Image guided vessel preparation

Abstract

The present disclosure provides systems and methods for Al model(s) that are trained to identify characteristics associated with lesions within a blood vessel and provide a suggested procedure associated with the lesion. The Al model may receive, as input, intravascular and / or extraluminal image data associated with the blood vessel and provide, as output, characteristics, e.g., characterizations, associated with the blood vessel. The characterizations may include, for example, form or tissue nature (e.g., microcalcification, punctate, fragment, sheet, nodular, or otherwise), location within the vessel (e.g., intimal, superficial, medial, deep, etc.), size (e.g., length, depth, thickness, etc.), severity, and / or the like. The characterizations may be used as input to the same and / or a different model for providing a suggested procedure associated with the lesion. The suggested procedure may include, for example, orbital atherectomy, intravascular lithotripsy, rotational atherectomy, cutting and / or scoring balloons, and / or the like.

Description

Image Guided Vessel PreparationCross-Reference To Related Applications

[0001] The present application claims the benefit of the filing date of U.S. Provisional Application No. 63 / 667,166, filed July 3, 2024, entitled Image Guided Vessel Preparation, the disclosure of which is hereby incorporated herein by reference.Technical Field

[0002] The technology disclosed herein is directed to the use of intravascular imaging to provide vessel condition data for determining vessel preparation and treatment strategies.Background

[0003] Percutaneous coronary intervention (PCI) planning uses visualization techniques such as intravascular ultrasound (IVUS), intravascular optical coherence tomography (OCT), micro-OCT, and near- infrared spectroscopy (NIRS). OCT is an intravascular imaging modality that uses near-infrared light to provide high-definition, cross-sectional and three-dimensional images of the vessel microstructure during PCI. An OCT catheter emits near-infrared light to the vessel to produce high-resolution real-time images of the vessel. To perform an imaging scan known as an OCT pullback, an OCT catheter is inserted into the vessel and an infrared laser is used to scan the vessel wall in a helical or spiral-like manner, such as rotating about an axis while moving longitudinally along the axis. OCT components include an OCT designed catheter and an imaging system for receiving the image data captured by the OCT probe, processing the image data, and providing the images for display.

[0004] Patients with inadequate stent expansion are at high risk for adverse outcomes, such as stent thrombosis and in-stent restenosis. Several publications propose an OCT based quantitative scoring system to predict future stent expansion. See Fujino, A. et al., A New Optical Coherence Tomography-based Calcium Scoring System to Predict Stent Under-expansion, EuroIntervention 13, e2182-e2189 (2018). The use of machine learning models and development have also been proposed to identify vessels at risk of under-expansion that would be candidates for atherectomy lesion preparation. See Gharaibeh et al, Prediction of Stent Under-expansion in Calcified Coronary Arteries Using Machine Learning on Intravascular Optical Coherence Tomography Images, Sci Rep 13, 18110 (2023). https: / / doi.org / 10.1038 / s41598-023-44610-9.

[0005] Artificial intelligence (Al) models can be trained to identify calcification of the vessel. Lesion calcification can be characterized based on the vascular bed or location within the body, including such vascular beds as coronary, peripheral, carotid, renal, and others. In general, the types of treatment regimens for particular calcified lesions include orbital atherectomy, rotational atherectomy, laser, and intravascular lithotripsy. Each treatment is typically chosen based on vessel characteristics.

[0006] Intimal calcification is one of two main categories of vascular calcification. Intimal calcification is associated with atherosclerotic plaques that are believed to result from modified lipid accumulation, pro- inflammatory cytokines, and apoptosis within the plaque that induce osteogenic cell differentiation. The result of atherosclerosis is typically luminal stenosis. Intimal calcification, being associated withatherosclerosis, is characterized by appearance as patchy or spotty areas of calcification on the luminal side of internal elastic lamina of the vessel and commonly affects coronary, carotid, and larger arteries. Advanced forms of intimal calcification can lead to vessel obstruction or occlusion.

