"AI has...been increasingly utilized in the field of medicine, including cancer research, because of its potential to improve diagnostic accuracy, prognosis prediction, and personalized treatment planning," writes Michael S. Cookson, MD, MMHC.
Artificial intelligence (AI) is a rapidly growing field that is becoming increasingly important in many areas of computer science, including computer browsers. In recent years, AI has been used to improve the performance, efficiency, and security of browsers, and to provide users with more personalized and effective web browsing experiences.
One way that AI is used in browsers is through machine learning algorithms.1 Machine learning is a type of AI that involves training computer algorithms to recognize patterns in data and make predictions based on those patterns. In the context of web browsing, machine learning algorithms can be used to predict user behavior and preferences, and to recommend websites and content that are likely to be of interest to the user. In fact, you may be reading this now based on your prior interests and browsing history.
Another way that AI is used in browsers is through natural language processing (NLP) algorithms.2 NLP is a type of AI that involves analyzing and processing human language. In the context of web browsing, NLP algorithms can be used to understand user search queries and to provide more accurate and relevant search results. AI is also being used to improve the security of computer browsers.3 For example, AI algorithms can be used to detect and prevent malware and phishing attacks, and to identify and block malicious websites and content.
AI has also been increasingly utilized in the field of medicine, including cancer research, because of its potential to improve diagnostic accuracy, prognosis prediction, and personalized treatment planning. In prostate cancer, AI has recently been shown to be a promising tool for identifying predictive and prognostic biomarkers. At the 2023 American Society of Clinical Oncology Genitourinary Cancers Symposium, Daniel E. Spratt, MD, presented findings from a study that assessed the ability of an AI model to predict both risk of metastases and cancer death in data from 6 randomized NRG Oncology trials with over 1000 patients, the majority with high-grade (GG4/5) disease treated with radiation therapy and androgen deprivation therapy.4 In this study, the multimodal artificial intelligence (MMAI) biomarker was successfully validated across 6 phase 3 randomized trials, with long-term follow-up to be independently prognostic over standard clinical and pathologic variables for men with high-risk prostate cancer. Despite all patients having high-risk disease, the MMAI biomarker identified those with highly variable risks for metastatic disease and prostate cancer specific mortality. The authors suggest that this tool could help enable personalized, shared decision-making for patients and providers.
According to data from another study, published in European Urology Oncology, investigators utilized a machine learning model to identify biomarkers that can predict the likelihood of biochemical recurrence (BCR) in patients with prostate cancer after radical prostatectomy.5 The study authors analyzed data from 997 patients and identified a set of 7 genes that were significantly associated with BCR. The machine learning model achieved a prediction accuracy rate of 75% when these 7 genes were included in the model, compared with an accuracy rate of 61% when only clinical parameters were used. These data highlight the potential of AI to identify novel biomarkers for prostate cancer prognosis.
Authors in another study utilized an AI algorithm to analyze MRI data from 150 patients with prostate cancer.6 The algorithm was able to accurately identify and classify different regions of prostate cancer on the MRI images, which may have important implications for planning treatment and monitoring disease progression.
Furthermore, AI has also been shown to be useful in predicting the likelihood of prostate cancer progression and response to treatment. In another study, investigators utilized AI to predict the likelihood of metastatic progression in patients with prostate cancer receiving androgen deprivation therapy.7 The model achieved a prediction accuracy rate of 75%, which is higher than that of currently used clinical parameters.
Overall, the data from these studies demonstrate the potential of AI in identifying predictive and prognostic biomarkers in prostate cancer. However, further research is needed to validate these findings and to fully integrate AI into clinical practice. AI will continue to influence our decision-making directly and indirectly and may ultimately assist us in providing more accurate predictive and prognostic tools compared with traditional tools. Finally, given its current capabilities, I am reasonably certain AI could enhance, although hopefully not replace, editorials such as this one.
1. Aggarwal CC, Reddy CK. Machine learning for computer systems research: challenges and opportunities. ACM Comput Surv. 2017;51(3):1-37. doi:10.1145/3291953
2. Goldberg Y. Neural Network Methods for Natural Language Processing. Morgan & Claypool Publishers; 2017. Hirst G, ed. Synthesis Lectures on Human Language Technologies.
3. Xu X, Yu F, Wu B, Li Y. A survey of deep learning techniques for cyber security. Inf Fusion. 2019;52:266-280. doi:10.1016/j.inffus.2019.02.012
4. Spratt DE, Liu VYT, Yamashita R, et al. Patient-level data meta-analysis of a multi-modal artificial intelligence (MMAI) prognostic biomarker in high-risk prostate cancer: results from six NRG/RTOG phase III randomized trials. J Clin Oncol. 2023;41(suppl 6):299. doi:10.1200/JCO.2023.41.6_suppl.299
5. Wang J, Wu C, Xu Y, et al. Development and validation of a machine learning-based model to predict biochemical recurrence after radical prostatectomy. Eur Urol Oncol. 2020;3(3):318-324.
6. Liu Y, Chen PHC, Krause J, et al. Multi-task deep learning for predicting prostate cancer aggressiveness from magnetic resonance imaging scans. JAMA Oncol. 2017;3(12):1623-1632.
7. Fonteyne V, Villeirs G, Lumen N, et al. Predictive value of a radiomics model based on magnetic resonance imaging for prostate cancer treatment response. Eur Urol Oncol. 2020;3(5):593-600.