Prospects and challenges for clinical decision support in the era of big data

Artificial Intelligence, Big Data, and Regulation of Immunity: Challenges and Opportunities

The immune system is regulated by a complex set of genetic, molecular, and cellular interactions. Rapid advances in the study of immunity and its network of interactions have been boosted by a spectrum of "omics" technologies that have generated huge amounts of data that have reached the status of big data (BD). With recent developments in artificial intelligence (AI), theoretical and clinical breakthroughs could emerge. Analyses of large data sets with AI tools will allow the formulation of new testable hypotheses open new research avenues and provide innovative strategies for regulating immunity and treating immunological diseases. This includes diagnosis and identification of rare diseases, prevention and treatment of autoimmune diseases, allergic disorders, infectious diseases, metabolomic disorders, cancer, and organ transplantation. However, ethical and regulatory challenges remain as to how these studies will be used to advance our understanding of basic immunology and how immunity might be regulated in health and disease. This will be particularly important for entities in which the complexity of interactions occurring at the same time and multiple cellular pathways have eluded conventional approaches to understanding and treatment. The analyses of BD by AI are likely to be complicated as both positive and negative outcomes of regulating immunity may have important ethical ramifications that need to be considered. We suggest there is an immediate need to develop guidelines as to how the analyses of immunological BD by AI tools should guide immune-based interventions to treat various diseases, prevent infections, and maintain health within an ethical framework.

Assessing the decision quality of artificial intelligence and oncologists of different experience in different regions in breast cancer treatment

Background

AI-based clinical decision support system (CDSS) has important prospects in overcoming the current informational challenges that cancer diseases faced, promoting the homogeneous development of standardized treatment among different geographical regions, and reforming the medical model. However, there are still a lack of relevant indicators to comprehensively assess its decision-making quality and clinical impact, which greatly limits the development of its clinical research and clinical application. This study aims to develop and application an assessment system that can comprehensively assess the decision-making quality and clinical impacts of physicians and CDSS.

Methods

Enrolled adjuvant treatment decision stage early breast cancer cases were randomly assigned to different decision-making physician panels (each panel consisted of three different seniority physicians in different grades hospitals), each physician made an independent "Initial Decision" and then reviewed the CDSS report online and made a "Final Decision". In addition, the CDSS and guideline expert groups independently review all cases and generate "CDSS Recommendations" and "Guideline Recommendations" respectively. Based on the design framework, a multi-level multi-indicator system including "Decision Concordance", "Calibrated Concordance", " Decision Concordance with High-level Physician", "Consensus Rate", "Decision Stability", "Guideline Conformity", and "Calibrated Conformity" were constructed.

Results

531 cases containing 2124 decision points were enrolled; 27 different seniority physicians from 10 different grades hospitals have generated 6372 decision opinions before and after referring to the "CDSS Recommendations" report respectively. Overall, the calibrated decision concordance was significantly higher for CDSS and provincial-senior physicians (80.9%) than other physicians. At the same time, CDSS has a higher " decision concordance with high-level physician" (76.3%-91.5%) than all physicians. The CDSS had significantly higher guideline conformity than all decision-making physicians and less internal variation, with an overall guideline conformity variance of 17.5% (97.5% vs. 80.0%), a standard deviation variance of 6.6% (1.3% vs. 7.9%), and a mean difference variance of 7.8% (1.5% vs. 9.3%). In addition, provincial-middle seniority physicians had the highest decision stability (54.5%). The overall consensus rate among physicians was 64.2%.

Conclusions

There are significant internal variation in the standardization treatment level of different seniority physicians in different geographical regions in the adjuvant treatment of early breast cancer. CDSS has a higher standardization treatment level than all physicians and has the potential to provide immediate decision support to physicians and have a positive impact on standardizing physicians' treatment behaviors.