[0007] Medial calcification is considered the second of two main categories of vascular calcification. Medial calcification tends to be more widespread in the lower abdominal region and limbs and results from the osteogenic differentiation of vascular smooth muscle cells within the medial layer of the vessel wall. Calcium accumulation begins as an amorphous mineral deposit and undergoes progressive remodeling, and ultimately may progress to mineralizing into mature bone. Medial calcification is more common in small and medium-sized arteries and can result in increased arterial stiffness. Although medial calcification is generally not associated with luminal obstruction, the decrease in the arterial vessel wall elasticity and compliance may ultimately lead to atherosclerosis, reduced perfusion, and eventually peripheral arterial disease (PAD), and in advanced forms, left ventricular hypertrophy and / or heart failure . Medial calcification is commonly associated with diabetes, chronic kidney disease, and metabolic disfunction and is characterized as above along with elastin layer disruption and calcification. Such medial calcification can appear as smooth continuous areas of calcification, typically seen as parallel lines or “railroad” tracks on x-rays.

[0008] The presence of vascular calcification, whether intimal or medial, can hinder achievement of optimal outcomes following conventional interventional therapies. Intimal calcification can obstruct or occlude vessel lumen passage and make it difficult to cross that portion of the vessel, and crossing such obstructed or occluded lesions can result in over-stretch injury that predisposes to restenosis. Both intimal and medial calcification can also impair the ability to achieve full device expansion for restoring flow and contribute to higher vessel recoil, both of which result in higher residual stenosis. . Vascular calcification is a barrier to drug diffusion further contributing to the potential for restenosis.

[0009] It can be difficult to fully expand a stent in a region of the vessel in which there is a heavily calcified lesion. Once implanted in tissue highly resistant to dilation, there is no opportunity to apply a pre-stent treatment option or regimen such as laser ablation, rotational atherectomy, a scoring balloon, an angioplasty balloon or orbital atherectomy. Post stent implantation, the only option is risky, high-pressure (e.g. up to 30 atm) balloon dilations with the attendant risk of vessel rupture.

[0010] Specialized balloon-based technologies include cutting and scoring balloons and are generally used in eccentric calcium. Cutting balloons consist of a number of micro-blades mounted on a balloon, and scoring balloons consist of a semi-compliant balloon around which several nitinol wires are wrapped. Both make incisions into the calcium and improve vessel compliance, allowing dilation. Their designs allow them to grip the calcium, resulting in less slippage, also known as “melon seeding”, which avoids dissection of the adjacent vessel. However, in the presence of severe calcification, cutting balloons have been found to have less procedural success than rotational atherectomy (RA), although they do have utility when used as an adjunct to RA. Very-high-pressure balloons consist of a twin-layered, noncompliant balloon with a rated burst pressure of approximately 30-35 atm. This technology has its place as an adjunct to other techniques, such as along with adjunctive RA.

[0011] Devices for intravascular lithotripsy (IVL) consist of a balloon-based delivery system containing a number of emitters that generate short high-energy electric sparks. Although the balloon itself is dilated to only 4 atm, each short-lived pulse delivers a localized spike in pressure just outside the balloon. To date, IVL has been more efficient in concentric calcification lesions. However, LVL can still be effective in eccentric calcification, but it requires more pulses when the calcification lesions are eccentric only.

[0012] IVL devices can include electrodes or lithotripsy emitters that create acoustic waves by arcing discharges between electrode components but may also include devices that create acoustic energy within the balloon via laser energy sources. Examples of such laser systems are described in U.S. Pat. Nos. 11,058,492 and 11,246,569 (the entire contents of which are incorporated by reference). Examples of electrically induced systems are described in U.S. Pat. Nos. 8,728,091, 9,642,673 and 10,850,078 and Published U.S. Pat. Appl. No. 2022-0054194 (the entire contents of which are incorporated by reference).

