Minimally invasive spine surgery with endoscopy, navigation, robotics, and artificial intelligence: Clinical evidence and future directions

Abstract

In recent years, minimally invasive spine surgery (MISS) has advanced rapidly in both technological capability and clinical adoption. By leveraging endoscopic visualization, minimally invasive working corridors, image-guided navigation, and robot-assisted instrumentation, MISS aims to achieve adequate decompression and stabilization while minimizing iatrogenic soft-tissue disruption and enhancing intraoperative accuracy and safety. This systematic critical review synthesizes contemporary innovations in MISS, including high-definition and three-dimensional endoscopy, intraoperative computed tomography/three-dimensional fluoroscopy-based navigation, and intelligent surgical instruments, and rigorously appraises clinical evidence across common spinal disorders (degenerative disc disease, spinal stenosis, deformity) and complex pathologies (tumors, infections). A structured literature search (2013-2023) adhering to PRISMA guidelines was conducted to identify randomized controlled trials, cohort studies, and systematic reviews/meta-analyses. Available data generally support meaningful perioperative advantages over conventional open surgery, including reduced blood loss, less postoperative pain, shorter hospitalization, faster functional recovery, and lower complication rates in appropriately selected patients, with expanding indications as techniques mature. However, these benefits are contextual, heterogeneity in surgeon experience, case complexity, and technology readiness influence outcomes. Key limitations include a steep learning curve, imaging-related radiation exposure, cost considerations, and ongoing debates regarding long-term effectiveness in multilevel disease or deformity correction. Future progress will depend on artificial intelligence-assisted planning/navigation, augmented visualization, personalized implants, and emerging biomaterials, alongside standardized training and high-quality prospective studies to define value, safety, and durable outcomes. Critical appraisal of evidence strength (GRADE system) and risk of bias (Cochrane Tool, Newcastle-Ottawa Scale) highlights where data remain inconclusive, emphasizing the need for targeted pragmatic randomized controlled trials and standardized outcome reporting.

Key Words: Artificial intelligence; Biomaterials; Clinical outcomes; Endoscopic techniques; Image-guided navigation; Individualized treatment; Minimally invasive decompression; Minimally invasive spine surgery; Robot-assisted surgery

Core Tip: Minimally invasive spine surgery uses navigation, endoscopy, and robotics to minimize soft-tissue damage, enhance precision, and improve safety, achieving clear benefits in degenerative disc disease, spinal stenosis, deformity, and related disorders. Compared with open surgery, minimally invasive spine surgery offers less trauma, quicker recovery, and particular advantages for older or comorbid patients. Yet its broader adoption is limited by a steep learning curve, radiation exposure, and evolving indications. Future priorities include artificial intelligence, individualized medicine, novel biomaterials, and strengthened training and standardized clinical research.

INTRODUCTION

Spinal disorders are among the leading causes of pain and functional disability worldwide, imposing a substantial burden on individual health and socioeconomic systems[1,2]. With population aging and changes in lifestyle, the incidence of lumbar disc herniation, degenerative spinal diseases, and spinal deformities has been increasing year by year[3]. Patients often present with chronic pain and restricted mobility, which severely impair quality of life[4]. In addition, spinal disorders are associated with intensive healthcare utilization, including long-term pharmacologic treatment, physical therapy, and surgical interventions[5,6].

Conventional open spine surgery typically requires extensive soft-tissue dissection and wide exposure to allow direct visualization of the operative field. Although such procedures have achieved notable success in relieving neural compression and stabilizing the spine, their limitations have become increasingly apparent. First, extensive tissue trauma leads to significant postoperative pain and greater intraoperative blood loss, thereby increasing the risk of infection and other complications. Second, the prolonged rehabilitation period delays early functional recovery, lengthens hospital stay, and increases overall healthcare costs[7]. Moreover, for elderly patients and those with multiple comorbidities, tolerance to open surgery is often reduced, and perioperative risks are substantially higher[8].

To overcome the shortcomings of traditional open procedures, minimally invasive spine surgery (MISS) has emerged as an important alternative[9]. MISS employs advanced image-guided navigation, endoscopic techniques, and specialized instruments to accomplish surgical goals through small incisions or percutaneous approaches, thereby minimizing injury to surrounding soft tissues and paraspinal muscles[10,11]. Compared with conventional open surgery, MISS is associated with reduced intraoperative blood loss, less postoperative pain, and faster recovery[12]. These advantages not only improve patient satisfaction and health-related quality of life, but also decrease the incidence of surgery-related complications[13,14]. As techniques continue to mature, the indications for MISS have expanded progressively to encompass a wide spectrum of conditions, ranging from disc herniation and spinal stenosis to various forms of spinal deformity[9,11,13].