A clinical decision support system for AI-assisted decision-making in response-adaptive radiotherapy (ARCliDS)

Involvement of many variables, uncertainty in treatment response, and inter-patient heterogeneity challenge objective decision-making in dynamic treatment regime (DTR) in oncology. Advanced machine learning analytics in conjunction with information-rich dense multi-omics data have the ability to overcome such challenges. We have developed a comprehensive artificial intelligence (AI)-based optimal decision-making framework for assisting oncologists in DTR. In this work, we demonstrate the proposed framework to Knowledge Based Response-Adaptive Radiotherapy (KBR-ART) applications by developing an interactive software tool entitled Adaptive Radiotherapy Clinical Decision Support (ARCliDS). ARCliDS is composed of two main components: Artifcial RT Environment (ARTE) and Optimal Decision Maker (ODM). ARTE is designed as a Markov decision process and modeled via supervised learning. Given a patient's pre- and during-treatment information, ARTE can estimate treatment outcomes for a selected daily dosage value (radiation fraction size). ODM is formulated using reinforcement learning and is trained on ARTE. ODM can recommend optimal daily dosage adjustments to maximize the tumor local control probability and minimize the side effects. Graph Neural Networks (GNN) are applied to exploit the inter-feature relationships for improved modeling performance and a novel double GNN architecture is designed to avoid nonphysical treatment response. Datasets of size 117 and 292 were available from two clinical trials on adaptive RT in non-small cell lung cancer (NSCLC) patients and adaptive stereotactic body RT (SBRT) in hepatocellular carcinoma (HCC) patients, respectively. For training and validation, dense data with 297 features were available for 67 NSCLC patients and 110 features for 71 HCC patients. To increase the sample size for ODM training, we applied Generative Adversarial Networks to generate 10,000 synthetic patients. The ODM was trained on the synthetic patients and validated on the original dataset. We found that, Double GNN architecture was able to correct the nonphysical dose-response trend and improve ARCliDS recommendation. The average root mean squared difference (RMSD) between ARCliDS recommendation and reported clinical decisions using double GNNs were 0.61 [0.03] Gy/frac (mean [sem]) for adaptive RT in NSCLC patients and 2.96 [0.42] Gy/frac for adaptive SBRT HCC compared to the single GNN's RMSDs of 0.97 [0.12] Gy/frac and 4.75 [0.16] Gy/frac, respectively. Overall, For NSCLC and HCC, ARCliDS with double GNNs was able to reproduce 36% and 50% of the good clinical decisions (local control and no side effects) and improve 74% and 30% of the bad clinical decisions, respectively. In conclusion, ARCliDS is the first web-based software dedicated to assist KBR-ART with multi-omics data. ARCliDS can learn from the reported clinical decisions and facilitate AI-assisted clinical decision-making for improving the outcomes in DTR.

Precision radiotherapy via information integration of expert human knowledge and AI recommendation to optimize clinical decision making

In the precision medicine era, there is a growing need for precision radiotherapy where the planned radiation dose needs to be optimally determined by considering a myriad of patient-specific information in order to ensure treatment efficacy. Existing artificial-intelligence (AI) methods can recommend radiation dose prescriptions within the scope of this available information. However, treating physicians may not fully entrust the AI's recommended prescriptions due to known limitations or at instances when the AI recommendation may go beyond physicians' current knowledge. This paper lays out a systematic method to integrate expert human knowledge with AI recommendations for optimizing clinical decision making. Towards this goal, Gaussian process (GP) models are integrated with deep neural networks (DNNs) to quantify the uncertainty of the treatment outcomes given by physicians and AI recommendations, respectively, which are further used as a guideline to educate clinical physicians and improve AI models performance. The proposed method is demonstrated in a comprehensive dataset where patient-specific information and treatment outcomes are prospectively collected during radiotherapy of 67 non-small cell lung cancer (NSCLC) patients and are retrospectively analyzed.

Can artificial intelligence improve cancer treatments?

Artificial intelligence (AI) powered by the accumulating clinical and molecular data about cancer has fueled the expectation that a transformation in cancer treatments towards significant improvement of patient outcomes is at hand. However, such transformation has been so far elusive. The opacity of AI algorithms and the lack of quality annotated data being available at population scale are among the challenges to the application of AI in oncology. Fundamentally however, the heterogeneity of cancer and its evolutionary dynamics make every tumor response to therapy sufficiently different from the population, machine-learned statistical models, challenging hence the capacity of these models to yield reliable inferences about treatment recommendations that can improve patient outcomes. This article reviews the nominal elements of clinical decision-making for precision oncology and frames the utility of AI to cancer treatment improvements in light of cancer unique challenges.

