Fundraise

Xijia Medical raises tens of millions of yuan for AI brain-mapping tech

What's the deal? Chinese brain-science AI company Xijia MedicalDealroom has a profile for this one. Try Dealroom → has raised tens of millions of yuan in an early-stage round led by Guozhong Capital, with participation from Xi Venture CapitalDealroom has a profile for this one. Try Dealroom →, Qinzhi CapitalDealroom has a profile for this one. Try Dealroom →, Huichuang Medical, and Xingfu CapitalDealroom has a profile for this one. Try Dealroom →. The funding will go toward clinical research, market expansion, and product iteration.

What's the endgame? Founded in 2019, Xijia uses AI and cloud computing to analyse MRI data and generate individualised brain-network maps for patients. Its core software, NuraTome, builds a 379-region atlas — combining structural and functional connectivity — to help doctors locate targets in the brain.

Why now? In May 2025, Xijia disclosed a tens-of-millions-yuan angel round from Fengqi QiyunDealroom has a profile for this one. Try Dealroom →. Chief financial officer Yuan Quan said the company had "made the business model work" in 2025 and aims to standardise and scale in 2026.

The product: Beyond its atlas software, Xijia has developed a transcranial magnetic stimulation (TMS) navigation device that integrates targeting and guidance into one system. Yuan compared the setup to shooting: "TMS is the gun, the navigation device is the scope, and NuraTome identifies the high-value target." For depression treatment, the company says its precision-targeting approach improved patient symptom scores markedly against baseline. Traditional anatomy-based TMS targeting can be effective as little as 20% to 30% of the time.

Beyond the brain map: Xijia has expanded into brain-computer interfaces, positioning NuraTome as upstream navigation infrastructure for electrode implantation. The company says it has supported more than 20 clinical implants, working with several leading brain-computer interface firms on a per-case fee model.

The technology traces to a 2016 study by Matthew F. Glasser and colleagues, who divided the cortex into 360 regions using data from the Human Connectome Project. But that atlas maps an "average" brain, Yuan noted, and cannot be applied directly to individual patients whose anatomy varies or whose tissue is displaced by tumours. Xijia's surgical-planning software has now been used in nearly 20,000 neurosurgical procedures across more than 180 medical institutions worldwide.

The signal: As brain-computer interfaces move toward clinical use, precise, patient-specific brain mapping is emerging as critical infrastructure. Xijia is betting that owning the "map" layer — from surgery to neuromodulation to implants — gives it a defensible position across multiple neuroscience markets.

Read more: 36kr.com

Image credit: foltzwerk

More top stories