microRNAs—tiny molecules deeply involved in the onset and progression of cancer—have long been a focus of medical research. In 2024, their importance was globally recognized when they were featured in the Nobel Prize in Physiology or Medicine.
However, detecting microRNAs has historically been a major challenge. They exist in only trace amounts in body fluids and are extremely fragile, making them difficult to capture and analyze with precision.
The key to overcoming this challenge lies in exosomes—microscopic capsules secreted by cells. These exosomes act as protective carriers, shielding microRNAs from degradation and preserving them in a stable state.
Craif has developed a proprietary technology that efficiently isolates these exosomes from urine and analyzes the microRNAs they contain. This innovation enables the accurate detection of trace microRNAs that were previously difficult to detect in sufficient quantities, delivering highly reliable and reproducible data to support early disease detection.
At Craif, we have developed a highly accurate Bio-AI for predicting cancer risk, based on tens of thousands of test samples analyzed using next-generation sequencing (NGS).
As of April 2025, this AI is capable of detecting 10 different types of cancer: esophageal, gastric, pancreatic, lung, colorectal, ovarian, breast, kidney, bladder, and prostate cancers. Notably, for early-stage pancreatic cancer, the AI has demonstrated world-leading performance, a sensitivity of 93% (correctly identifying cancer) as published in a peer-reviewed scientific journal.*
Based on this large-scale data, we have developed AI-powered algorithms that analyze these patterns to predict an individual’s risk of cancer.
In collaboration with Hokkaido University Hospital and Iwanai Kyokai Hospital, Craif conducted a real-world study in Iwanai Town, Hokkaido, with the support of local residents.
In this study, 100 healthy residents underwent screening using miSignal Scan, leading to the successful detection of stage 0 lung cancer (carcinoma in situ) in one participant.
Detecting cancer before symptoms appear has long been a major challenge in clinical practice. This achievement marks an important milestone toward the real-world implementation of miSignal Scan. By working hand in hand with local communities and healthcare institutions, we are beginning to make early cancer detection not just a possibility, but a reality.
Craif is developing a novel medical device that supports the diagnosis of pancreatic cancer by combining urinary microRNA profiling with advanced AI (artificial intelligence) analysis.
This device calculates a risk score based on the expression levels and patterns of microRNAs and other biomarkers found in urine, and classifies the result as either “positive” or “negative” to assess the likelihood of pancreatic cancer.
Similar to tumor marker tests and abdominal ultrasound, this diagnostic tool is intended for individuals with risk factors for pancreatic cancer. It serves as a decision-support tool to determine whether further diagnostic testing is warranted.
The device is currently undergoing a pivotal clinical trial in collaboration with multiple medical institutions across Japan to evaluate its effectiveness, and Craif is preparing to file for manufacturing and marketing approval.
This innovative technology aims to become a new option for the early detection and appropriate diagnosis of pancreatic cancer.
In today's health checkups and comprehensive medical exams, early cancer detection requires combining different tests for different parts of the body. This places significant physical and financial burdens on patients, while also requiring a substantial time commitment—and even then, comprehensive coverage of all cancer types remains difficult.
To address this unmet need, Craif is developing technology that can detect multiple cancers simultaneously from urine alone. Small RNAs found in urine—a class of biomolecules that includes microRNAs—carry rich biological information reflecting the state of cells throughout the body. By analyzing these small RNAs using Craif's proprietary Bio-AI technology platform, we aim to enable detailed assessment of multiple cancer types by organ from a single urine sample, helping physicians make clinical decisions as quickly as possible.
miSignal Dx Multi-Cancer is being developed with the goal of obtaining regulatory approval as software as a medical device (SaMD) for use in medical institutions at physicians’ discretion.
We are currently working to establish detection performance across at least 10 cancer types, advancing the collection and analysis of clinical data as well as the development of a quality management system.
a test that can cover
more than 95%
of all cancers
*Excluding rare cancers
