Report Date: 2026-11-15 | Industry Stance: Cautiously Bullish (12-24 months) | Research Scope: Implantable (invasive + semi-invasive) and non-invasive analyzed separately, focusing on clinical evidence, device approval, and payment pathways
A brain-computer interface (BCI) is a technology that establishes a direct information pathway between the brain and external devices. By degree of invasiveness, this report divides three technology routes and analyzes them separately throughout:
One-sentence industry chain map: Upstream (flexible electrode materials, dedicated chips, encoding/decoding algorithms) → Midstream (implant systems and surgical robots, e.g., Neuralink R1) → Downstream (hospital implant surgeries, rehabilitation training, lifelong programming services). Medical-grade BCI's downstream is tightly bound to neurosurgery centers (in China, concentrated at Huashan, Xuanwu, Tiantan, West China and other top-tier tertiary hospitals), a channel structure completely different from pharmaceuticals.
Industry coordinates today: The technology sits at the transition point on the S-curve from "invention phase" to "introduction phase"—cumulative global human implants total roughly 250-280 cases (approximately 176 on a long-term implant basis; Precision's 100+ cases are ≤30-day temporary implants), penetration below 0.001% of the addressable patient pool; commercialization revenue comes almost entirely from research funding and equipment sales, with end-to-end implantable product revenue at only one company, Treater, totaling RMB 6.175 million cumulative.
Bottom-up estimation (patient pools drawn from primary epidemiology such as GBD 2021; sources: GBD/Lancet/WHO literature series):
| Indication | Addressable Patient Pool (estimated basis) | Optimistic Penetration | Full-Cycle Price per Case | Optimistic TAM |
|---|---|---|---|---|
| Spinal cord injury tetraplegia | 15.4 million global traumatic SCI prevalence; cervical × severe addressable ≈ 3.66 million | 15% | $100,000 | $54.9 billion |
| ALS | ~350,000 globally | 20% | $150,000 | $10.5 billion |
| Post-stroke aphasia | 93.8 million survivors × 30% aphasia × 25% severe ≈ 7.03 million | 5% | $80,000 | $28.1 billion |
| Post-stroke hemiplegia | 93.8 million × 50% upper-limb impairment × 15% severe ≈ 7.03 million | 3% | $80,000 | $16.9 billion |
| Epilepsy (invasive monitoring/closed-loop) | ~10.1 million surgical candidates | 8% | $30,000 | $24.2 billion |
| Disorders of consciousness | ~1.75 million globally (estimated) | 5% | $100,000 | $8.8 billion |
| Optimistic Total | ≈ $143.4 billion |
It must be emphasized: the penetration assumptions in the table above have no precedent anchor and represent an optimistic upper bound. Using the most mature comparable implantable neuromodulation—DBS (deep brain stimulation, approved 1997)—as a mirror: only about 244,000 cumulative global implants over three decades, a cumulative penetration rate of about 2.4% against a Parkinson's patient pool in the tens of millions. If discounted according to DBS's penetration history (penetration rate falling to the 0.5%-3% tier), the implantable endgame TAM converges to on the order of ~$30 billion. This report uses the formulation of "a range from tens of billions to over a hundred billion dollars, on a 20-30 year horizon" and does not use a single TAM figure as an investment basis.
Non-invasive: medical-grade is currently the main revenue source (narrow definition ~$431 million in 2026, Grand View Research, aggregated figures, to be verified; ~$774 million by 2033, CAGR ~8.7%); consumer headbands globally ~$230 million (Fact.MR, third-hand aggregated figures, to be verified). Non-invasive is a breadth business of "penetration × repeat purchase"; the ceiling depends on whether consumer-grade brain monitoring scenarios can stand up, with penetration currently below 0.1% of wearables.
Base year 2025 global core revenue ~$580 million (non-invasive devices ~$430 million + implantable research/clinical procurement ~$150 million, the latter an estimate, to be verified). Five drivers itemized: non-invasive devices $430M→$800M (CAGR ~13%); implantable post-approval commercialization 2030: 600 cases × $100,000 = $600 million (basis: Synchron pivotal launch, Neuralink's plan to scale to several hundred cases/year, China's multi-center registrational trials starting 2026-27); installed base ~15,000 cases × $15,000 average annual service fee = $225 million (benchmarked against RNS/DBS remote programming); registrational clinical trial equipment procurement +$50 million (219 BCI-related trials registered on ClinicalTrials.gov, 134 on China's ChiCTR—a doubling-scale trial pipeline expansion over two years is a typical leading indicator of the introduction phase); military/government +$50 million (DARPA neurotechnology budget ~$80 million/year). Total 2030E ~$1.875 billion, implied CAGR ~26.4%.