Given MISS’s clinical importance and rapid evolution, a systematic critical synthesis of its latest advances, technological innovations, and clinical outcomes is warranted. This review aims to: Detail the structured literature search strategy, inclusion/exclusion criteria, and study selection process to mitigate selection bias; synthesize technological innovations [endoscopy, navigation, robotics, artificial intelligence (AI), biomaterials] with clear differentiation of technology readiness levels; rigorously appraise clinical evidence for MISS across heterogeneous indications, with explicit linkage of outcomes to specific techniques; integrate balanced discussions of benefits, complications, controversies, and evidence gaps within each clinical/technological subsection; define key concepts (short/long-term outcomes, high-risk patients) operationally; articulate concrete research priorities to address unresolved questions. An overview of MISS’s technological foundations, clinical indications, outcomes, and future directions is summarized in Figure 1.

Figure 1 Innovation and clinical efficacy of minimally invasive spine surgery. The central illustration shows the human spine, surrounded by five key domains related to minimally invasive spine surgery (MISS): (1) Technological innovations: The development of spinal endoscopy, image-guided navigation, intraoperative computed tomography/three-dimensional fluoroscopy, robotics, and artificial-intelligence-assisted planning enables more precise and less invasive spinal procedures; (2) Procedures: Representative MISS techniques include percutaneous endoscopic discectomy, endoscopic decompression for spinal canal stenosis, minimally invasive transforaminal lumbar interbody fusion, lateral and extreme lateral interbody fusion, and percutaneous tumor ablation combined with vertebral cement augmentation; (3) Clinical indications: Common indications comprise degenerative disc disease, lumbar spinal stenosis, spinal deformity, and spinal tumors or infections; (4) Clinical outcomes: Compared with conventional open surgery, MISS is associated with reductions in pain visual analogue scale, disability (Oswestry Disability Index), intraoperative blood loss, length of hospital stays, and procedure-related complications, together with improved quality of life; and (5) Challenges and future directions: Current limitations include a steep learning curve, radiation exposure, indication selection, and high costs, whereas future development is expected to rely on artificial-intelligence-based navigation, augmented/virtual-reality and telesurgery platforms, personalized three-dimensional-printed implants, and biomaterials- and tissue-engineering-based strategies. CT: Computed tomography; 3D: Three-dimensional; AI: Artificial intelligence; PED: Percutaneous endoscopic discectomy; MIS-TLIF: Minimally invasive surgery-transforaminal lumbar interbody fusion; LLIF/XLIF: Lateral and extreme lateral interbody fusion; DDD: Degenerative disc disease; VAS: Visual analogue scale; ODI: Oswestry Disability Index; QoL: Quality of life; AR/VR: Augmented/virtual-reality.

LITERATURE SEARCH STRATEGY AND STUDY SELECTION

Search methods

This review was conducted in accordance with the PRISMA guidelines. A comprehensive literature search was performed across PubMed, EMBASE, Cochrane Library, and Web of Science databases from January 2013 to December 2023. The search terms included combinations of “minimally invasive spine surgery”, “MISS”, “endoscopic spine surgery”, “robot-assisted spine surgery”, “image-guided navigation”, “artificial intelligence”, “spinal degenerative disease”, “spinal stenosis”, “spinal deformity”, “clinical outcome”, and “randomized controlled trial”.

Inclusion criteria: (1) Original clinical studies [randomized controlled trials (RCTs), cohort studies, case-control studies] or systematic reviews/meta-analyses; (2) Studies comparing MISS with open surgery or evaluating standalone MISS techniques; (3) Studies reporting key outcomes including intraoperative blood loss, length of hospital stay, visual analogue scale (VAS) score, Oswestry Disability Index (ODI), complication rate, or long-term fusion rate; and (4) English-language publications with full-text availability.

Exclusion criteria: Preclinical studies, basic science research, or technical notes without clinical outcomes; case reports, editorials, or narrative reviews without systematic evidence synthesis; studies with sample size < 30 patients (for observational studies) or < 50 patients (for RCTs); studies with follow-up duration < 12 months for long-term outcome assessment.

Study selection and quality assessment: Two independent reviewers (Dong Li and Xiao-Dong Tang) screened titles/abstracts and full texts, with discrepancies resolved by a third reviewer (Rui-Bo Wang). The risk of bias was assessed using the Cochrane Risk of Bias Tool for RCTs and the Newcastle-Ottawa Scale for observational studies. The GRADE system was employed to evaluate the quality of evidence for key clinical outcomes.

Technological innovations in MISS

Advances in surgical instrumentation: MISS evolution is closely tied to surgical instrumentation innovations, improving precision, safety, and indication scope. Key breakthroughs include high/ultra-high-definition endoscopes, minimally invasive working channels, and high-precision tools, though their clinical translation varies by technology readiness.

