**4. Discussion**

The introduction of i-OCT-integrated surgical microscopes might represent a further step toward a safer and more efficient surgery. To date, different i-OCT systems can be integrated into an ocular microscope, providing useful feedback for the surgeons both the anterior and posterior segmen<sup>t</sup> surgeons [22,23].

When dealing with AS procedures, this technology provides direct visualization of anatomic structures before, during, and after surgical maneuvers, allows for an analysis of surgical planes, guides surgical steps, and helps to detect intraoperative complications, eventually impacting surgical decision-making [5,24,25].

In their study, Tañá-Sanz et al. demonstrated that some AS parameters (ACD, CCT, LT, and WTW) obtained with the AS-SD-OCT integrated into the Catalys femtosecond laser platform differed from those derived from SS-OCT biometers (IOLMaster 700 and Anterion biometer). According to the authors, these differences could be related to patients' position and the mydriasis required for the surgery [11].

In addition to these parameters, which can be easily acquired by traditional biometry, a greater understanding of lens anatomy (including the dimensions of the aged crystalline lens and its capsule) could be useful for surgeons and the development of new IOL formulas and technologies. However, most studies have been conducted in research settings, applying customized devices and not commonly available instruments (such as Magnetic Resonance Imaging), consequently, results couldn't be easily applied to clinical practice. Moreover, the ability of more commonly available biometric data to predict LD and LV is quite limited. The introduction of i-OCT integrated on femtosecond laser platforms has facilitated the study of lens anatomy in larger data sets making new lens parameters, consequently, available [26]. In their work, Waring et al. showed that i-OCT could detect LV, LD, and LT and they provided regression equations to predict LD and LV from conventionally available parameters. The authors stated this additional info could help in effective lens position (ELP) estimate/ion (consequently improving IOL power calculation

and enhancing refractive predictability) and, in new IOL technologies development, such as capsule refilling [12].

Indeed, the prediction of the IOL position after surgery still represents one of the main issues when dealing with IOL power calculation [27,28].

Hirnshall et al. analyzed if intraoperative lens measurements, instead of preoperative ACD measurements, could improve ELP evaluation. It was found that the position of the anterior capsule after the insertion of a CTR represented an excellent predictor of ACD before surgery. However, it must be stated that to acquire these values, the use of a CTR was required (which is not an ordinary step in uncomplicated surgery) and that preoperative AL and ACD were also associated with high VIP for the prediction of ACD measured three months after surgery [13].

In cataract surgery, i-OCT might represent a valid device both for standard phacoemulsification procedures and for FLACS. Its main applications include the visualization of corneal incisions and the stromal hydration, the assessment of hydro-dissection, perception of the trenching depth, and identification of lens positioning [10,29]. Thus, the use of i-OCT might allow a safer surgical procedure, decreasing the rate of postoperative wound leak and hypotony and preventing any iatrogenic capsular rupture during hydro-dissection and phacoemulsification [22,24].

Titiyal et al. compared the morphology of CCIs in conventional phacoemulsification and FLACS using i-OCT; they noticed that a ragged slit morphology was a significant predictive factor for incision site DMD and it occurred more frequently during conventional surgery. Interestingly, the authors stated that all DMDs detected by i-OCT were also detectable under the operating microscope before stromal hydration; however, an increase in the extent of DMD or the occurrence of DMD after stromal hydration (which represented the phase in which the higher rate of DMD occurred—83.7% cases) were only detected by i-OCT. At any rate, all DMDs solved spontaneously in one month without requiring additional surgery [16]. The ability to detect an early subclinical DMD, an epithelial disruption, or a microtear in the inner or outer lip of the wound intraoperatively could be of grea<sup>t</sup> value for the surgeon not only to modify the subsequent steps of surgery but also to manage the early post-operative period [10].