Quantum deep reinforcement learning for clinical decision support in oncology: application to adaptive radiotherapy

Subtle differences in a patient's genetics and physiology may alter radiotherapy (RT) treatment responses, motivating the need for a more personalized treatment plan. Accordingly, we have developed a novel quantum deep reinforcement learning (qDRL) framework for clinical decision support that can estimate an individual patient's dose response mid-treatment and recommend an optimal dose adjustment. Our framework considers patients' specific information including biological, physical, genetic, clinical, and dosimetric factors. Recognizing that physicians must make decisions amidst uncertainty in RT treatment outcomes, we employed indeterministic quantum states to represent human decision making in a real-life scenario. We paired quantum decision states with a model-based deep q-learning algorithm to optimize the clinical decision-making process in RT. We trained our proposed qDRL framework on an institutional dataset of 67 stage III non-small cell lung cancer (NSCLC) patients treated on prospective adaptive protocols and independently validated our framework in an external multi-institutional dataset of 174 NSCLC patients. For a comprehensive evaluation, we compared three frameworks: DRL, qDRL trained in a Qiskit quantum computing simulator, and qDRL trained in an IBM quantum computer. Two metrics were considered to evaluate our framework: (1) similarity score, defined as the root mean square error between retrospective clinical decisions and the AI recommendations, and (2) self-evaluation scheme that compares retrospective clinical decisions and AI recommendations based on the improvement in the observed clinical outcomes. Our analysis shows that our framework, which takes into consideration individual patient dose response in its decision-making, can potentially improve clinical RT decision-making by at least about 10% compared to unaided clinical practice. Further validation of our novel quantitative approach in a prospective study will provide a necessary framework for improving the standard of care in personalized RT.

Artificial intelligence in oncology: Path to implementation

In recent years, the field of artificial intelligence (AI) in oncology has grown exponentially. AI solutions have been developed to tackle a variety of cancer‐related challenges. Medical institutions, hospital systems, and technology companies are developing AI tools aimed at supporting clinical decision making, increasing access to cancer care, and improving clinical efficiency while delivering safe, high‐value oncology care. AI in oncology has demonstrated accurate technical performance in image analysis, predictive analytics, and precision oncology delivery. Yet, adoption of AI tools is not widespread, and the impact of AI on patient outcomes remains uncertain. Major barriers for AI implementation in oncology include biased and heterogeneous data, data management and collection burdens, a lack of standardized research reporting, insufficient clinical validation, workflow and user‐design challenges, outdated regulatory and legal frameworks, and dynamic knowledge and data. Concrete actions that major stakeholders can take to overcome barriers to AI implementation in oncology include training and educating the oncology workforce in AI; standardizing data, model validation methods, and legal and safety regulations; funding and conducting future research; and developing, studying, and deploying AI tools through multidisciplinary collaboration.

Impact of a machine learning based decision support for Urinary Tract Infections: Prospective observational study in 36 primary care practices (Preprint)

Background

There is increasing attention on machine learning (ML)-based clinical decision support systems (CDSS), but their added value and pitfalls are very rarely evaluated in clinical practice. We implemented a CDSS to aid general practitioners (GPs) in treating patients with urinary tract infections (UTIs), which are a significant health burden worldwide.

Objective

This study aims to prospectively assess the impact of this CDSS on treatment success and change in antibiotic prescription behavior of the physician. In doing so, we hope to identify drivers and obstacles that positively impact the quality of health care practice with ML.

Methods

The CDSS was developed by Pacmed, Nivel, and Leiden University Medical Center (LUMC). The CDSS presents the expected outcomes of treatments, using interpretable decision trees as ML classifiers. Treatment success was defined as a subsequent period of 28 days during which no new antibiotic treatment for UTI was needed. In this prospective observational study, 36 primary care practices used the software for 4 months. Furthermore, 29 control practices were identified using propensity score-matching. All analyses were performed using electronic health records from the Nivel Primary Care Database. Patients for whom the software was used were identified in the Nivel database by sequential matching using CDSS use data. We compared the proportion of successful treatments before and during the study within the treatment arm. The same analysis was performed for the control practices and the patient subgroup the software was definitely used for. All analyses, including that of physicians’ prescription behavior, were statistically tested using 2-sided z tests with an α level of .05.

Results

In the treatment practices, 4998 observations were included before and 3422 observations (of 2423 unique patients) were included during the implementation period. In the control practices, 5044 observations were included before and 3360 observations were included during the implementation period. The proportion of successful treatments increased significantly from 75% to 80% in treatment practices (z=5.47, P<.001). No significant difference was detected in control practices (76% before and 76% during the pilot, z=0.02; P=.98). Of the 2423 patients, we identified 734 (30.29%) in the CDSS use database in the Nivel database. For these patients, the proportion of successful treatments during the study was 83%—a statistically significant difference, with 75% of successful treatments before the study in the treatment practices (z=4.95; P<.001).

