Objectives This study aimed to develop a classification model to detect and distinguish apathy and depression based on text, audio, and video features and to make use of the shapely additive explanations (SHAP) toolkit to increase the model interpretability. Methods Subjective scales and objective experiments were conducted on 319 mild cognitive impairment (MCI) patients to measure apathy and depression. The MCI patients were classified into four groups, depression only, apathy only, depressed‐apathetic, and the normal group. Speech, facial and text features were extracted using the open‐source data analysis toolkits. Multiclass classification and SHAP toolkits were used to develop a classification model and explain the contribution of specific features. Results The macro‐averaged f1 score and accuracy for overall model were 0.91 and 0.90, respectively. The accuracy for the apathetic, depressed, depressed‐apathetic, and normal groups were 0.98, 0.88, 0.93, and 0.82, respectively. The SHAP toolkit identified speech features (Mel‐frequency cepstral coefficient (MFCC) 4, spectral slopes, F0, F1), facial features (action unit (AU) 14, 26, 28, 45), and text feature (text 6 semantic) associated with apathy. Meanwhile, speech features (spectral slopes, shimmer, F0) and facial expression (AU 2, 6, 7, 10, 14, 26, 45) were associated with depression. Apart from the shared features mentioned above, new speech (MFCC 2, loudness) and facial (AU 9) features were observed in the depressive‐apathetic group. Conclusions Apathy and depression shared some verbal and facial features while also exhibited distinct features. A combination of text, audio, and video could be used to improve the early detection and differential diagnosis of apathy and depression in MCI patients.

No abstract

No abstract

UNSTRUCTURED Surgery is still far from being completely safe and reliable. Surgical safety has, therefore, been the focus of considerable attention over the last few decades, and there are a growing number of national drives to improve it. There are also a number of large surgical complication reporting systems and system-based interventions, both of which have made remarkable progress in the past two decades. These systems, however, have either mainly focused on reporting complications and played a limited role in guiding practice or have provided nonselective interventions to all patients, perhaps imposing unnecessary burdens on frontline medical staff. We have, therefore, developed an evidence-based stratified surgical safety information system based on a multicenter surgical safety improvement program. This study discusses some critical issues in the process of developing this information system, including (1) decisions about data gathering, (2) establishing and sharing knowledge, (3) developing functions for the system, (4) system implementation, and (5) evaluation and continuous improvement. Using examples drawn from the surgical safety improvement program, we have shown how this type of system can be fitted into day-to-day clinical practice and how it can guide medical practice by incorporating inherent patient-related risk and providing tailored interventions for patients with different levels of risk. We concluded that multidisciplinary collaboration, involving experts in health care (including senior staff in surgery, nursing, and anesthesia), data science, health care management, and health information technology, can help build an evidence-based stratified surgical patient safety improvement system. This can provide an information-intensified surgical safety learning platform and, therefore, benefit surgical patients by delivering tailored interventions and an integrated workflow.