In recent years, advances in artificial intelligence (AI) technology have led to the rapid clinical implementation of devices with AI technology in the medical field. More than 60 AI-equipped medical devices have already been approved by the Food and Drug Administration (FDA) in the United States, and the active introduction of AI technology is considered to be an inevitable trend in the future of medicine. In the field of oncology, clinical applications of medical devices using AI technology are already underway, mainly in radiology, and AI technology is expected to be positioned as an important core technology. In particular, “precision medicine,” a medical treatment that selects the most appropriate treatment for each patient based on a vast amount of medical data such as genome information, has become a worldwide trend; AI technology is expected to be utilized in the process of extracting truly useful information from a large amount of medical data and applying it to diagnosis and treatment. In this review, we would like to introduce the history of AI technology and the current state of medical AI, especially in the oncology field, as well as discuss the possibilities and challenges of AI technology in the medical field.

Glioblastoma is one of the most devastating human malignancies for which a novel efficient treatment is urgently required. This pre–clinical study shows that eribulin, a specific inhibitor of telomerase reverse transcriptase (TERT)‐RNA‐dependent RNA polymerase, is an effective anticancer agent against glioblastoma. Eribulin inhibited the growth of 4 TERT promoter mutation‐harboring glioblastoma cell lines in vitro at subnanomolar concentrations. In addition, it suppressed the growth of glioblastoma cells transplanted subcutaneously or intracerebrally into mice, and significantly prolonged the survival of mice harboring brain tumors at a clinically equivalent dose. A pharmacokinetics study showed that eribulin quickly penetrated brain tumors and remained at a high concentration even when it was washed away from plasma, kidney or liver 24 hours after intravenous injection. Moreover, a matrix‐assisted laser desorption/ionization mass spectrometry imaging analysis revealed that intraperitoneally injected eribulin penetrated the brain tumor and was distributed evenly within the tumor mass at 1 hour after the injection whereas only very low levels of eribulin were detected in surrounding normal brain. Eribulin is an FDA‐approved drug for refractory breast cancer and can be safely repositioned for treatment of glioblastoma patients. Thus, our results suggest that eribulin may serve as a novel therapeutic option for glioblastoma. Based on these data, an investigator‐initiated registration‐directed clinical trial to evaluate the safety and efficacy of eribulin in patients with recurrent GBM (UMIN000030359) has been initiated.

Machine learning models for automated magnetic resonance image segmentation may be useful in aiding glioma detection. However, the image differences among facilities cause performance degradation and impede detection. This study proposes a method to solve this issue. We used the data from the Multimodal Brain Tumor Image Segmentation Benchmark (BraTS) and the Japanese cohort (JC) datasets. Three models for tumor segmentation are developed. In our methodology, the BraTS and JC models are trained on the BraTS and JC datasets, respectively, whereas the fine-tuning models are developed from the BraTS model and fine-tuned using the JC dataset. Our results show that the Dice coefficient score of the JC model for the test portion of the JC dataset was 0.779 ± 0.137, whereas that of the BraTS model was lower (0.717 ± 0.207). The mean Dice coefficient score of the fine-tuning model was 0.769 ± 0.138. There was a significant difference between the BraTS and JC models (p < 0.0001) and the BraTS and fine-tuning models (p = 0.002); however, no significant difference between the JC and fine-tuning models (p = 0.673). As our fine-tuning method requires fewer than 20 cases, this method is useful even in a facility where the number of glioma cases is small.

In this study, we focus on participants called lurkers, who do not post any messages in an online community such as interactive mailing lists and bulletin board systems. We propose a method of classifying participants including lurkers based on two criteria: what types of actions they take outside the online community, and whether or not the online community affects their thoughts. In addition, based on the results of interviews, we propose hypotheses regarding factors that characterize the categories of lurkers. We conduct a questionnaire survey of all participants in two in-house online communities to verify our method and test the hypotheses. There are a considerable number of lurkers who have a strong and wide influence outside the online community. We conclude that such lurkers cannot be neglected in an evaluation of online communities within a company. We also discuss the possibility of online community management by focusing on not only posters but also lurkers who are indirect contributors to increasing the influence of an online community on its outside environment.

In medical imaging, the characteristics purely derived from a disease should reflect the extent to which abnormal findings deviate from the normal features. Indeed, physicians often need corresponding images without abnormal findings of interest or, conversely, images that contain similar abnormal findings regardless of normal anatomical context. This is called comparative diagnostic reading of medical images, which is essential for a correct diagnosis. To support comparative diagnostic reading, content-based image retrieval (CBIR) that can selectively utilize normal and abnormal features in medical images as two separable semantic components will be useful. In this study, we propose a neural network architecture to decompose the semantic components of medical images into two latent codes: normal anatomy code and abnormal anatomy code. The normal anatomy code represents counterfactual normal anatomies that should have existed if the sample is healthy, whereas the abnormal anatomy code attributes to abnormal changes that reflect deviation from the normal baseline. By calculating the similarity based on either normal or abnormal anatomy codes or the combination of the two codes, our algorithm can retrieve images according to the selected semantic component from a dataset consisting of brain magnetic resonance images of gliomas. Moreover, it can utilize a synthetic query vector combining normal and abnormal anatomy codes from two different query images. To evaluate whether the retrieved images are acquired according to the targeted semantic component, the overlap of the ground-truth labels is calculated as metrics of the semantic consistency. Our algorithm provides a flexible CBIR framework by handling the decomposed features with qualitatively and quantitatively remarkable results.

Glioblastomas (GBMs) often acquire resistance against temozolomide (TMZ) after continuous treatment and recur as TMZ-resistant GBM (TMZ-R-GBM). Lomustine (CCNU) and nimustine (ACNU), which were previously used as standard therapeutic agents against GBMs before TMZ, have occasionally been employed for the salvage therapy of TMZ-R-GBMs; however, their efficacy has not yet been thoroughly examined. We therefore investigated the antitumor effects of CCNU and ACNU against TMZ-R-GBMs. As a model of TMZ-R-GBMs, TMZ resistant clones of human GBM cell lines (U87, U251MG, and U343MG) were established (TMZ-R-cells) by the culture of each GBM cells under continuous TMZ treatment, and the antitumor effects of TMZ, CCNU or ACNU against these cells were analyzed in vitro and in vivo. As a result, although growth arrest and apoptosis were triggered in all TMZ-R-cells after the administration of each drug, the antitumor effects of TMZ against TMZ-R-cells were significantly reduced as compared to those of parental cells, whereas CCNU and ACNU demonstrated efficient antitumor effects on TMZ-R-cells as well as parental cells. In addition, it was also demonstrated that TMZ resistance of TMZ-R-cells was regulated at the level of DNA damage response initiation. Furthermore, survival in mice was significantly prolonged by systemic treatment with CCNU or ACNU but not TMZ after implantation of TMZ-R-cells. These findings suggest that CCNU or ACNU may serve as a therapeutic agent in salvage treatment against TMZ-R-GBMs.