Pre‐clinical evaluation of a laser system with emission control based on stone vs soft tissue recognition Authors: Olivier Traxer, Federico Zorzi, IIya V. Yaroslavsky, Anastasia Kovalenko, Viktoriya Andreeva, James Childs, Alexey G. Martov, Gregory Altshuler Published: 2026 ABSTRACT ObjectiveTo evaluate the feasibility and performance of the StoneSense™ module (SSM; IPG Medical, Marlborough, MA, USA), a prototype real‐time target recognition unit integrated into a IPG Thulium Fiber Laser (TFL).Materials and MethodsThe SSM analyses the spectral signature of back reflected light‐emitting diode (LED) light from the target, stopping laser energy emission when incoming signal is originating from tissue or when the fibre‐to‐target distance is larger than pre‐defined threshold (typically, ≥2–4 mm). The ex vivo tests were conducted on artificial BegoStone phantoms ((Bego GmbH & Co. KG, Bremen, Germany) embedded in porcine kidney tissue using energy settings of 0.1–2 J and 5–100 Hz at two fibre‐scanning speeds (1 and 3 mm/s) through a translation stage movement. Human urinary stones were subsequently tested at 2 and 5 mm/s with setting of 0.5 J × 20 Hz. The in vivo trial was conducted by lithotripsy of pre‐implanted human stones in porcine urinary tract. Primary outcome measures included ablation efficiency as well as specificity and sensitivity of stone recognition.ResultsSoft tissue ablation was completely prevented in all the ex vivo trials while maintaining high lithotripsy efficiency. Mean stone ablation reduction was 2.75% and 3% at fibre scanning speeds of 1 and 3 mm/s, respectively. The system achieved 100% sensitivity and up to 98% specificity in inhibiting laser emission to soft tissue. In vivo, complete dusting with fragments <500 μm was achieved for all studied stone types without observable mucosal injury. Specificity ranged from 85% to 100%, while sensitivity ranged from 90% to 100%.ConclusionsIntegration of back reflected‐light‐spectra‐based laser emission control into a TFL‐based lithotripsy system is feasible and effective. The SSM enables precise real‐time stone recognition, minimises unnecessary lasing, and preserves dusting efficiency. Abbreviations artificial intelligence Animal Research: Reporting of In Vivo Experiments artificial stones brushite calcium oxalate monohydrate cystine flexible ureterorenoscopy Good Laboratory Practice holmium: yttrium‐aluminium‐garnet human stones stone ablation StoneSense™ module struvite thulium fiber laser uric acid Introduction Urolithiasis is a major global health problem associated with significant morbidity and economic burden, with >106 million incident cases reported worldwide in 2021 [1]. In Europe and North America, between 10% and 15% of the population will experience urinary stone disease during their lifetime, with recurrence rates may reaching up to 40% [2, 3]. The impact is particularly significant in working‐age adults, where recurrent episodes increase hospital admissions, impair productivity, and diminish quality of life [4]. Management of urinary stones has evolved substantially over the past decades. According to international guidelines, percutaneous nephrolithotomy remains the ‘gold standard’ for kidney stones larger than 2 cm, while smaller stones and ureteric fragments are predominantly managed with flexible ureterorenoscopy (fURS) [5, 6]. Since its introduction, the holmium: yttrium‐aluminium‐garnet (Ho:YAG) laser has become the predominant energy source for lithotripsy [7]. Recent advances in the understanding of laser–stone interaction have led to the development of new solid‐state laser systems. Among these, the Thulium Fiber Laser (TFL) emerged as a promising alternative to Ho:YAG. The TFL enables higher pulse frequencies, lower energy per pulse, longer pulse durations, and a more homogeneous temporal beam profile [8]. These technical advantages translate into reduced retropulsion, improved efficiency of energy transfer, and a greater capacity of fine dust production compared to Ho:YAG [9]. Particularly, dusting techniques are clinically relevant because they minimise the need for fragment retrieval, shorten procedure times, and allow treatment of larger stone burdens within a single fURS session [10]. Moreover, dusting reduces the likelihood of clinically significant residual fragments, thereby lowering recurrence risk [11]. Despite these advantages, laser lithotripsy continues to present safety challenges. High‐frequency settings and prolonged lasing may increase intrarenal temperatures secondary to excessive thermal dissipation [12]. Moreover, inadvertent firing on urothelium due to impaired vision during dusting