Commercialization-side adjudication of this bridge (adopted by this report): the probability of the implantable 600-case line item being delivered is only ~20-25%—US PMA at earliest 2029-2030, China's annual implant volume constrained by payment; global 2030 commercial implants more likely land at the ~1,500-case level, with implant revenue of $150-300 million; the probability of the overall $1.875 billion figure being delivered is ~35-45% (the non-invasive and research expansion portions are more solid).
Implantable revenue today comes almost entirely from research/government funding (trial-phase patients pay nothing); consumer-grade is 100% out-of-pocket. The truth on the insurance side is in I4: China's medical insurance has so far only priced the surgical procedure fee (RMB 6,552/case); the RMB 475,000 consumable is essentially fully out-of-pocket; US Medicare has neither an NTAP nor a CPT code for invasive BCI. Payer coverage depth is the entire distance between the hundred-billion-dollar endgame narrative and hundred-million-dollar-level actual revenue.
Note: Precision's 100+ cases are ≤30-day intraoperative/temporary implants, not counted in the long-term implant basis; Synchron was reported by industry media in June 2026 as having "100 cumulative cases," but this lacks first-hand company confirmation (to be verified); this report uses the company-reported figure of ~20 cases throughout (Australia 4 + US 10 + Canada 6).
This is where the industry is most easily misread; the three classes of evidence must be kept separate:
① Academic peaks (n≤2, Utah array, non-commercialized systems): Stanford team's speech decoding at 62 wpm (2023-08, Nature) → ~82 wpm, 97% accuracy (2024-11); a single ALS patient achieved a 9.1% word error rate on a 50-word vocabulary, 23.8% on a 125,000-word vocabulary. "Academic research has reached conversation level" holds, but the sample is single-digit patients.
② Productized metrics (real commercial/registered systems): NeuroXess Chinese syllable decoding accuracy 71.2%, word error rate 28.8%, cursor control 5.2 BPS, latency <50ms—still one generation from conversation level; Neuralink's first patient achieved cursor control >10 BPS in the first week (healthy humans with a mouse: ~8-10 BPS), ten-finger keyboard at 40 wpm; Synchron has achieved calibration-free home use (safety data based on 6 cases).
③ Long-term safety: Neuralink's first patient had 85% thread retraction (disclosed 2024-05, compensated via algorithm); since then the company reports "zero serious device-related adverse events," with the first implant in-body 34 months—company-reported basis, sample ~27 cases, limited statistical power; individual Utah array cases have been in-body 7+ years and DBS for decades, serving as usable precedents for lifelong implants; Synchron's lead issue in 2026-11 is another reminder: a single performance/safety event can freeze regulatory, financing, and hospital-adoption processes—this is the industry's biggest tail risk.
The direction of the technology generational shift is clear: channel counts multiply 3-10× per generation (96-128 → 1024/1600 → 3000/10000+), output leaps from "cursor control" to "speech/multimodal," and in-body lifespan evolves from 5-7 years toward "lifetime." But the generational leap still awaits multi-center, n≥10 replicated validation—until then, "conversation-level decoding" can only be treated as a post-2030 option and does not enter the 12-24 month delivery logic.
Cost anchors (converted at 7.15 FX for USD figures): DBS full-cycle average ~$47,600 (meta-analysis of 26 studies) ≈ RMB 341,000; US cochlear implant full-cycle $50,000-100,000; Neuralink trial-phase per-case surgical cost third-party estimated ~$10,500; Musk claims mass-production device cost of $1,000-2,000, ~$5,000 total including surgery (≈RMB 36,000)—this is a stated target, not a measured cost, ~89.5% below the DBS anchor, with delivery dependent on robotic automation and consumer-electronics-grade supply chains. Non-invasive has already provided a cost-reduction precedent: BrainCo has compressed bionic hand cost to 1/5-1/7 of imported equivalents.
Assessment of the ramp-up bottleneck (corrected by multi-source evidence): it is not surgical capability—China performed nearly 10,000 DBS surgeries in 2024, surpassing the US; stereotactic/robot-assisted implantation capability is mature; it is payment willingness and site processes. Treater's 13 units sold against 70+ units of intent orders (conversion rate ~18.6%) shows "intent ≠ payment"; single-site annual implant capacity is currently only single digits, constrained by ethics-approval pacing and processes rather than device manufacturing. Observation metrics should lock onto quarterly implant volume and intent-to-order conversion rate, not site counts.