[0013] Rotational atherectomy (RA) uses a diamond-tipped burr rotating at very high speeds (140,000- 160,000 rpm) and resulting in differential ablation of calcified lesions. RA was previously used for aggressive debulking of the calcium, which led to a number of complications, including no-reflow (from embolization of particulate matter) and vessel perforation. However, modifications to RA techniques such as shorter RA runs, the use of a pecking motion at the lesion, smaller burr sizes, and the combination of adjunctive, complementary techniques, inter alia, improve usage.

[0014] Orbital atherectomy (OA) comprises an eccentrically mounted, diamond-coated crown that uses centrifugal force to orbit (at 80,000 or 120,000 rpm), resulting in preferential calcium sanding while flexing away from elastic healthy tissue. With OA, a reduction in diameter stenosis to 50% in > 98% of lesions is desired. OA is preferred to RA in larger vessels with concentric or nodular calcification due to the wider rotational orbit and deeper calcium modification achieved with OA.

[0015] Greater specificity and techniques for using OCT are described within published US patent applications 2020 / 0294659 Al published on September 17, 2020, 2021 / 0042927 Al published on February 11, 2021, and 2023 / 0054891 Al published on February 23, 2023, the entire contents of each of these publications being fully incorporated by reference herein.Brief Summary

[0016] The disclosure is generally directed to Al model(s) that are trained to identify characteristics associated with lesions within a blood vessel and provide a suggested procedure associated with the lesion. The Al model may receive, as input, intravascular and / or extraluminal image data associated with the blood vessel and provide, as output, characteristics associated with the blood vessel. The characterizations may include, for example, form or tissue nature, location within the vessel, size, severity, and / or the like. The characteristics may be used as input to the same and / or a different model for providing a suggested procedure associated with the lesion. The suggested procedure may include, for example, orbital atherectomy, intravascular lithotripsy, rotational atherectomy, cutting and / or scoring balloons, and / or the like. Treatment of coronary artery disease (CAD) and peripheral artery disease (PAD) requires proficiency in using all types of modification techniques, an understanding of which procedures and / or tools are most appropriate in a given patient, and familiarity with intracoronary imaging use to guide the procedure.Increasingly, a combination of lesion modification techniques are being used in clinical practice. As such, the systems and methods described herein can provide suggested procedures that are efficient and patientspecific in a quick and accurate manner.

[0017] One aspect of the disclosure is directed to a system, comprising one or more processors to receive, from an imaging modality, vessel data associated with at least a portion of a vessel, to determine, based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel, to provide as input into a model the determined initial characteristics of the lesion, to identify, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, and to provide for output the suggested procedure.

[0018] The one or more processors of the system may further receive, from the imaging modality after the suggested procedure is completed, post-procedural vessel data, and determine, based on the postprocedural vessel data, updated characteristics of the lesion.

[0019] The one or more processors of the system may further store, in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics, and update the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

[0020] In some examples, the imaging modality may be optical coherence tomography or intravascular ultrasound.

[0021] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0022] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0023] In some examples, the lesion may be a calcium lesion.

[0024] The one or more processors of the system may further determine, based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and provide for output, based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0025] Another aspect of the disclosure is directed to a method, comprising: receiving, by one or more processors from an imaging modality, vessel data associated with at least a portion of a vessel, determining, by the one or more processors based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel, providing, by the one or more processors, as input into a model the determined initial characteristics of the lesion, identifying, by the one or more processors by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, and providing for output, by the one or more processors, the suggested procedure.

[0026] The method may further comprise receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, post-procedural vessel data, and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

[0027] The method may further comprise storing, by the one or more processors in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics, and updating, by the one or more processors, the model based on the initial characteristics, the suggestedprocedure, and the updated characteristics.

[0028] In some examples, the imaging modality may be optical coherence tomography or intravascular ultrasound.

[0029] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0030] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0031] In some examples, the lesion may be a calcium lesion.

[0032] The method may further comprise determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0033] Yet another aspect of the disclosure is directed to one or more non-transitory computer-readable storage media encoding instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: receiving, from an imaging modality, vessel data associated with at least a portion of a vessel, determining, based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel, providing as input into a model the determined initial characteristics of the lesion, identifying, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, and providing for output the suggested procedure.