procedures can result in mucosal injury, perforation, and impaired renal function [13]. To address these issues, optical monitoring technologies have been investigated as potential safeguards. Previous studies demonstrated that autofluorescence signals emitted by urinary calculi upon illumination with visible light differ significantly from those of urothelial tissue, enabling real‐time stone–tissue discrimination [14]. Building on this principle, laser systems with detection of an autofluorescence signal from a tissue and stone have been developed for Ho:YAG lithotripsy, showing promising reductions in energy delivery and urothelial tissue injury [15]. However, no stone vs tissue recognition has been tested with TFL, which owing to its pulse characteristics and superior dusting capabilities, is particularly well suited for integration with real‐time recognition systems. The aim of the present study was to provide the first pre‐clinical evaluation of the StoneSense™ module (SSM; IPG Medical, Marlborough, MA, USA), a prototype light‐guided unit integrated into an IPG TFL platform. Through ex vivo and in vivo experiments, we sought to assess its accuracy, safety, and impact on stone ablation (SA) efficiency. Materials and Methods The SSM Technology The SSM is a prototype optical recognition module integrated in the IPG TFL laser system (Fig. S1). The system relies on the ability of the surgical laser fibre 150–940 μm core diameter range to transmit treatment energy and collect back reflected‐light signal from the target. These emissions are guided back through the same fibre pathway and processed by the analyser module. Before the lithotripsy procedure, the system requires a calibration step. During this phase, the distal fibre tip is directed toward healthy urothelium and continuously moved along the soft tissue for 10 s, keeping up to 2 mm distance between fibre tip and soft tissue. The system records its reflected light signal, which is stored as the ‘non‐stone’ reference data. After calibration, the module continuously monitors optical signal reflected from the interrogated target after activation of laser pedal by the surgeon. Based on detected back reflecting amplitude, the system classifies the target in real time (<10 ms) as either ‘Stone’ or ‘Not stone’ (i.e., tissue or fluid) (Fig. 1). When a ‘Stone’ signal is identified, laser emission is enabled; conversely, if the reflected signal corresponds to tissue or is below a pre‐determined threshold – indicating that the laser fibre tip is ≥2–4 mm from the target – energy delivery is automatically suspended. Before lithotripsy, the laser fibre tip and the scope light‐emitting diode (LED) light are directed toward healthy urothelium to acquire back‐reflected optical signals, which are stored as a ‘non‐stone’ reference. This calibration step enables subsequent real‐time discrimination between stone and tissue. Energy delivery resumes immediately when the stone is re‐detected by the module. From the surgeon's perspective, system operation is fully integrated with the conventional foot pedal. When SSM is active, the pedal can remain continuously depressed, but energy is gated by the recognition system and released only when stone reflected light is detected (Fig. 2). The system can be overridden at any time, allowing the operator to regain direct manual control. This design ensures that the fibre functions both as a therapeutic conduit for TFL energy and as an optical feedback channel, enabling real‐time gating of laser emission. The fibre–stone interaction during activation of the SSM (A), and blockage of the laser emission by the system when fibre is not facing the stone (B) or is distanced >2–4 mm (C). Study Design The study comprised two phases: (i) ex vivo testing and (ii) in‐vivo trial. In the first phase, artificial stones (AS) of 1 cm3, prepared from BegoStone plaster (Bego GmbH & Co. KG, Bremen, Germany) and water according to published methods [16], were embedded centrally in fresh porcine kidney tissue samples. Subsequently, four types of human stones (HS) were positioned in the same manner and tested using higher velocities of the translation stage. The second phase represented an in vivo validation. HS chosen from five different compositions were surgically placed in the bladder, ureter, and kidney of a female Yorkshire pig under general anaesthesia. The stones were treated with a 7.5‐F flexible ureteroscope until they were reduced to fine dust. The primary endpoint of the study was to assess the capability of the SSM for real‐time stone recognition and preventing energy emission when fibre is facing tissue or is distant ≥2–4 