Endoscopic systems: High/ultra-high-definition (4K) endoscopes enhance fine-detail imaging[15,16], while three-dimensional (3D) endoscopy improves depth perception and spatial awareness[17,18]. Fluorescence imaging enables real-time visualization of vascular/neural structures[19]. Limitations: Narrow field of view, technical complexity, and steep learning curves, novice surgeons require 30-50 cases to achieve proficiency. Controversy: While 4K/3D endoscopy reduces iatrogenic injury in expert hands[18], low-quality studies (small sample sizes, unblinded outcome assessment) overstate benefits in non-specialized centers[16].

Working channels and dilators: Sequential muscle dilators access the surgical field via intermuscular planes, minimizing retraction injury[20,21]. New-generation channels use flexible, modular designs and lightweight alloys[22,23]. Limitation: Applicable only to specific approaches and ineffective for complex deformity correction requiring wide exposure[23]. High-precision tools: Miniaturized drills, ultrasonic osteotomes, and plasma knives enable precise bone/soft-tissue resection with hemostasis[24]. Intelligent instruments with haptic feedback improve manipulation accuracy. Robot-assisted systems integrate mechanical control with navigation. Critical appraisal: High-quality RCTs (GRADE moderate) show haptic feedback reduces pedicle screw misplacement[25], but cost (1 to 2 million dollars per system) limits dissemination.

Application of advanced imaging technologies: Advanced imaging is critical for MISS, providing anatomical localization and real-time navigation[26,27], though radiation exposure remains a key concern. Intraoperative computed tomography (CT) navigation: Integrates real-time cross-sectional images with navigation, improving pedicle screw accuracy[28]. Meta-analyses (GRADE high) show CT navigation reduces screw misplacement rate [relative risk (RR) = 0.28, 95% confidence interval (CI): 0.18-0.44, I² = 32%] vs conventional fluoroscopy[29-32]. Limitations: Increased radiation exposure (surgeon hand dose: 0.3-0.8 mSv/case vs 0.1-0.3 mSv for fluoroscopy) and operative time (+15-20 minutes)[28].

3D fluoroscopy and real-time navigation: C-arm-based 3D reconstruction provides dynamic anatomical guidance[26,27,31,33,34] [technology readiness level (TRL) 8: Widely used]. Evidence (GRADE moderate) shows reduced screw malposition and fluoroscopy time[28,32,35], but radiation exposure remains higher than open surgery. Augmented/virtual-reality (AR/VR) technologies. Superimpose patient-specific anatomy onto the surgical field[36,37]. Limitations: High cost, limited long-term data on outcomes, and surgeon adaptation requirements[36]. Controversy: While AR improves spatial perception in simulation[37], no RCTs demonstrate superiority over CT navigation in clinical settings. When coupled with AI-based algorithms for image reconstruction and instrument tracking optimization, these imaging technologies collectively provide MISS with higher levels of safety and procedural success[38,39].

Robot-assisted surgery: Robot-assisted surgery enhances MISS precision and safety via mechanical control, navigation, and intelligent planning, with clear benefits but contextual limitations[40]. Robotic positioning and guidance: Spinal robotic systems use preoperative/intraoperative imaging to construct 3D anatomical models, guiding instruments along planned paths[41-43]. Clinical evidence (GRADE moderate): Robotic pedicle screw placement accuracy (> 98%) exceeds conventional fluoroscopy[42,43], with reduced radiation exposure and operative time in complex cases[44]. Benefits in deformity correction: Shortens learning curve for surgeons[45] and avoids pre-existing implants in revision surgery[46]. Limitations: High equipment cost, dedicated training, and inconsistent outcomes in low-volume centers[47]. Controversy: A cohort study reported no difference in complication rates vs fluoroscopy in unselected cases[46], highlighting the role of case complexity in determining value.

AI in surgical planning: AI optimizes medical image analysis, trajectory planning, and individualized strategies[48]. Deep learning algorithms automate spinal anatomy segmentation[48] and optimize screw paths[49]. Real-time intraoperative AI alerts reduce neural injury risk[50]. Critical appraisal: No large-scale RCTs (GRADE very low) validate AI’s clinical superiority, current evidence is limited to small feasibility studies. AI cannot replace surgeon judgment but serves as a decision support tool[51]. Future direction: Integration of multi-omics data with AI to predict patient-specific outcomes[52]. The main MISS-related technological modalities, together with their advantages, limitations, and appropriate clinical scenarios, are summarized in Table 1.

Table 1 Summary of advantages and disadvantages of minimally invasive spine-related technologies.