| Tier | Definition | Players and Revenue |
|---|---|---|
| L1 Research funding phase | Revenue from research procurement | Blackrock (Utah research installed base), Precision (>100 temporary implants), global research procurement ~$150M/year (estimate, to be verified) |
| L2 Registrational clinical phase | No revenue | StairMed (WRS02), NeuroXess (GCP, 15 tertiary hospitals), Treater ANS epilepsy indication, five US IDE holders |
| L3 Approved + out-of-pocket + Huimin insurance | The only active tier globally | Treater NEO: list price RMB 475,000/unit, listed in 27 provinces, first sales of 13 units (RMB 6.175 million), 8 end-site implants |
| L4 Insurance/commercial coverage phase | Reimbursement in place | Only China has a prototype: consumable code (2026-03-22) + Shanghai Huihui Bao 30% payout (annual cap RMB 150,000) + Zhejiang inclusion in medical insurance; US blank |
| L5 Volume ramp phase | No player has arrived | Benchmark: NeuroPace RNS received PMA in 2013; 13 years later annual revenue ~$100 million (2026 guidance $99.5 million), still loss-making |
Order verification (Treater IPO first-round inquiry response, 2026-09-29, disclosed by the SSE): list price of RMB 475,000/unit confirmed; "13 units sold since July 2026" confirmed; 8 end-site implants (a lag exists between sale and implant, ~5 units in transit); "on-hand + intent orders exceed 70 units" confirmed but intent orders are unsigned, intent : sales = 5.4 : 1; Shanghai Huihui Bao's 30% payout on list price with an annual cap of RMB 150,000 confirmed (475,000 × 30% = 142,500, just within the cap).
RMB 475,000 ≈ $66,400: 1.39× the DBS full-cycle cost, 0.89× the US cochlear implant median—not anomalous pricing, but a first-mover sole-supply premium. The medium-term price-reduction driver is not supply cost, but payer volume-for-price negotiation (Huimin insurance renewals, provincial Class B listing, national reimbursement negotiation).
China payment-chain timeline (all from first-hand government documents): 2025-03-11 NHSA's "Pricing Project Guide for Neurology Medical Service Items" established three items: invasive implant/removal + non-invasive fitting → 2025-03-31 Hubei became the first province with government-guided pricing (implant RMB 6,552 / removal RMB 3,139 / fitting RMB 966) → 2025-08-26 Shanghai piloted project pricing → 2025-12 Wuhan completed the nation's first surgery billed at guided prices → 2026-03-13 Treater received its certificate → 2026-03-22 received the world's first BCI medical insurance consumable code (9 days after approval) → 2026-07 Shanghai Huihui Bao inclusion (30% payout); most provinces have priced implant fees at RMB 6,000-6,600 (NHSA, 2026-04).
But a popular misreading must be dispelled: "fastest closed loop" ≠ "payment connected." The RMB 6,552 surgical fee is only 1.38% of the consumable's list price; the consumable is not in basic medical insurance, and the patient's out-of-pocket ≈ RMB 340,000 (475,000 × 70% + implant + fitting fees) = 8.2 years of national per-capita disposable income. Huihui Bao's 30% payout is only a marginal improvement. The real trigger for volume growth is the consumable substantively entering medical insurance (provincial Class B / national negotiation) or a sharp rise in Huimin insurance payout ratios—base-case projection: Huimin insurance expansion 2026-2028, provincial Class B 2027-2029 (probability ~60%), national negotiation 2030-2032 (probability ~40-50%, conditional on the list price falling to the RMB 300,000 level).
The US pathway's timeline is starker: no approved product → no NTAP/CPT; precedent anchor—cochlear implants took 21 years from 1984 approval to Medicare's 2005 national coverage determination (NCD); RNS has been approved 13 years with still no Medicare-specific NCD. Combined with Synchron's pause and Neuralink's stall, this report shifts the entire US timeline right by 12 months: Synchron pivotal submission 2027-2028, PMA 2029-2030E; Neuralink submission 2028-2029, PMA 2030E or later; within 12-24 months the US line does not constitute a tradeable revenue delivery event.
Base-case path (out-of-pocket + Shanghai Huihui Bao → Huimin insurance expansion → provincial Class B pilot → national negotiation preparation → national negotiation with 40-50% reimbursement; 2027 volume base case 40-60 units, list price RMB 450,000-475,000, already revised down for payment constraints). 2030E China revenue ~RMB 245 million (700 units × RMB 350,000)—only ~5.7% of the global implantable commercialization figure ($600 million); this disparity is itself a sobering dose for the "global volume ramp narrative." List price path: 2026Q4 flat at RMB 475,000 → 2027H1 flat (out-of-pocket RMB 320,000-340,000) → 2027H2 probing RMB 460,000-475,000 before second approval lands → 2028 provincial Class B volume-for-price RMB 420,000-450,000. Cumulative 2026-2030 Chinese manufacturer revenue ~RMB 441 million.