mm from the stone surface. Secondary endpoints included: (i) evaluation of ablation efficiency, quantified by SA efficiency with SSM activated vs deactivated; (ii) evaluation of collateral soft tissue damage caused by false‐positive laser activation on tissue; (iii) qualitative assessment of stone debris based on the completeness of dusting; and (iv) evaluation of the system's diagnostic performance in distinguishing stone from tissue, expressed as sensitivity and specificity. Sensitivity reflects the system's ability to correctly detect stone surfaces (true‐positive recognitions among all actual stone targets), whereas specificity indicates its ability to correctly identify soft tissue (true‐negative detections in which laser emission was appropriately inhibited). All procedures were performed in compliance with the relevant animal care legislation (Guide for the Care and Use of Laboratory Animals and the Animal Research: Reporting of In Vivo Experiments [ARRIVE] guidelines). Ethical approval was obtained from the United States Department of Research Institutional Animal Care and Use Committee. Given the pilot design, sample size was not based on hypothesis testing. The ex vivo and in vivo sample sizes were selected to demonstrate technical feasibility across representative scenarios and to inform precision‐based planning of a subsequent confirmatory validation study. The Ex Vivo Experimental Setup A 9.0‐F flexible ureteroscope (Innovex Medical Co., Ltd., Shanghai, China) with 0.2 mm fibre diameter was mounted on a two‐dimensional motorised stage (Standa, Vilnius, Lithuania) at a fixed 90° angle to the stone surface (Fig. S2). Irrigation was maintained at 20 mL/min (gravity). The fibre tip was cleaved before each test. In the first trial, duplicates of the AS were treated using the following laser settings generating an average power of 10 W: 0.1 J × 100 Hz, 0.5 J × 20 Hz, 1 J × 10 Hz and 2 J × 5 Hz. For each setting, one of the duplicates was treated at a fibre scanning speed of 1 mm/s and the other duplicate was treated at 3 mm/s. The scope tip traced five, vertically separated horizontal scans over both stone and tissue while the foot pedal was continuously depressed. High‐resolution images of the samples were acquired using a macro‐lens digital camera immediately after scan completion. Ablated areas of both stone and tissue were quantified. In the second trial, four types of HS—uric acid (UA), calcium oxalate monohydrate (COM), brushite (BR) and cystine (CY)—were embedded centrally within porcine tissue. For each composition, two scanning velocities (2 and 5 mm/s) were tested using a laser setting of 0.5 J × 20 Hz (10 W). Five cuts were performed on stone surface for each sample. Ablation patterns were recorded with SSM deactivated (OFF) and re‐activated (ON). At the end of the procedure, stone and tissue surfaces were analysed and compared to assess ablated areas. The In Vivo Trial In a Good Laboratory Practice (GLP)‐compliant animal study, the same SSM‐equipped TFL system was evaluated in two female Yorkshire pigs under general anaesthesia. Different typologies of HS—comprising UA, COM, BR, CY, and struvite (STR)—were surgically implanted into the bladder, bilateral ureters, and kidneys. Laser lithotripsy was performed using a 7.5‐F flexible ureteroscope (ZhuHai Pusen Medical Technology Co., Ltd., Guangdong, China) coupled with the TFL system integrated with the SSM and a 200‐μm laser fibre. Each anatomical site was treated with one of two laser settings delivering 10 W and 5 W of power, corresponding to high‐frequency (0.1 J x 100 Hz) and low‐frequency (1 J × 5 Hz) configurations, respectively. The objective was complete fragmentation of the samples into dust particles with dimensions of <500 μm with standard dusting technique [17]. All dust was evacuated via the cystoscope outflow channel. At the end of each procedure, the bladder and upper urinary tract were re‐examined endoscopically by independent surgeons to document residual fragments and assess mucosal injury. Sensitivity and specificity were determined for each stone/organ/laser setting by comparing SSM's decision to fire the laser with a ground truth determination based on three independent, experienced endourologists viewing the endoscopic video post procedure and assessing whether the laser should fire or not on a frame‐by‐frame basis, that is, if the fibre is in quasi‐contact with the stone. The output of SSM's laser status, to fire the laser or not, occurred at 500 Hz output and was synchronised with the video. The SSM laser status for each video frame was determined by the status majority and the result compared