Feasibility of New Method for the Prediction of Clinical Trial Results Using Compressive Sensing of Artificial Intelligence


  • Yasunari Miyagi Representative, Medical Data Labo, 289-48 Yamasaki, Naka Ward, Okayama City, Okayama Prefecture 703-8267, Japan Director, Department of Gynecology, Miyake Ofuku Clinic, 393-1 Ofuku, Minami ward, Okayama City, Okayama Prefecture 701-0204, Japan
  • Keiichi Fujiwara Department of Gynecologic Oncology, Saitama Medical University International Medical Center, Hidaka city, Saitama prefecture, Japan, Department of Obstetrics and Gynecology, International University of Health and Welfare, Narita city, Chiba prefecture, Japan
  • Hisanaga Nomura Department of Data Science / Pharmacy, National Cancer Center Hospital East, Kashiwa City, Chiba prefecture, Japan
  • Koji Yamamoto Department of Biostatistics, Yokohama City University School of Medicine, Yokohama city, Kanagawa prefecture, Japan
  • Robert L Coleman US Oncology Research, The Woodlands, TX, USA




Clinical Trial, Artificial Intelligence, Compressive Sensing, Computer Science, Study Design


To develop a new method for interim analysis to obtain the p-values distribution profile of the log-rank test and the statistical power at various information fraction by iteratively predicting the lacking time data of censored and uncensored cases respectively for each arm at the time when the determined sample size will be achieved. A compressive sensing algorithm derived from artificial intelligence was developed for comparing two groups of clinical trial data from real-world databases of deidentified individual participant-level data. The judgement of the interim analysis was compared to the conventional method, with our method demonstrating that it was at least comparable to the conventional method. The power by our method was higher than the conditional power of the conventional method. When α-error by log-rank test was focused, our method seemed to have made the right decision earlier than the conventional method. It may be possible to expand the options of applicable methods for interim analysis. It would be helpful in establishing efficacy and futility judgement in clinical trials.




How to Cite

Miyagi, Y., Keiichi Fujiwara, Hisanaga Nomura, Koji Yamamoto, & Robert L Coleman. (2023). Feasibility of New Method for the Prediction of Clinical Trial Results Using Compressive Sensing of Artificial Intelligence. British Journal of Healthcare and Medical Research, 10(1), 237–267. https://doi.org/10.14738/bjhmr.101.14061