Conclusions

The introduction of the CDSS as an intervention in the 36 treatment practices was associated with a statistically significant improvement in treatment success. We excluded temporal effects and validated the results with the subgroup analysis in patients for whom we were certain that the software was used. This study shows important strengths and points of attention for the development and implementation of an ML-based CDSS in clinical practice.

Trial Registration

ClinicalTrials.gov NCT04408976; https://clinicaltrials.gov/ct2/show/NCT04408976

Global updates in the treatment of gastric cancer: a systematic review. Part 2: perioperative management, multimodal therapies, new technologies, standardization of the surgical treatment and educational aspects

Gastric cancer is the fifth malignancy and the third cause of cancer death worldwide, according to the global cancer statistics presented in 2018. Its definition and staging have been revised in the eight edition of the AJCC/TNM classification, which took effect in 2018. Novel molecular classifications for GC have been recently established and the process of translating these classifications into clinical practice is ongoing. The cornerstone of GC treatment is surgical, in a context of multimodal therapy. Surgical treatment is being standardized, and is evolving according to new anatomical concepts and to the recent technological developments. This is leading to a massive improvement in the use of mini-invasive techniques. Mini-invasive techniques aim to be equivalent to open surgery from an oncologic point of view, with better short-term outcomes. The persecution of better short-term outcomes also includes the optimization of the perioperative management, which is being implemented on large scale according to the enhanced recovery after surgery principles. In the era of precision medicine, multimodal treatment is also evolving. The long-time-awaited results of many trials investigating the role for preoperative and postoperative management have been published, changing the clinical practice. Novel investigations focused both on traditional chemotherapeutic regimens and targeted therapies are currently ongoing. Modern platforms increase the possibility for further standardization of the different treatments, promote the use of big data and open new possibilities for surgical learning. This systematic review in two parts assesses all the current updates in GC treatment.

Emotions as Social Information in Shared Decision-Making in Oncology

Emotions play an important role in decision-making and they can impact individual as well as shared decisions. With increasing complexity of the decision, the potential for emotions to influence the outcome increases. Emotions are thus an influential factor in oncological decision-making which is a complex and high-stakes situation. As the shared decision-making process is at the center of patient-centric decisions, we model emotions as social information that inform the shared decision-making process. We present and explain a range of emotional concepts, together with a specific clinical example, that can impact the shared decision-making process. Our process model shows that emotions are experienced in various combinations before, during, and after a shared decision is made and how patients' and physicians' emotions interact and spill over during a shared decision situation. Overall, our process model and specific example show how emotions can impact shared decision-making in oncology in a multitude of ways. With this paper, we want to raise awareness of the role of emotions in the shared decision-making process, as emotions are often not explicitly recognized as decision criteria. Increased awareness of emotions may help their optimal utilization and reduce their influence as a bias in shared decision-making.

Balancing accuracy and interpretability of machine learning approaches for radiation treatment outcomes modeling

Radiation outcomes prediction (ROP) plays an important role in personalized prescription and adaptive radiotherapy. A clinical decision may not only depend on an accurate radiation outcomes’ prediction, but also needs to be made based on an informed understanding of the relationship among patients’ characteristics, radiation response and treatment plans. As more patients’ biophysical information become available, machine learning (ML) techniques will have a great potential for improving ROP. Creating explainable ML methods is an ultimate task for clinical practice but remains a challenging one. Towards complete explainability, the interpretability of ML approaches needs to be first explored. Hence, this review focuses on the application of ML techniques for clinical adoption in radiation oncology by balancing accuracy with interpretability of the predictive model of interest. An ML algorithm can be generally classified into an interpretable (IP) or non-interpretable (NIP) (“black box”) technique. While the former may provide a clearer explanation to aid clinical decision-making, its prediction performance is generally outperformed by the latter. Therefore, great efforts and resources have been dedicated towards balancing the accuracy and the interpretability of ML approaches in ROP, but more still needs to be done. In this review, current progress to increase the accuracy for IP ML approaches is introduced, and major trends to improve the interpretability and alleviate the “black box” stigma of ML in radiation outcomes modeling are summarized. Efforts to integrate IP and NIP ML approaches to produce predictive models with higher accuracy and interpretability for ROP are also discussed.