[0034] The one or more non-transitory computer-readable storage media may further comprise receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, postprocedural vessel data, and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

[0035] The one or more non-transitory computer-readable storage media may further comprise storing, by the one or more processors in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics, and updating, by the one or more processors, the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

[0036] In some examples, the imaging modality may be optical coherence tomography or intravascular ultrasound.

[0037] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0038] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0039] In some examples, the lesion may be a calcium lesion.

[0040] The one or more non-transitory computer-readable storage media may further comprise: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0041] Yet another aspect of the disclosure is directed to a system, comprising one or more processors, the one or more processors to receive, from an imaging modality, intravascular imaging data of a vessel, to determine, based on the intravascular imaging data, information associated with the vessel, to determine, based on the intravascular imaging data, a presence of a lesion within the vessel, to determine, based on the intravascular imaging data, characteristics of the lesion, to provide as input, into a model, the characteristics of the lesion, wherein the model is trained to identify a suggested procedure based at least in part on the information associated with the vessel and the determined characteristics of the lesion, to identify, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, to provide for output the suggested procedure, to store, in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database, and to update, based on the database, the model.

[0042] The one or more processors of the system may further receive, from the imaging modality after the suggested procedure is completed, post-procedural vessel data, and determine, based on the postprocedural vessel data, updated characteristics of the lesion.

[0043] In some examples, the imaging modality may be optical coherence tomography or intravascular ultrasound.

[0044] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0045] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0046] In some examples, the lesion may be a calcium lesion.

[0047] The one or more processors of the system may further determine, based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and provide for output, based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0048] Yet another aspect of the disclosure is directed to a method, comprising: receiving, by one or more processors from an imaging modality, intravascular imaging data of a vessel, determining, by the one or more processors based on the intravascular imaging data, information associated with the vessel, determining, by the one or more processors based on the intravascular imaging data, a presence of a lesion within the vessel, determining, by the one or more processors based on the intravascular imaging data, characteristics of the lesion, providing as input, by the one or more processors, into a model, the characteristics of the lesion, wherein the model is trained to identify a suggested procedure based at least in part on the information associated with the vessel and the determined characteristics of the lesion, identifying, by the one or more processors executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, providing for output, by the one or more processors, the suggested procedure, storing, by the one or more processors in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database, and updating, by the one or more processors based on the database, the model.

[0049] The method may further comprise receiving, by the one or more processors from the imagingmodality after the suggested procedure is completed, post-procedural vessel data, and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

[0050] In some examples, the imaging modality may be optical coherence tomography or intravascular ultrasound.

[0051] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0052] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0053] In some examples, the lesion may be a calcium lesion.

[0054] The method may further comprise determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0055] Yet another aspect of the disclosure is directed to one or more non-transitory computer-readable storage media encoding instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: receiving, an imaging modality, intravascular imaging data of a vessel, determining, based on the intravascular imaging data, information associated with the vessel, determining, based on the intravascular imaging data, a presence of a lesion within the vessel, determining, based on the intravascular imaging data, characteristics of the lesion, providing as input into a model, the characteristics of the lesion, wherein the model is trained to identify a suggested procedure based at least in part on the information associated with the vessel and the determined characteristics of the lesion, identifying, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, providing for output the suggested procedure, storing in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database, and updating, based on the database, the model.

[0056] The one or more non-transitory computer-readable storage media may further comprise receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, postprocedural vessel data, and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

[0057] In some examples, the imaging modality may be optical coherence tomography or intravascular ultrasound.

[0058] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0059] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0060] In some examples, the lesion may be a calcium lesion.

[0061] The one or more non-transitory computer-readable storage media may further comprise determining, by the one or more processors based on the vessel data, whether the lesion comprises a mediallesion or an intimal lesion, and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0062] Another aspect of the disclosure is directed to a method, comprising: receiving, by one or more processors, vessel data associated with at least a portion of a vessel, determining, by the one or more processors based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel, providing, by the one or more processors, as input into a model the determined initial characteristics of the lesion, identifying, by the one or more processors by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, and providing for output, by the one or more processors, the suggested procedure.