Technology	Advantages	Disadvantages	Typical clinical scenarios
Intraoperative CT navigation	Accurate localization of bony structures; enhances surgical safety; reduces intraoperative errors	Increases operative cost and radiation exposure; requires high operator proficiency	Spinal fixation; pedicle screw placement; complex fracture reconstruction
Endoscopic systems (HD/4K)	High-definition visualization; reduces soft-tissue trauma; improves clarity of the operative field	Limited field of view; narrow working space; technically demanding	Discectomy; spinal decompression; laminectomy
Robotic systems	High accuracy; shorter operative time; reduces intraoperative surgeon fatigue	High equipment cost; system complexity; requires dedicated training	Spinal deformity correction; complex multilevel procedures
Sequential muscle dilators	Decrease muscle retraction; reduce postoperative pain; shorten recovery time	Applicable only to specific approaches; cannot fully replace conventional open procedures	Lumbar fusion surgery; disc procedures
Fluorescence imaging (e.g., ICG)	Improves intraoperative identification of vascular and neural structures; reduces iatrogenic injury	Requires additional equipment and contrast agents; increases procedural complexity	Complex spinal surgery in vascular-rich regions

CT: Computed tomography; HD: High definition; 4K: 4K ultra–high definition; ICG: Indocyanine green.

The introduction of robot-assisted surgery has brought about a revolutionary change in MISS. By combining precise mechanical control, advanced image-guided navigation, and intelligent surgical planning, robotic technology markedly improves procedural accuracy and safety and expands the range of indications for MISS[53-55]. The major MISS-related technical approaches, along with their advantages, limitations, and suitable clinical scenarios, are summarized in Table 2.

Table 2 Comparison of surgical techniques and indications.

Technique	Indications	Advantages	Disadvantages
Endoscope-assisted decompression	DDD; spinal canal stenosis	Minimal tissue trauma, rapid recovery, and reduced intraoperative blood loss; high-resolution imaging improves surgical visualization	Limited indications; steep learning curve; high operative complexity
Percutaneous endoscopic discectomy	Mild to moderate disc herniation	Small incision, less postoperative pain, shorter hospital stay; reduced soft-tissue disruption	Restricted endoscopic field of view and relatively narrow indications; requires high technical precision
Robot-assisted surgery	Spinal deformity correction; multilevel spinal fixation	High surgical accuracy and reduced human error; improves pedicle screw placement accuracy and lowers complication rates	High equipment cost and demanding training requirements; complex system maintenance
3D fluoroscopy and real-time navigation	Spinal fixation procedures; pedicle screw placement	High accuracy; lowers the risk of screw misplacement; real-time imaging shortens operative time and reduces radiation exposure	Requires substantial radiation exposure; depends on expensive imaging equipment

DDD: Degenerative disc disease; 3D: Three-dimensional.

Biomaterials and tissue engineering: Biomaterials and tissue engineering offer novel therapeutic avenues for spinal disorders, with varying clinical readiness. Disc regeneration: Mesenchymal stem cell transplantation with hydrogel scaffolds improves nucleus pulposus function[56-59]. Exosome-based approaches modulate cellular signaling[58]: Long-term cell survival, immune responses, and lack of phase III trials (GRADE very low). Bone graft substitutes: Calcium phosphate ceramics and biodegradable magnesium alloys have favorable biocompatibility[60,61]. Bioactive glass promotes new bone formation[62]. Benefits: Reduce complications of autologous bone grafting (GRADE low, limited cohort studies). Controversy: Magnesium alloy degradation rate varies by patient, with some studies reporting implant loosening[61].

Clinical applications of MISS

Degenerative disc disease: Degenerative disc disease is a leading cause of low back pain[63], with MISS techniques [percutaneous endoscopic discectomy (PED), laser/plasma-based procedures] demonstrating perioperative benefits contextualized by indication selection and surgeon experience[64-67]. PED: Removes herniated disc material via small incisions[66,68]. Meta-analyses (GRADE high) show PED reduces intraoperative blood loss [mean difference (MD) = -150 mL, 95%CI: -180 to -120], length of stay (LOS) (MD = -3.2 days, 95%CI: -4.1 to -2.3), and VAS at 1 week (MD = -2.1, 95%CI: -2.7 to -1.5) vs open discectomy[66,68,69]. Long-term outcomes (1-2 years): Similar ODI improvement (MD = -3.5, 95%CI: -7.2 to 0.2, I² = 68%)[69] (GRADE moderate, heterogeneity due to surgeon experience). Limitations: Steep learning curve (30-50 cases for proficiency) and restricted indications (mild-moderate disc herniation)[67]. Controversy: A RCT (high risk of bias) reported higher reoperation rates for PED (8% vs 3% for open surgery)[69], attributed to incomplete decompression in novice hands.

Laser and plasma-based techniques: Percutaneous laser disc decompression and plasma radiofrequency ablation reduce intradiscal pressure[64,70-72]. Evidence (GRADE moderate): Plasma ablation achieves VAS reduction (MD = -1.8, 95%CI: -2.5 to -1.1) at 6 months[73], with low complication rates (< 2%)[70]. Limitations: Ineffective for large disc herniations[74] and variable outcomes based on disc pathology[75].

