemic chemotherapy is considered as a prognostic marker for patients with advanced pancreatic cancer. Neutrophil-to-lymphocyte ratio (NLR) has been extensively studied as a simple and useful indicator of prognosis in various cancers including pancreatic cancer. AIM: To assess the prognostic significance of NLR and CA19-9 in patients with advanced pancreatic adenocarcinoma received first-line chemotherapy according to CA19-9 positivity. METHODS: We retrospectively analyzed patients diagnosed with advanced pancreatic cancer who received first-line chemotherapy between January 2010 and July 2017 at the Catholic University of Seoul St. Mary's Hospital. Patients were divided according to CA19-9 positivity (CA19-9-positive vs -negative groups) and pre-and post-treatment NLR levels. To determine cut-off value of NLR and CA19-9 reduction, time-dependent receiver operating characteristic curve was applied. We evaluated overall survival (OS) and progression-free survival (PFS) for each group using Kaplan-Meier method, and we performed multivariate analyses on the entire cohort. RESULTS: We included 271 patients in this study. Cut-off value of NLR and CA19-9 reduction was determined as 2.62 and 18%. Multivariate analysis showed that post-treatment NLR < 2.62 and reduction of ≥ 18% of baseline CA19-9 were significantly associated with OS and PFS. Post-treatment NLR ≥ 2.62 showed hazard ratio (HR) of 2.47 [95% confidence interval (CI): 1.84-3.32, P < 0.001] and CA19-9 decline (≥ 18%) showed HR of 0.51 (95%CI: 0.39-0.67, P < 0.001) for OS. When CA19-9-positive patients were divided into groups according to CA19-9 response (responder vs non-responder) and post-treatment NLR (< 2.62 vs ≥ 2.62), CA19-9 responder and post-treatment NLR < 2.62 group showed better survival than CA19-9 non-responder and post-treatment NLR ≥ 2.62 group (OS 11.0 mo vs 3.9 mo, P < 0.001; PFS 6.3 mo vs 2.0 mo, P < 0.001). The combination of CA19-9 decline and post-treatment NLR showed a significant correlation with clinical response in CA 19-9 positive group. Within the CA19-9-negative group, the post-treatment NLR < 2.62 group showed better survival than the post-treatment NLR ≥ 2.62 group (OS 12.7 mo vs 7.7 mo, P < 0.001; PFS 6.7 mo vs 2.1 mo, P < 0.001), and post-treatment NLR showed correlation with clinical response. CONCLUSION: In advanced pancreatic cancer patients positive for CA19-9 and treated with systemic chemotherapy, the combination of post-treatment NLR < 2.62 and 18% decline of CA19-9 at the first response evaluation is a good prognostic marker. Post-treatment NLR < 2.62 alone could be used as a prognostic marker and an adjunctive tool for response evaluation in CA19-9-negative patients.