[0063] The method may further comprise receiving, by the one or more processors after the suggested procedure is completed, post-procedural vessel data, and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

[0064] The method may further comprise storing, by the one or more processors in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics, and updating, by the one or more processors, the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

[0065] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0066] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0067] In some examples, the lesion may be a calcium lesion.

[0068] The method may further comprise determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

[0069] Yet another aspect of the disclosure is directed to a method, comprising: updating, by one or more processors based on post-procedural intravascular imaging data of a vessel and based on a database, a model. The model may be trained to identify a suggested procedure based at least in part on determined vessel characteristics and determined characteristics of the lesion. Updating the model may include: receiving, by the one or more processors, the post-procedural intravascular imaging data of the vessel, determining, by the one or more processors based on the post-procedural intravascular imaging data, vessel characteristics, determining, by the one or more processors based on the post-procedural intravascular imaging data, a presence of a lesion within the vessel, determining, by the one or more processors based on the post-procedural intravascular imaging data, characteristics of the lesion, providing as input, by the one or more processors, into the model, the characteristics of the lesion, identifying, by the one or more processors executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel, providing for output, by the one or more processors, the suggested procedure, and storing, by the one or more processors in memory, the determined vessel characteristics, the determinedcharacteristics of the lesion, and the suggested procedure in a database.

[0070] The method may further comprise receiving, by the one or more processors after the suggested procedure is completed, post-procedural vessel data, and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

[0071] In some examples, the characteristics associated with the lesion may comprise at least one of a total angle, a thickness, or a length of the lesion.

[0072] In some examples, the suggested procedure may be an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

[0073] In some examples, the lesion may be a calcium lesion.

[0074] The method may further comprise determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion, and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.Brief Description of the Drawings

[0075] Figure 1 is an example data collection system for use in respect of a blood vessel according to aspects of the disclosure.

[0076] Figure 2 is an example therapy device that can be used in conjunction with the data collection system of Figure 1 and / or based on the suggested procedure determined by the data collection system of Figure 1 , according to aspects of the disclosure.

[0077] Figure 3 is an example method for determining information associated with the vessel and / or characteristics of a lesion based on vessel data received by the data collection system of Figure 1 , according to aspects of the disclosure.

[0078] Figures 4 and 5A are example methods of using OCT as an imaging based modality for training a model, according to aspects of the disclosure.

[0079] Figure 5B is an example flow diagram for training a model to generate suggested vessel preparation options, based at least in part on the vessel data received by the data collection system of Figure 1, according to aspects of the disclosure.

[0080] Figure 6 is an example flow diagram for creating a database of vessel preparation procedures and outputs, based at least in part on the vessel data received by the data collection system of Figure 1, according to aspects of the disclosure.

[0081] Figure 7 is an example flow diagram for creating a database of vessel preparation procedures and outputs, based at least in part on the vessel data received by the data collection system of Figure 1, and executing a model trained on the database, according to aspects of the disclosure.

[0082] Figure 8 illustrates suggested procedures that can be determined by executing the model trained on the database of the data collection system of Figure 1, according to aspects of the disclosure.

[0083] Figures 9-15 illustrate example interface screens which may be implemented, for example, in the context of the data collection system of Figure 1 according to aspects of the disclosure.

[0084] Figure 16 depicts a flow diagram of an example process of providing a suggested procedure using the data collection system of Figure 1, according to aspects of the disclosure.

[0085] Figure 17 depicts a flow diagram of another example process of providing a suggested procedure using the data collection system of Figure 1 , according to aspects of the disclosure.Detailed Description

[0086] The technology is generally directed to Al model(s) that are trained to identify characteristics associated with lesions within a blood vessel and provide a suggested procedure associated with the lesion. The Al model may receive, as input, intravascular and / or extraluminal image data associated with the blood vessel and provide, as output, characteristics, e.g., characterizations, associated with the blood vessel. The characterizations may include, for example, form or tissue nature (e.g., microcalcification, punctate, fragment, sheet, nodular, or otherwise), location within the vessel e.g., intimal, superficial, medial, deep, etc.), size (e.g., length, depth, thickness, etc.), severity, and / or the like. The characterizations may be used as input to the same and / or a different model for providing a suggested procedure associated with the lesion. The suggested procedure may include, for example, orbital atherectomy, intravascular lithotripsy, rotational atherectomy, cutting and / or scoring balloons, and / or the like.