Spinal canal stenosis: Spinal canal stenosis management includes minimally invasive decompression [endoscope/microscope-assisted, unilateral laminotomy for bilateral decompression (ULBD)] and indirect decompression with fusion[76] [extreme lateral lumbar interbody fusion, transforaminal lumbar interbody fusion (TLIF), anterior lumbar interbody fusion (ALIF)], with benefits modulated by patient anatomy and comorbidities.

Minimally invasive decompression techniques: Endoscope-assisted decompression. Transforaminal/interlaminar approaches resect hypertrophic bone/ligamentum flavum[77]. Meta-analysis (GRADE moderate) shows similar symptom relief vs open surgery, with reduced blood loss (MD = -120 mL, 95%CI: -150 to -90) and LOS (MD = -2.8 days, 95%CI: -3.6 to -2.0)[78]. Limitations: Limited working space in severe stenosis[77] (GRADE low, small sample sizes).

Microscope-assisted decompression. Tubular retractor-based targeted decompression preserves paraspinal muscles[20,21]. Evidence (GRADE high): Superior to open decompression in operative time (MD = -30 minutes, 95%CI: -45 to -15), blood loss, and LOS[79], with similar long-term functional outcomes[79]. ULBD: Unilateral approach for bilateral decompression[80]. RCT (GRADE moderate) shows faster recovery (return to work: 4 weeks vs 6 weeks for open) and higher patient satisfaction (85% vs 72%)[81], with low complication rates (< 4%)[81]. Controversy: ULBD may increase spinal instability in multilevel stenosis (GRADE very low, limited data).

Indirect decompression and fusion: Indirect decompression and fusion aim to relieve neural compression by restoring disc height and sagittal alignment, while simultaneously achieving segmental stabilization.

Extreme lateral lumbar interbody fusion. Kambin lumbar interbody fusion: Lateral transpsoas approach restores disc height[82]. Evidence (GRADE moderate): Improves low back/leg pain (VAS reduction: 3.5-4.0 points) and reduces LOS (3-4 days vs 7-10 days for posterior fusion)[83]. Limitations: Lumbar plexus injury risk (2%-5%)[82] and contraindications for severe psoas hypertrophy[83].

TLIF. Minimally invasive TLIF: Posterolateral transforaminal approach preserves contralateral facet joints[84]. RCT (GRADE high) shows similar fusion rates (90% vs 88%) vs open TLIF, with reduced blood loss and postoperative pain[85]. Controversy: No difference in long-term ODI (MD = -2.1, 95%CI: -5.3 to 1.1) in elderly patients[85], highlighting age-related outcome heterogeneity.

ALIF. ALIF reaches the disc space through an anterior retroperitoneal approach, thus avoiding disruption of posterior spinal muscles and neural elements[86]. Insertion of a large interbody cage restores disc height and sagittal alignment, achieving indirect decompression while providing robust segmental stability. Kapustka et al[86] reported that ALIF has unique advantages in symptom relief and restoration of spinal alignment. For patients with concomitant sagittal imbalance, ALIF may represent a particularly suitable option.

Overall, minimally invasive decompression techniques and indirect decompression with fusion have demonstrated significant efficacy and advantages in the treatment of spinal canal stenosis[87]. Minimally invasive decompression achieves adequate neural decompression with minimal tissue disruption, enabling rapid recovery and high patient satisfaction. Indirect decompression combined with fusion restores anatomical alignment and simultaneously accomplishes decompression and stabilization. As surgical techniques and instrumentation continue to advance, MISS is expected to play an even more prominent role in the management of spinal canal stenosis[87].

Spinal fusion surgery: Minimally invasive surgery (MIS)-TLIF and lateral lumbar interbody fusion (LLIF) are mainstream spinal fusion techniques, offering reduced trauma but contextual benefits. MIS-TLIF: Minimally invasive posterior approach combining TLIF biomechanics with small incisions[88,89] (TRL 8: Widely used). Evidence (GRADE high): Less intraoperative blood loss (MD = -200 mL, 95%CI: -250 to -150), lower VAS at 2 weeks (MD = -1.9, 95%CI: -2.5 to -1.3), and shorter LOS (MD = -3.5 days, 95%CI: -4.3 to -2.7) vs open TLIF[90-94]. Limitations: Steep learning curve[95] and higher cost (+5000-8000 dollars). Controversy: Low-volume centers report higher complication rates (10% vs 4% in high-volume centers)[95] (GRADE moderate, cohort studies). LLIF: Lateral approach avoiding posterior muscles[96,97] (TRL 7: Growing use). Evidence (GRADE moderate): Fewer complications (RR = 0.65, 95%CI: 0.43-0.98) and faster recovery vs MIS-TLIF in single-level degenerative disease[96], but limited data on multilevel fusion (GRADE low).

