XNef Know the Brain, Understanding by AI and Change the mind.

XNef is a startup company established in 2017, which develops clinical applications
based upon more than 20 years of research and development to diagnose and treat psychiatric disorders.




We are pleased to announce that we got the business license of Second Class Marketing Authorization Holder on April 10, 2023.
We will make a further development of various machine-learning softwares, so called “Brainalyzer” series, for supporting the diagnosis of not only major depressive disorder.


We are pleased to announce that we applied for Medical Devices Reviews to use machine-learning software as a medical device for supporting the diagnosis of major depressive disorder to PMDA (Pharmaceuticals and Medical Devices Agency) on March 30, 2023, and it was accepted.
We will make a further announcement when the application is approved through the review process.


Our advertisement feature was published on 9 March 2023, as part of Nature Index Japan in the online version of Nature, a weekly international journal publishing the finest peer-reviewed research in science and technology.
Supporting depression diagnosis through machine learning, XNef


Machine Learning Identifies New Brain Network Signature of Major Depression in collaboration with ATR and Hiroshima University.(PLoS Biology)
XNef is working for practical clinical application of this brain network marker for major depressive disorder (MDD) that is useful at any imaging site.


Research findings about PTSD neurofeedback were published in the Nikkei Sangyo Shimbun.


A collaboration with KDDI about smartphone dependence was announced in the Nikkei Sangyo Shimbun.


The article about "Rewriting the memory of fear: a way to treat PTSD" was published in the Asashi Shimbun.

XNef's Business

XNef develops new diagnostic methods for psychiatric disorders.

Neuropsychiatric disorders create the greatest societal burden of all diseases. For individuals afflicted with these diseases, this burden is quantified in terms of disability-adjusted life years, the number of years of life that the patient loses due to morbidity, disability, and premature death. However, the burden extends well beyond the individuals afflicted with these diseases. Among neuropsychiatric disorders, depression causes the most significant social loss (*1). In economic terms, in Japan that loss amounts to 2 trillion yen annually (*2).

Counterintuitively, despite the cost of depression and other diseases, development of treatments for neuropsychiatric disorders has been stagnant. Partly, this is because diagnosis of neuropsychiatric disorders is based solely on subjective assessments of patient symptoms. As a result, diagnostic accuracy for some neuropsychiatric disorders is quite low.

XNef employs a standard MRI machine to visualize brain activity with fMRI (*3). fMRI images are used to measure functional connectivity between brain regions. Artificial intelligence (AI) technology is used to analyze this information to produce brain network markers (biomarkers) that serve as objective indicators for diagnosis of psychiatric disorders such as depression.

XNef supports development of drugs to treat psychiatric disorders.

Treatment of psychiatric disorders has been hindered by the fact that patients with potentially different neurobiological factors are often given the same diagnosis. As a result, it is impossible to optimize treatment based on individual diagnoses or to conduct appropriate clinical development using homogenous groups.

However, using the "brain network markers" that XNef generates, it is possible to identify different subtypes of patients (stratification of patients) among those currently grouped with the same diagnosis. Once the clinical significance of each subtype is known, it will become possible to develop therapeutic agents for each type.

XNef develops advanced treatments for psychiatric disorders.

Psychiatric disorders are thought to be disorders of brain networks (*4). Even though XNef’s "brain network markers" can be used to evaluate a patient's condition, drug therapy cannot necessarily normalize the brain network.

XNef will develop a neurofeedback treatment device that can directly restore the brain network to a healthy state, using advanced neurofeedback technology (*5).

  • (*1) Hay et al., The Lancet, 2017; Kassebaum et al., The Lancet, 2016
  • (*2) Sado et al., Psychiatry and Clinical Neurosciences, 2011
  • (*3) functional magnetic resonance imaging
  • (*4) Insel & Cuthbert, Science, 2015
  • (*5) Shibata et al, Science, 2011; Koizumi et al., Nat Hum Behav, 2016; Watanabe et al., Trends Cog Sci, 2017