Implantable devices with no commercial revenue make CR3 inapplicable; ranked by "approval + implant cases + funding": First tier — BrainCo (global first certificate + sales in transit), Neuralink (largest trial scale and widest technology pipeline), JieTi Medical (most implant cases globally for a single company); Second tier — Synchron (endovascular route + $200 million Series D, but currently paused), Precision (only US-market-approved device + over 100 temporary implants + Medtronic partnership), NeuroXess (Chinese language decoding + GCP registered clinical trial). The true hierarchy of moats: regulatory approvals and in-vivo data (decade-scale) > hospital channels and neurosurgery KOLs > implant-grade manufacturing and packaging > capital (top players raising $200 million to $1 billion each); chips and algorithms are actually not the scarcest link — all top players develop them vertically in-house.
(BrainCo is on the non-invasive route; BoRuiKang (博睿康) is converted at an expected IPO market cap of RMB 3.5–4.0 billion, pending pricing confirmation at listing.)
Bargaining power structure: upstream (electrodes/chips/materials) has weak bargaining power — top players are vertically integrated; downstream (hospitals/neurosurgeons/payers) has strong bargaining power — at this stage, value is captured by hospital channels and clinical KOLs, and equipment makers' profits are eaten by "certification + hospital adoption" costs; only after payment codes are implemented will power shift back to equipment makers.
SW Medical Devices (Level II) PE 32.36x (32nd percentile since 2014), PB 2.39x (2nd percentile — historically extremely low, 2026-11-13); while at the BCI concept layer: Tom Cat PB 16.8x (97th percentile over 10 years), Chuangxin Medical PB 4.59x (89th percentile, company loss-making), Nanjing Panda PE(TTM) around 126x and loss-making in the first three quarters of 2026. A-share mapping screen: Real business — BoRuiKang (IPO application under review), Nurotron (genuine neuro-implant cash flow + auditory brainstem implant ABI granted NMPA innovative-device registration on 2026-10-20), Sanbo Brain Hospital (private neurosurgery platform, participated in the first case, but PE 155x already fully priced in); Concept/semi-concept — Tom Cat, Nanjing Panda, Chengyitong, Chuangxin Medical, all with zero BCI revenue contribution.
China (for implantable clinical-grade: full tailwinds): Approvals — NMPA innovative-device special review green channel; from 2025-06, invasive BCI is explicitly regulated as Class III; two certificates already issued (BoRuiKang NEO 2026-03-13, global first; JieTi 2026-03), JieTi/NeuroXess similar systems in the channel. Industry — MIIT and six other departments' "Implementation Opinions on Promoting Innovative Development of the Brain-Computer Interface Industry" (MIIT-Linked-Sci-Tech [2025] No. 164, 2025-08): key technology breakthroughs and 2–3 clusters by 2027; 2–3 global leaders cultivated by 2030; the "15th Five-Year Plan" outline (2026-03) features a dedicated BCI column (electrodes/specialized chips/codec algorithms/Chinese-language corpus R&D); Shanghai/Beijing action plans set measurable targets (Shanghai: 5 products complete type testing + clinical trials by 2027; Beijing: 3–5 global leaders by 2030). Payment — see I4.3's five-step chain. Standards — China-led ISO/IEC 8663:2025 (BCI terminology) is the first international standard; in 2026-11 MIIT approved nine industry standards for initiation.
US (approval tailwind, payment absent): Invasive BCI regulated under Class III PMA+IDE (FDA's 2021-05 dedicated guidance remains the current version); Breakthrough Device designations: Neuralink (2025-02, speech), Precision (2023-10); Paradromics entered the TAP accelerator. Neuralink's regulatory history (first IDE rejected in 2022 → approved 2023-05 → thread retraction in 2024 without a halt) shows FDA's approach is "issue-driven, case-by-case control." The payment side is a fault line: NTAP is premised on market authorization, and FY2026 IPPS contains no BCI entry — CMS has "nothing to authorize."
Geopolitics and ethics: The US BIS has listed BCI on the emerging-technology export control assessment list since 2021 (86 FR 59621), with no dedicated rule implemented to date — "deterrence in transit, controls not yet landed"; DARPA (NESD/N3/RAM/RESTORE) and military demand are the core drivers of its dual-use nature. Neural data privacy: Colorado/California/Montana have classified neural data as sensitive data; UNESCO adopted the global first "Recommendation on the Ethics of Neurotechnology" in 2025-11 — constraining consumer-grade data monetization more than clinical-grade. US research funding contraction (NIH BRAIN fell from its FY2023 peak of $680 million to $321 million in FY2025) contrasts with China's expansion.