[0087] The world’s population will increasingly require treating for more and more complex coronary artery disease (CAD) and peripheral artery disease (PAD). Treatment of CAD and PAD require proficiency in using all types of modification techniques, an understanding of which procedures and / or tools are most appropriate in a given patient, and familiarity with intracoronary imaging use to guide the procedure. Increasingly, a combination of lesion modification techniques are being used in clinical practice, and again, can benefit from methods and systems described herein. In particular, the use of vessel data to characterize lesions and the subsequent use of the characterizations when determining a suggested procedure associated with the lesion provides for various workflows and options to facilitate a process of treating CAD and PAD, amongst other illnesses.

[0088] By using vessel data, e.g., intravascular and / or extraluminal image data, captured for the region of interest of the vessel, the system can dynamically determine information associated with the region of interest in the vessel, including characterizations of the lesion, that can be used to identify a suggested procedure for treating the lesion. The suggested procedure may be provided automatically upon receipt of the vessel data. Computational efficiency of the system is increased as processing and network overhead is decreased as fewer user inputs may be received by the system when identifying a suggested procedure. Further, when characterizing the vessel and / or the lesion with the vessel, the determined characterizations may be provided for output along with the suggested procedure. This allows for an efficient and almost fully automatic fusion of information between the image data and the information determined based on the vessel data, including the suggested procedure. For example, the information associated with the vessel can include information regarding the plaque burden, fractional flow reserve (“FFR”) measurements at one or more locations along the vessel, calcium angles, external elastic lamina (“EEL”) detections, calcium detections, proximal frames, distal frames, EEL-based metrics, stent / no stent decisions, lumen diameter, mean diameter, percent stenosis, types of lesions, and / or the like. Greater specificity and techniques forimaging and assessing vessels are described within U.S. Pat. No. 11,819,309, issued November 21, 2023, the entire contents of which being fully incorporated by reference herein.

[0089] Additional data, such as patient historical data, and / or indications of subsequent vessel preparation procedures performed on an imaged vessel can also be provided as input to the model. This additional data can be stored in a database for generating additional training data for the Al models, which in turn can be used for additional fine-tuning for improving model performance.

[0090] The display may include one or more representations including the information associated with the vessel. For example, the information associated with the vessel may be provided as numerical representations, classifications or labels, color-coded visualization, and / or co-axial rings, etc. The representations can include the image data, a two-dimensional representation generated based on the image data, three-dimensional representations, and / or the like.

[0091] The information associated with the vessel and the characteristics associated with the lesion may be used when determining a suggested procedure. Accordingly, the information associated with the vessel that is used to generate representations of the vessel that are provided for output can also be used for determining characterizations associated with the lesion as well as identifying suggested procedures. By using the vessel data for multiple purposes, e.g., determining information associated with the vessel, determining characteristics of the lesion, and / or determining suggested procedures associated with the lesions, the efficiency of the system is increased as multiple procedures no longer have to be performed to capture vessel data for each specific use.

[0092] According to some examples, the suggested procedure may include lesion modification procedures and / or treatment procedures. For example, the Al model(s) may be trained to identify suggested modification and / or treatment procedures. The modification treatment may be, for example, atherectomy, IVL, various forms of ablation, and / or balloons, etc. In yet another example, the suggested procedure may be a suggested implant device, e.g., a type of stent, a size of stent, and / or the like, and / or a suggested landing zone, e.g., suggested stent location, and / or the like. In some examples, a suggested procedure can include a suggestion to perform no procedure at all. As another example, a suggested procedure can include deferring treatment and / or a vessel preparation procedure until a later time.

