ABSTRACT
In recent years, variational quantum algorithms such as the Quantum Approximation Optimization Algorithm (QAOA) have gained popularity as they provide the hope of using NISQ devices to tackle hard combinatorial optimization problems. It is, however, known that at low depth, certain locality constraints of QAOA limit its performance. To go beyond these limitations, a non-local variant of QAOA, namely recursive QAOA (RQAOA), was proposed to improve the quality of approximate solutions. The RQAOA has been studied comparatively less than QAOA, and it is less understood, for instance, for what family of instances it may fail to provide high-quality solutions. However, as we are tackling NP-hard problems (specifically, the Ising spin model), it is expected that RQAOA does fail, raising the question of designing even better quantum algorithms for combinatorial optimization. In this spirit, we identify and analyze cases where (depth-1) RQAOA fails and, based on this, propose a reinforcement learning enhanced RQAOA variant (RL-RQAOA) that improves upon RQAOA. We show that the performance of RL-RQAOA improves over RQAOA: RL-RQAOA is strictly better on these identified instances where RQAOA underperforms and is similarly performing on instances where RQAOA is near-optimal. Our work exemplifies the potentially beneficial synergy between reinforcement learning and quantum (inspired) optimization in the design of new, even better heuristics for complex problems.
ABSTRACT
BACKGROUND: Numerous antineoplastic drugs such as chemotherapeutics have cardiotoxic side effects and can lead to long QT syndrome (LQTS). When diagnosed and treated in time, the potentially fatal outcomes of LQTS can be prevented. Therefore, regular electrocardiogram (ECG) assessments are critical to ensure patient safety. However, these assessments are associated with patient discomfort and require timely support of the attending oncologist by a cardiologist. OBJECTIVE: This study aimed to examine whether this approach can be made more efficient and comfortable by a smartphone app (QTc Tracker), supporting single-lead ECG records on site and transferring to a tele-cardiologist for an immediate diagnosis. METHODS: To evaluate the QTc Tracker, it was implemented in 54 cancer centers in Germany. In total, 266 corrected QT interval (QTc) diagnoses of 122 patients were recorded. Moreover, a questionnaire on routine ECG workflow, turnaround time, and satisfaction (1=best, 6=worst) was answered by the centers before and after the implementation of the QTc Tracker. RESULTS: Compared to the routine ECG workflow, the QTc Tracker enabled a substantial turnaround time reduction of 98% (mean 2.67, 95% CI 1.72-2.67 h) and even further time efficiency in combination with a cardiologic on-call service (mean 12.10, 95% CI 5.67-18.67 min). Additionally, nurses and patients reported higher satisfaction when using the QTc Tracker. In particular, patients' satisfaction sharply improved from 2.59 (95% CI 2.41-2.88) for the routine ECG workflow to 1.25 (95% CI 0.99-1.51) for the QTc Tracker workflow. CONCLUSIONS: These results reveal a significant improvement regarding reduced turnaround time and increased user satisfaction. Best patient care might be guaranteed as the exposure of patients with an uncontrolled risk of QTc prolongations can be avoided by using the fast and easy QTc Tracker. In particular, as regular side-effect monitoring, the QTc Tracker app promises more convenience for patients and their physicians. Finally, future studies are needed to empirically test the usability and validity of such mobile ECG assessment methods. TRIAL REGISTRATION: ClinicalTrials.gov NCT04055493; https://classic.clinicaltrials.gov/ct2/show/NCT04055493.