Technical features. LLIF is performed via a retroperitoneal corridor and reaches the disc space through the psoas muscle or along its anterior border[97]. Minimally invasive LLIF employs tubular retractors and an operating microscope to minimize retraction-related injury to the psoas and surrounding soft tissues[98]. In recent years, the incorporation of intraoperative neuromonitoring has effectively reduced the risk of lumbar plexus injury and improved procedural safety[94].

Indications and clinical outcomes. LLIF is indicated for conditions such as degenerative lumbar scoliosis, disc degeneration, and spinal canal or foraminal stenosis[99]. Nazierhan et al[96] reported that, compared with MIS-TLIF, crenel lateral interbody fusion, a specific LLIF technique, resulted in fewer complications, less surgical trauma, and faster postoperative recovery in patients with single-level degenerative lumbar disease. However, larger patient cohorts and longer follow-up are still needed to further validate these findings.

Spinal deformity correction: MISS for scoliosis and kyphosis offers reduced trauma but contextual correction efficacy[100]. Key techniques: Percutaneous fixation endoscope-assisted correction; LLIF percutaneous pedicle screws[101,102]. Evidence (GRADE low): Adult scoliosis correction (Cobb angle reduction: 15°-20°) with reduced blood loss and LOS[101,102]. Limitations: Inferior correction vs open surgery in adolescent idiopathic scoliosis (Cobb angle reduction: 45% vs 65%)[103] (GRADE moderate, cohort study). Controversy: No RCTs compare MISS vs open surgery in severe deformity (Cobb angle > 60°), with limited long-term data on curve progression[102].

Spinal tumors and infections. MISS improves diagnostic and therapeutic outcomes for spinal tumors and infections, with benefits in selected cases. Diagnostic biopsy: Percutaneous CT/magnetic resonance imaging-guided biopsy with > 90% accuracy and < 2% complication rate[104] (GRADE high).

Therapeutic techniques: Endoscopic tumor resection (benign tumors)[104]; radiofrequency ablation cement augmentation (malignant lesions)[104]; percutaneous fixation for infection-related instability[105,106]. Evidence (GRADE low): Pain relief (VAS reduction: 3-4 points) and stability restoration[104,106], but limited long-term survival data for malignant tumors. Controversy: MISS is contraindicated for extensive tumor infiltration (≥ 3 vertebral levels) due to inadequate resection[107]. Representative minimally invasive treatment strategies for different categories of spinal disorders and their principal outcome measures are summarized in Table 3.

Table 3 Summary of minimally invasive treatment modalities for spinal disorders.

Disease type	Minimally invasive method	Specific procedure	Outcome evaluation
Degenerative disc disease	Percutaneous endoscopic discectomy	Small skin incision; herniated disc tissue is removed under direct endoscopic visualization to relieve neural compression	Pain relief in > 85% of patients; rapid postoperative recovery; short length of hospital stay
Spinal canal stenosis	Endoscope-assisted decompression	Resection of hypertrophic bone and ligament via an endoscopic approach to enlarge the spinal canal	Improved walking capacity; reduced pain; relatively short postoperative recovery period
Spinal scoliosis	Robot-assisted surgery	Robot-guided precise pedicle screw placement combined with corrective maneuvers to restore normal spinal alignment	Significant deformity correction; low postoperative complication rate; high patient satisfaction
Spinal tumors	Percutaneous biopsy, radiofrequency ablation, and cement augmentation	CT- or MRI-guided biopsy followed by radiofrequency ablation of the tumor and subsequent cement injection to stabilize the spine	Marked pain relief; short recovery period; applicable to benign and selected malignant lesions

CT: Computed tomography; MRI: Magnetic resonance imaging.

Clinical outcomes and evidence appraisal

Operational definitions: Short-term outcomes: ≤ 3 months [intraoperative blood loss, LOS, acute pain (VAS ≤ 3 months), early complications]. Long-term outcomes: ≥ 12 months [fusion rate, chronic pain (VAS ≥ 12 months), ODI, reoperation rate]. High-risk patients: Age ≥ 75 years, ≥ 2 comorbidities (cardiopulmonary disease, diabetes, renal insufficiency), or American Society of Anesthesiologists physical status ≥ 3.

Short- and long-term clinical outcomes: Perioperative outcomes (GRADE high): MISS reduces intraoperative blood loss (< 50-100 mL vs 200-500 mL for open surgery), LOS (2-4 days vs 7-10 days), and early complication rate (< 5% vs 10%-20%)[108-110]. VAS score reduction (50%-60% at 1 week) is faster than open surgery (pain relief at 4 weeks)[108]. Long-term outcomes (GRADE moderate): MISS and open surgery achieve similar ODI improvement (30%-40% at 6 months) and fusion rates (85%-90% at 2 years)[111]. Heterogeneity drivers: Surgeon learning curve (complication rates decrease by 50% after 50 cases), case complexity (multilevel disease has worse outcomes vs single-level)[87], and center volume (high-volume centers have 30% lower reoperation rates)[95]. The key differences between traditional open surgery and MISS in terms of approach and perioperative outcomes are summarized in Figure 2.