with the evaluators’ majority decision for the corresponding video frame. Data interpretation and statistical analyses were performed by the academic investigators independently of the industrial partner involved in the technological development of the system. Sensitivity and specificity estimates were complemented with 95% CIs calculated using the Wilson method. Inter‐rater agreement among the three independent reviewers was assessed using Fleiss’ kappa coefficient. Results The Ex Vivo Tests Overall, eight experiments were performed on the AS phantoms (Table 1). With SSM activated, soft‐tissue ablation was completely avoided in all experiments. Table 1Results of the ex vivo trials on AS showing the reduction of soft‐tissue damage and ablated stone surface area under continuous laser emission with activation of the SSM.Laser settingSpeed, mm/sReduction of soft tissue damage, %Reduction of ablated stone surface area, %0.1 J × 100 Hz = 10 W110040.1 J × 100 Hz = 10 W310050.5 J × 20 Hz = 10 W110030.5 J × 20 Hz = 10 W310031.0 J × 10 Hz = 10 W110021.0 J × 10 Hz = 10 W310022.0 J × 5 Hz = 10 W110022.0 J × 5 Hz = 10 W31002Each laser setting was tested in four replicates. With the SSM ON, soft‐tissue injury was completely avoided (100% reduction) across all laser parameters. Results of the ex vivo trials on AS showing the reduction of soft‐tissue damage and ablated stone surface area under continuous laser emission with activation of the SSM. Each laser setting was tested in four replicates. With the SSM ON, soft‐tissue injury was completely avoided (100% reduction) across all laser parameters. The SA efficiency was only slightly reduced with the SSM activated. At 0.1 J × 100 Hz, the ablated surface area decreased by 4% at 1 mm/s and by 5% at 3 mm/s. At 0.5 J × 20 Hz, the reduction was 3% at both speeds. For 1 J × 10 Hz and 2 J × 5 Hz, the reduction was 2% across both velocities. Overall, the mean SA reduction with the SSM activated was 2.75% at 1 mm/s and 3% at 3 mm/s. For the second ex vivo trial, sensitivity and specificity values for the four tested HS are reported in Table 2. Sensitivity was consistently 100% with the SMM ON vs OFF. Specificity was slightly reduced, ranging from 98% for UA to 95% for BR during ablation. High resolution images permitted to evaluate the difference between the ablation patterns (Fig. 3). Table 2Sensitivity and specificity of the SSM for real‐time stone recognition across four stone compositions ex vivo (UA, COM, BR, CY) at scanning speeds of 2 mm/s and 5 mm/s.Stone typeUACOMBRCY2 mm/s5 mm/s2 mm/s5 mm/s2 mm/s5 mm/s2 mm/s5 mm/sSensitivity, %100100100100100100100100Specificity, %9898979795959898 Sensitivity and specificity of the SSM for real‐time stone recognition across four stone compositions ex vivo (UA, COM, BR, CY) at scanning speeds of 2 mm/s and 5 mm/s. Comparison of ablation patterns between two different stone types (COM and CY) with and without activation of the SSM. The translation stage performed the same automatically programmed motion across the stone surface, completing five passes for each stone sample. The In Vivo Trial Endoscopic SA was successfully performed for all implanted stone samples, with the SSM activated throughout each procedure. Inter‐rater agreement among the three reviewers was high, with a Fleiss’ kappa coefficient of 0.84. Complete ablation was achieved in every case, and collateral mucosal injury was consistently graded as none to minimal. Laser application was continued until only dust and residual fragments <500 μm remained. Overall, in the in vivo model, sensitivity ranged from 90% (95% CI 87.9–91.4%) to 100% (95% CI 99.5–100%), while specificity ranged from 85% (95% CI 82.3–86.6%) to 100% (95% CI 99.7–100%) (Table 3). When pooling all evaluated frames across conditions, overall sensitivity was 98.5% and overall specificity was 94.6%. The CIs were generally narrow, reflecting the large number of evaluated frames. Table 3Summary of the in vivo experimental results obtained using the SSM equipped TFL system in two female Yorkshire pigs.Anatomical siteLaser settingStone compositionSpecificity, % (95% CI)Sensitivity, % (95% CI)Bladder1 J × 5 HzUA98 (96.9–98.3)100 (99.8–100)0.1 J × 100 HzUA100 (99.7–100)96 (94.9–96.9)Right ureter1 J × 5 HzBR85 (82.3–86.6)100 (99.7–100)0.1 J × 100 HzCY86 (85.3–87.6)90 (87.9–91.4)Left ureter1 J × 5 HzSTR99 (98.4–99.6)99 (98.2–99.1)0.1 J × 100 HzCOM97 (95.9–97.7)98 (96.7–98.4)Right kidney1 J × 5 HzCY100 (99.8–100)100 (99.8–100)0.1 J × 100 HzBR95 (93.7–95.4)99 (98.2–99.3)Left kidney1 J × 5 HzCOM100 (99.7–100)100 (99.7–100)0.1 J × 100 HzSTR95 (94.0–96.0)99 (98.3–99.2)The HS were surgically implanted into