Net direction: Implantable clinical-grade — China enjoys four-line tailwinds ("approval + industry + payment + standards"), the fastest globally; the US has approval tailwind but a payment gap. Non-invasive consumer-grade — neutral-to-cautious (no device-approval dividend + rising privacy compliance costs).
| Scenario | Probability | Narrative | Implications |
|---|---|---|---|
| Bear | 30% | HuiMinBao expansion stalls / consumables not in medical insurance, US line delayed again, or a serious safety event | China annual implants <40 units, industry revenue stuck at RMB 100M tier, systematic A-share concept-layer drawdown |
| Base | 55% | Policy chain proceeds on schedule: HuiMinBao expansion to 5–10 cities, second certificate late 2027–2028, provincial Class B pilots; US pivotal filed 2027–2028 | China 40–60 units in 2027, listing price RMB 450–475 thousand; real-business names earn earnings, concept stocks range-bound at highs |
| Bull | 15% | Multiple provinces implement Class B in 2027 or early national negotiation + Synchron resumes and files pivotal in 2027 | China 100–150 units/year in 2027–28, revenue bridge revised up 2–3x, sector repriced |
In one sentence: bullish momentum ≠ worth buying. The bottom-right quadrant (momentum up but poor returns) is typified by A-share concept stocks with zero BCI revenue, priced purely on narrative — Tom Cat at PB 97th percentile over 10 years, Chuangxin Medical at 89th percentile and loss-making — pricing in revenue two orders of magnitude away; moreover, the catalyst-dense 2026–2028 period (IPO, second certificate, pivotal) may sustain or even deepen thematic heat. There is no timetable for the premium itself to revert, so it is unsuitable as an allocation basis. True "core allocation" lies in the top right: Nurotron (real cochlear implant profits + the ABI-approved neuro-implant platform) and BrainCo (genuine non-invasive cash flow); Medtronic belongs in the top left as "stable rent collection + M&A consolidation option."
| Scenario | Most Benefited | Most Hurt |
|---|---|---|
| Bear (30%) | Nurotron, Medtronic (defensive existing cash flow); real-business names relatively resilient after thematic capital exits | Tom Cat / Nanjing Panda / Chuangxin Medical (systematic concept-layer drawdown); BoRuiKang IPO pricing under pressure |
| Base (55%) | BoRuiKang (exclusive first-certificate supply + research equipment subsidy), JieTi Medical (strongest contender for the second certificate) | Zero-revenue concept stocks drifting lower; Synchron (primary-market valuation markdown) |
| Bull (15%) | BoRuiKang + JieTi (volume directly monetized), NeuroXess (Chinese-decoding positioning), Sanbo Brain Hospital (implant surgery volume elasticity) | Conservative incumbents (valuations grabbed by narrative) |
Priority watch: BoRuiKang — global first certificate + listing in 27 provinces + No.1 EEG share cash-flow base, but must be priced on "2029–2030 profitability guidance + 3.5-year cumulative losses of RMB 328 million"; IPO valuation-overextension risk is explicit. JieTi Medical — most implants globally, sharpest execution, under A-share IPO tutoring. Nurotron — A-share-rare genuine neuro-implant cash flow (net profit +205%) + new ABI certificate. BrainCo — genuine non-invasive cash flow + HKEX filing; revenue quality needs verification once the prospectus is disclosed. (Neuralink/Precision/Synchron are private-market names, tracked for industry purposes only.)
Avoid: Tom Cat, Nanjing Panda, Chengyitong, Chuangxin Medical — zero BCI revenue contribution, valuation percentiles 89–97%, value-trap tier in the bottom-right quadrant.
Key industry-level risks: ① a single serious adverse event freezing the whole industry (Synchron's pause and thread retraction set the precedent); ② consumable medical-insurance access later than expected, volumes stuck at double digits per year; ③ US PMA and payment codes both delayed, global revenue bridge revised down; ④ sector-contagion drawdown amid extreme valuation divergence at the concept layer; ⑤ technology friction in cross-border cooperation from neural data privacy and export controls (under BIS review).
Data and methodology notes: implant counts per company announcements/regulatory filings/authoritative media (as of 2026-11-15); Synchron's "100 cases" not confirmed by primary company sources — company figures of ~20 cases used throughout; aggregated market-size estimates (Grand View/Fact.MR) serve only as cross-references and are flagged as TBC; FX assumption 7.15; valuation percentile data as of 2026-11-13/15. This report does not constitute investment advice.