Figure 2 Comparison between traditional open spine surgery and minimally invasive spine surgery. The left panel illustrates a traditional open posterior lumbar approach with a long midline incision, extensive paraspinal muscle stripping, and wide laminectomy exposure. These features are associated with increased muscle damage, higher intraoperative blood loss, more severe postoperative pain, longer hospital stay, and a higher risk of complications, leading to slow functional recovery, especially in elderly patients. The right panel depicts a minimally invasive spine surgery approach using a paramedian skin incision, tubular retractors, and percutaneous pedicle screw, assisted by endoscopy, image-guided navigation, intraoperative computed tomography/three-dimensional fluoroscopy, robotics, and artificial intelligence-based planning. Compared with open surgery, minimally invasive spine surgery enables small incisions, less soft-tissue trauma, and precise decompression, resulting in reduced pain (visual analogue scale), decreased blood loss and complication rates, faster postoperative mobilization, and improved quality of life, particularly for elderly and comorbid patients. CT: Computed tomography; AI: Artificial intelligence; VAS: Visual analogue scale; QoL: Quality of life.

Comparative studies with traditional open surgery. Meta-analyses (GRADE high) confirm MISS advantages in perioperative outcomes[109,110,112], with no significant long-term differences in pain or function[111]. Cost-effectiveness (GRADE moderate): MISS reduces healthcare costs (10000-15000 dollars per patient) via shorter LOS and faster return to work[113,114], with higher quality-adjusted life years vs open surgery[115]. Typical differences between MISS and traditional open surgery in perioperative parameters and early clinical outcomes are summarized in Table 4.

Table 4 Comparison between minimally invasive spine surgery and traditional open surgery.

Parameter	MISS	Traditional open surgery
Intraoperative blood loss	< 50-100 mL; minimal soft-tissue disruption	200-500 mL; requires extensive soft-tissue dissection
Length of hospital stay	Average 2-4 days	Average 7-10 days
Pain relief (VAS)	VAS score reduced by 50%-60% at 1 week postoperatively; further improvement by 2 weeks	Pain relief typically observed around 4 weeks postoperatively; recovery is comparatively slower
Complication rate	Overall complication rate < 5%; mainly minor infections and mild transient neurological deficits	Complication rate 10%-20%; includes bleeding, infection, and neurological injury
Functional recovery (ODI)	ODI improved by 30%-40% at 3 months; marked improvement by 6 months	ODI gradually improves, reaching approximately 30%-35% improvement at 6 months

MISS: Minimally invasive spine surgery; VAS: Visual analogue scale; ODI: Oswestry Disability Index.

Clinical outcomes in special populations. Elderly patients (≥ 75 years): MISS reduces postoperative complications (RR = 0.45, 95%CI: 0.32-0.63), LOS (MD = -3.8 days, 95%CI: -4.7 to -2.9), and mortality (RR = 0.31, 95%CI: 0.16-0.60) vs open surgery[116] (GRADE moderate). Limitations: Higher implant loosening rate (5% vs 2% in younger patients)[116]. High-comorbidity patients: MISS improves operative tolerance, with lower readmission rates (RR = 0.58, 95%CI: 0.41-0.83)[117] (GRADE low, limited RCTs).

Current challenges and controversies

Challenges and controversies are integrated into relevant sections above; key cross-cutting issues include. Learning curve: Most MISS techniques require 30-50 cases for proficiency[118-121], with novice surgeons having (2-3) × higher complication rates[95,122-124]. No standardized training pathways limit widespread adoption[121]. Radiation exposure: MISS increases surgeon/patient radiation dose vs open surgery, with cumulative surgeon dose exceeding annual limits in high-volume centers[125-127]. Low-radiation navigation technologies (e.g., electromagnetic tracking) are in early development[128]. Indication expansion: Overuse in complex multilevel disease or severe deformity leads to suboptimal outcomes[103]. Clear contraindications are inconsistently applied[67]. Evidence heterogeneity: Variable study designs, outcome reporting, and follow-up durations limit meta-analysis validity[111]. No standardized outcome sets hinder comparison[115].

Future directions and concrete research priorities

Key future directions: AI integration. Develop AI algorithms for preoperative planning (patient-specific screw trajectories), intraoperative real-time risk prediction (neural injury), and postoperative rehabilitation monitoring[129-131]. Personalized medicine. 3D-printed patient-specific implants for deformity correction; genomic profiling to predict fusion success[52]. Low-radiation navigation. Electromagnetic tracking and AI-based image reconstruction to reduce radiation exposure[128]. Biomaterials and regenerative medicine. Phase III trials of mesenchymal stem cell-hydrogel constructs for disc regeneration; biodegradable implants with controlled degradation rates[56,61].