the bladder, bilateral ureters, and kidneys. Laser lithotripsy was performed using high (0.1 J × 100 Hz) and low (1 J × 5 Hz) frequency settings. Sensitivity and specificity are reported with 95% CIs (Wilson method). Summary of the in vivo experimental results obtained using the SSM equipped TFL system in two female Yorkshire pigs. The HS were surgically implanted into the bladder, bilateral ureters, and kidneys. Laser lithotripsy was performed using high (0.1 J × 100 Hz) and low (1 J × 5 Hz) frequency settings. Sensitivity and specificity are reported with 95% CIs (Wilson method). The high‐frequency setting (0.1 J × 100 Hz) demonstrated lower specificity across most anatomical locations, with the exception of the bladder. Conversely, sensitivity remained high across all conditions, reaching 100% (95% CI 99.7–100%) with the low‐frequency (1 J × 50 Hz) setting and up to 96% (95% CI 94.9–96.9%) with the high‐frequency setting (Table 3). In the ureter, the system maintained satisfactory diagnostic accuracy despite a reduction in specificity, with values of 85% (95% CI 82.3–86.6%) for BR and 86% (95% CI 85.3–87.6%) for CY on the right side, increasing to 97% (95% CI 95.9–97.7%) and 99% (95% CI: 98.4–99.6%) for COM and STR on the left, respectively. Sensitivity remained high across all conditions, ranging from 90% (95% CI 87.9–91.4%) to 99% (95% CI 98.2–99.1%). In the kidney, diagnostic performance was consistently high across all compositions, with specificity ranging from 95% (95% CI 93.7–95.4%) to 100% (95% CI 99.8–100%), and sensitivity reaching up to 100% (95% CI 98.8–100%) across all tested conditions. Discussion This study represents the first pre‐clinical evaluation of a laser emission control system based on back reflected‐light spectral recognition—named StoneSense™—integrated into a TFL platform. The system continuously analyses optical signals emitted and collected through the surgical laser fibre, enabling real‐time identification of stone surfaces and automatically interrupting energy delivery when the fibre tip was either too distant from the target or directed toward tissue rather than stone. Our findings demonstrate that light‐guided lithotripsy using TFL energy is feasible, achieves effective dusting, and prevents collateral tissue injury in both ex vivo and in vivo settings. Previous studies have investigated optical feedback systems for stone‐tissue discrimination during laser lithotripsy. Lange et al. [18] first demonstrated that autofluorescence emitted by urinary calculi under specific laser excitation could reliably differentiate stones from urothelial tissue. Building on this principle, Schlager et al. [19] developed a Ho:YAG ‐based real‐time recognition system that superimposed excitation light onto the treatment beam and collected the returning signal through the same fibre. The in vitro experiments confirmed high accuracy in detecting stone contact, with laser firing automatically gated to valid stone signals only. The in vivo porcine trials showed effective prevention of tissue injury and a reduction in total laser energy use of up to 50% in the kidney and 27% in the ureter, without compromising lithotripsy efficiency [20]. Building on these earlier works, the present system expands the concept of optical feedback by employing reflected‐light spectral analysis rather than autofluorescence. This approach relies on measurement of characteristic wavelengths of the light reflected by the stone surface during laser activation, allowing selective emission control based on real‐time spectral feedback [21]. In the AS‐embedded setting, tissue ablation occurred in the ex vivo experiments, while stone dusting efficiency was only minimally affected, with a maximum reduction of 5% at 100 Hz. The system reliably interrupted TFL firing when the fibre deviated from the stone surface, even during continuous movement at speeds of 1–3 mm/s. For the HS samples, increasing the velocity of the translation stage up to 5 mm/s did not affect the precision of the SSM platform. Energy delivery was consistently interrupted when the SSM failed to detect back reflection from the stone surface. Conversely, specificity was slightly reduced to 95%, indicating only a negligible impact on the overall SA rate. In vivo, the SSM system demonstrated high level of safety, with no macroscopic or endoscopic evidence of mucosal injury observed in any treated site. The IPG TFL integrated SSM consistently achieved complete stone lithotripsy, producing fine stone dust with fragments <500 μm. Diagnostic accuracy remained robust across all anatomical locations, with both sensitivity and specificity exceeding 