The location of the Continuous curvilinear capsulorhexis (CCC) is critical for visual outcomes [30,31]. In conventional cataract surgery, the procedure is guided by the position of the PC and the LC, which are easily detected using a microscope; in FLACS, they can automatically be detected, together with an additional parameter called lens center. This represents very interesting data since the IOL center position will be similar to the center of the crystalline lens. A precisely sized and centered capsulotomy, enabled by this method, might improve predictability and control of the IOL placement reducing IOL tilting and decentration. Song et al. analyzed the relative location of and distance between the PC, the LC, and the lens center in patients who underwent FLACS. It was found that the PC was closer to the lens center than the LC whose X and Y coordinate position was significantly inferior and temporal compared to the PC [17,32].

Palanker et al. compared the size and the shape of laser capsulotomy to manual ones using a system combining FD-OCT with a femtosecond pattern scanning laser. They demonstrated that the former was characterized by size more similar to the intended one than the latter; moreover, they were more circular than manual ones [15].

Mastropasqua et al. analyzed the characteristics of capsulotomies obtained during two types of i-OCT guided FLACS platforms (Lensx and Lensar) and during a standard manual technique. Laser-made capsulotomies demonstrated significantly better circularity than the manual CCCs at seven days, their sizes were much more similar to the intended ones, and they showed greater IOLs centration than the manual group at all time points [18].

As for biometry images, optical opacity could also affect the quality of i-OCT images, leading to misleading analysis. During FLACS, precise detection of radiation sites is critical to correct the direction of spots and to avoid complications. Kurosawa et al. demonstrated

that LT inspection could guide the surgeon in adjusting laser settings and avoiding posterior capsule breaks [14].

Moreover, real-time visualization of the trenching depth during phacoemulsification could be very useful for surgeons in training to decide the exact location to crack the nucleus during divide and conquer techniques [10].

Many authors have underlined the importance of i-OCT in complicated cases [33]. When dealing with white cataracts, the direct visualization of lens anatomical features through i-OCT could help anticipate the intraoperative dynamics of spontaneous milky fluid release, thus letting the surgeon be ready to deal with possible complications, especially during capsulorhexis [19].

For traumatic cataracts or PPCs, i-OCT could identify a capsular defect preventing further complications for the surgeon [20,29,34]. PPCs still represent a surgical challenge [35], due to the high incidence of posterior capsular break. To prevent it, hydrodissection is commonly avoided, consequently requiring greater manipulations during cortical clean-up and longer surgical time. In their study, Titiyal et al. evaluated morphological characteristics and intraoperative dynamics of PPCs with i-OCT, demonstrating that in the case of an intact posterior capsule homogenously spaced from the posterior polar opacity (called "type I PPC") gentle hydrodissection could be safely performed. At any rate, the authors declared the preoperative AS-OCT features correlated with the intraoperative ones. Moreover, according to the authors, i-OCT use didn't reduce the incidence of posterior capsule dehiscence compared to not i-OCT-guided surgeries [20].

I-OCT could also be helpful in patients with ectopia lentis, preventing further corneal endothelium damage during lens removal [36,37].

Juergens et al. reported i-OCT to be crucial for the implantation of a two-part brown iris diaphragm, because of the poor contrast between the anterior lens capsule margin and the brown implant [21].

Interestingly, i-OCT could detect the presence of direct intraoperative communication between Berger space and anterior chamber, which might lead to excessive fluid flow through this segmen<sup>t</sup> causing anterior displacement of the posterior capsule thus increasing the risk for a posterior capsular break and iris prolapse. Anisimova et al. showed that i-OCT could identify the presence of lens micro fragments and cellular material within the Berger space for the discontinuity of the zonules and Wieger ligament. with a higher sensitivity than postoperative OCT. Furthermore, they hypothesized that Wieger ligament detachment was associated with increased zonular permeability. This observation could be useful to clarify the mechanism of acute aqueous misdirection syndrome also known as acute rock-hard eye syndrome (AIRES) [6].

Although i-OCT might represent a helpful and not invasive tool, its application in clinical practice presents several limitations for cataract surgery. Firstly, intraoperative measurements are still time-consuming. Secondly, OCT-friendly instruments to reduce shadowing and integrated calipers are still lacking [10,13]. Moreover, total cataracts or extremely dense nuclear sclerosis reduce the ability of i-OCT to visualize the posterior capsule [19,20]. Finally, the analysis of the intraoperative images is not automatic, and it is still influenced too much by the insights of the observer.