Concrete research priorities: Pragmatic RCTs. Compare MISS vs open surgery in understudied indications (multilevel stenosis, severe deformity, elderly high-comorbidity patients) using standardized outcomes (VAS, ODI, quality-adjusted life years, cost) and 5-year follow-up. Learning curve studies: Evaluate outcomes (complication rates, functional scores) stratified by surgeon case volume (0-30, 31-50, > 50 cases) to define proficiency thresholds and develop structured training curricula. Technology validation: RCTs comparing AR/VR navigation vs CT navigation in complex cases (deformity, revision surgery) to assess accuracy and radiation reduction. Standardized outcome reporting: Develop a core outcome set for MISS studies (including patient-reported outcomes, cost, and long-term reoperation rates) to reduce heterogeneity. Biomaterials trials: Phase III trials of novel bone graft substitutes (magnesium alloys, bioactive glass) vs autologous bone graft in spinal fusion, with focus on fusion rate and implant stability. A conceptual roadmap illustrating the evolution from current MISS to AI-driven, AR/VR-enabled, and biomaterials-assisted platforms is presented in Figure 3, and the key goals, anticipated clinical impact, and representative technological pathways of these frontier directions are summarized in Table 5.

Figure 3 From current minimally invasive spine surgery to future intelligent, regenerative, and personalized minimally invasive spine surgery platforms. The left panel depicts a representative current minimally invasive spine (MISS) surgery procedure using endoscopic visualization and fluoroscopy-based navigation. From this baseline, a forward arrow illustrates key future directions. In the upper middle, artificial intelligence- and data-driven surgery integrates multimodal preoperative and intraoperative imaging into machine-learning algorithms to assist surgical planning, predict risk, and provide decision support, including suggested pedicle screw trajectories and optimized fusion levels or alignment corrections. On the upper right, augmented/virtual-reality and remote surgery enable augmented reality-guided navigation, immersive virtual-reality training, and telerobotic MISS via high-speed 5G communication. The lower middle highlights personalized and precision implants, such as three-dimensional-printed cages, patient-specific implants, and instrumentation designed for individualized spinal alignment and segmental stability. The lower right illustrates biomaterials and regenerative medicine approaches, including disc regeneration, bone graft substitutes, bioactive scaffolds, stem cell- and exosome-based therapies, and tissue-engineering strategies. Together, these innovations aim to deliver safer MISS with more individualized treatment, better long-term biomechanical stability, and lower complication rates. MISS: Minimally invasive spine; AI: Artificial intelligence; AR/VR: Augmented/virtual-reality.

Table 5 Future directions and research priorities.

Research area	Objectives	Potential clinical impact	Technological applications
AI	Optimize preoperative planning, real-time intraoperative navigation, and postoperative complication prediction	Improve surgical precision and safety; reduce postoperative complications	AI-based imaging analysis; intelligent intraoperative monitoring systems
Personalized medicine	Develop individualized treatment strategies based on patient-specific genetic and anatomical characteristics	Increase surgical success rates; reduce risks associated with inter-individual variability	Genomic profiling; 3D-printed patient-specific implants
Novel biomaterials	Promote spinal tissue regeneration and enhance implant biocompatibility	Accelerate postoperative healing; reduce inflammation and procedure-related complications	Nanofiber scaffolds; bioactive glass; hydrogels and other regenerative materials
Tele-surgery and virtual reality	Enable remote surgical guidance and real-time training; shorten the learning curve	Improve healthcare quality in underserved regions; reduce training time and cost	Remote robotic surgery; virtual surgical simulation; augmented-reality-based navigation systems

AI: Artificial intelligence; 3D: Three-dimensional.

CONCLUSION

This systematic critical review synthesizes technological innovations and clinical evidence for MISS, adhering to PRISMA guidelines and rigorous evidence appraisal. MISS offers clear perioperative advantages (reduced blood loss, shorter LOS, faster recovery) over open surgery in appropriately selected patients, particularly degenerative disc disease, spinal stenosis, and single-level fusion. However, these benefits are contextual: Surgeon experience, case complexity, and technology readiness modulate outcomes. Key limitations include a steep learning curve, radiation exposure, and inconsistent long-term data in complex scenarios (multilevel disease, severe deformity).

Technological innovations (robotics, AI, biomaterials) hold promises, but clear differentiation between routine, emerging, and experimental technologies is critical to avoid overstating readiness. Balanced discussion of benefits, complications, and controversies, integrated across clinical/technological sections, highlights where MISS delivers value and where data remain inconclusive. Concrete research priorities (pragmatic RCTs, learning curve studies, standardized outcomes) will address evidence gaps, enabling more personalized and effective MISS implementation. As MISS evolves, balancing innovation with rigorous evidence will ensure safe, equitable, and cost-effective care for patients with spinal disorders.