85% under all tested conditions. A slight reduction in specificity was observed during ureteric lithotripsy at both energy settings. This finding likely reflects the intrinsic challenges of performing laser lithotripsy in the porcine ureter [22], as well as of assessing ground truth required to calculate sensitivity and specificity. The CIs remained within acceptable bounds, supporting the robustness of the system's performance. However, the combined effects of limited endoscope mobility, visibility, tissue proximity, and narrow luminal space can introduce ambiguity and error in the ground truth assessment. An additional contributing factor might be related to stone composition. BR and CY are known to be among the hardest urinary stone types for laser lithotripsy [23]. Their interaction with TFL energy in the ureter can generate fragments that are not immediately cleared, thereby reducing target visibility. These findings highlight the importance of anatomical context in optical recognition systems and suggest that further optimisation in challenging environments may enhance overall accuracy. The fibre–stone distance is critical for both Ho:YAG and TFL. At distances >2 mm, energy absorption by irrigation fluid dominates and reduces efficiency [24]. Although TFL produces low‐energy pulses that promote efficient dusting, these settings can also lead to elevated intrarenal temperatures during prolonged activation [25]. In addition, the generation of fine dust may impair endoscopic visibility and increase the risk of inadvertent tissue contact. By gating energy delivery to real‐time stone detection, light‐based feedback minimises unnecessary lasing in fluid‐filled gaps and prevents unintended tissue exposure, thereby reducing both thermal and mechanical injury. Safety remains a critical issue in endourological laser surgery. Dau et al. [26] showed that 90–96% of laser energy is dissipated as heat into irrigation fluid, leading to potentially hazardous temperature rise in poorly cooled environments. Ventimiglia et al. [27] demonstrated that powers above 20 W in the kidney, or 10 W in the ureter, exceeded safety thresholds, with tissue damage observed even at 1–5 mm from the fibre tip. Furthermore, no significant difference between Ho:YAG and TFL have been found in safety thresholds when irrigation was maintained. While increasing irrigation flow may reduce thermal rise, this must be balanced against the risk of elevated intrarenal pressure [28]. The applicability of light‐based recognition is particularly relevant in the ureter, where safety margins are narrow. Experimental studies have shown that high‐frequency, high‐energy TFL settings can produce more pronounced ureteric injury, especially in inexperienced hands. This emphasises the need for cautious parameter selection and appropriate training [29]. Taken together, current evidence indicates that SA efficiency and patient safety depend on a careful balance of laser power, irrigation, and intrarenal pressure. By selectively gating energy delivery with stone contact, SSM can provide an additional safeguard to reduce unnecessary lasing and to reduce the risk of thermal injury in the urinary system. The integration of optical feedback systems, such as the SSM, reflects a broader trend toward the incorporation of intelligent technologies in endourology. Recent advances in artificial intelligence (AI) and machine learning are increasingly being explored to assist clinical decision‐making, automate image interpretation, and improve procedural safety. AI‐driven systems have already demonstrated potential in urology for patient monitoring, diagnostic support, and surgical guidance, including applications such as AI‐driven chatbots and automated clinical assistance tools that may enhance patient care pathways and physician workflow efficiency [30]. In this context, real‐time optical recognition platforms represent an early step toward more autonomous or semi‐autonomous endoscopic systems capable of assisting surgeons during lithotripsy by optimising energy delivery and reducing procedural risks, particularly in the context of expanded indications of stone surgery through suction systems [31]. Several limitations of this approach should be acknowledged. First, illumination intensity is affected by fibre quality, and progressive tip degradation during lithotripsy may reduce detection accuracy over time [32]. Progressive fibre tip wear may alter light transmission and reflection properties, which could potentially influence the optical recognition accuracy of the SSM. Future investigations should therefore assess the stability of the system's performance over cumulative lasing time and delivered energy. Second, system performance was evaluated in controlled ex vivo and porcine models, which may not fully replicate the complexity of human endoscopic surgery, including variable intrarenal pressures, tissue optical properties, bleeding and severe turbidity, which may reduce optical signal quality. Third, the system currently requires calibration against mucosa at the start of each procedure, which may limit its practicality in emergency or highly dynamic intraoperative settings. Fourth, while light gating reduces unnecessary laser activation, it does not address all mechanisms of thermal injury, such as cumulative heating during prolonged fURS or excessive irrigation pressures required to maintain cooling. Fifth, the experimental pilot design of this study and the consequent limited sample size restrict the precision of the sensitivity and specificity estimates. Therefore, these results should be interpreted as preliminary exploratory findings that require further confirmation. In the present study, intraluminal temperature was not monitored during laser activation. Finally, although the emission‐gating mechanism of the SSM aims to reduce both off‐target laser firing and unnecessary energy dissipation into the irrigation fluid, the present experimental design did not include direct intraluminal temperature measurements. Therefore, the potential impact of this system on thermal dose and intrarenal temperature dynamics could not be directly quantified, which may limit the generalisability of these findings. Future clinical trials will be necessary to determine how light‐based systems integrate into routine practice, whether they reduce complication rates in complex ureteric and renal stone cases, and how they compare with other adjunctive safety strategies, such as temperature monitoring or pressure‐regulated irrigation. Conclusions Laser emission control based on reflected‐light spectral recognition is both feasible and effective when integrated into a TFL platform. Real‐time detection of stone surfaces enables automatic interruption of laser emission in the absence of a valid target, thereby reducing unnecessary energy delivery and minimising liquid overheating and potential tissue injury. The in vivo results highlight the system's translational potential, supporting its future clinical application. Further clinical studies are warranted to validate these findings and confirm the role of this technology in enhancing the safety and efficiency of endoscopic stone surgery. Author Contributions Study conception and design: Olivier Traxer, IIya V. Yaroslavsky, Gregory Altshuler; Experimental setup and data acquisition: Anastasia Kovalenko, Viktoriya Andreeva, James Childs, Alexey G. Martov. Data analysis and interpretation: IIya V. Yaroslavsky, Anastasia Kovalenko, Federico Zorzi, Viktoriya Andreeva, James Childs, Gregory Altshuler. Manuscript drafting: Federico Zorzi. Critical revision of the manuscript: Olivier Traxer, IIya V. Yaroslavsky, Alexey G. Martov. Supervision: Gregory Altshuler. Disclosure of Interests Olivier Traxer is a consultant for the IPG Medical Corporation. IIya V. Yaroslavsky, Anastasia Kovalenko, James Childs and Gregory Altshuler are employees of the IPG Medical Corporation and shareholders of IPG Photonics. The other authors do not declare any conflicts of interest in writing this paper. Funding This research received no external funding. The prototype laser system IPG Medical Thulium Fibre Laser with StoneSense™ module was provided by IPG Medical Corporation (Marlborough, MA, USA) for pre‐clinical testing. Ethics Statement All experimental procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals and the ARRIVE guidelines. The in vivo animal trial was carried out on female Yorkshire pigs under general anaesthesia at a GLP‐compliant facility (IRCAD ‐ Institute de Recherche sur les Cancers de l’Appareil Digestif; 1, Place de l'Hôpital 67 091 Strasbourg, France). The in vivo protocol was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC). Supporting information Fig. S1. Interaction between surgeon, the SSM and TFL energy source. Laser emission is enabled only when surgeon intent through the footswitch pedal and real‐time optical stone detection are simultaneously confirmed. Fig. S2 The ex vivo experimental setup. The HS—embedded at the centre of a fresh porcine kidney tissue sample—were scanned with a ureteroscope and a 200‐μm laser fibre mounted on the two‐dimensional (2D) motorised stage under constant irrigation. The experiment was repeated with the SSM ON and OFF to assess the system sensitivity and specificity.