[0093] The suggested procedure, device, and / or landing zone, etc. (collectively referred to herein as “suggested procedure”) is determined based on data information determined from the received vessel data. For example, the system receives vessel data, such as imaging data. The imaging data may be intravascular imaging data and / or extraluminal imaging data. In some examples, the suggested procedure may be further determined based on information stored in a database. The database is built over time from previous procedures including pre-treatment data, patient history data, the specific details of the procedure chosen and performed, and / or data related to the results of the treatment determined post-treatment.

[0094] The suggested procedure may be identified based on the characterizations associated with the lesion. For example, there are treatments that have been developed specifically for the expansion, breaking up, and / or removal of specific type(s), size(s), and / or location(s) of plaque. Such plaque includes, for example, lipids, fibrin, cholesterol, and / or calcification. The plaque can take many different forms aboutthe vessel wall and / or within different layers of the vessel wall. Accordingly, to identify the suggested procedure associated with the lesion within the vessel, the system determines characteristics associated with the lesion, e.g., the type, shape, and / or location (along a vessel wall and / or within a vessel layer), etc., and uses the determined lesion characteristics when identifying a suggested procedure.

[0095] According to some examples, the characteristics associated with the lesion, e.g., the characterization of the lesion, can be determined based on histopathology and / or microscopic examination of the tissue of a lesion. For example, the system can receive information from the histopathology and / or microscopic examination of the tissue. The information may include, for example, the form or tissue nature, such as microcalcification, punctate, fragment, sheet, nodular, and / or other forms of calcification. In some examples, the system may be configured to perform the histopathology and / or microscopic examination of the tissue. The histopathology or microscopic examination of the tissue of a lesion may be conducted on cadaveric and / or living coronary and / or peripheral arteries.

[0096] In some examples, the characteristics associated with the lesion may include the location of the lesions, e.g., intimal (closest to the vessel lumen), superficial, medial, and / or deep. In some examples, the Al model generates an output at least partially characterizing identified lesions and their locations. The output can form part of the input to the Al model for outputting a suggested procedure. In some examples, the Al model is split into separate models, at least one model for classifying lesions from input images, and at least one other model for suggesting a vessel preparation procedure based on the lesion classification output and potentially other vessel data. Before providing the lesion classification output, the system can receive user input to confirm or correct the classification, for example through a user interface. Output classifications can include intimal or medial calcification, including microcalcification, punctate calcification, fragment calcification, sheet calcification, and / or nodular calcification. The classification can include information as to the location of the lesion, e.g., intimal, superficial, medial, deep, and so on; size, e.g., length, depth, thickness, and so on; and / or severity. The output classification can be updated, if needed, before being provided as input to the model trained to suggest a vessel preparation procedure. In other examples, the Al model is trained end-to-end, e.g., from vessel data to output suggested vessel preparation. In those examples, the Al model may output a lesion classification. Other characteristics can be determined based upon any number of distinctions of one lesion as compared to another.

[0097] According to some examples, the characteristics associated with the lesion may be determined using an Al model, for example, a machine learning (ML) model, such as a convolutional neural network (CNN). The Al model may be trained to receive, as input, the vessel data and provide, as output, characteristics, e.g., characterizations, associated with the lesion. In some examples, the same Al model may be trained to determine information associated with the vessel in addition to characteristics associated with the lesion.

[0098] As an example, the model may be trained to identify a marker or markers as a detectable characteristic of a lesion. The lesion may be, for example, calcium within or along a vessel wall. The identified marker(s) may provide an indication as to the location of the lesion, whether or not the lesion has changed or been disrupted, and / or the like. The model can be trained on temporal data, e.g., vessel datacorresponding to the same vessel captured at different points in time. For example, the model can receive vessel data to generate a suggested procedure. The suggested procedure can then be performed, e.g., autonomously or manually. The suggested procedure can be paused, and new vessel data can be taken, for example from an OCT pullback taken of the vessel that has been partially prepared according to the procedure. In some examples, additional imaging, e.g., extraluminal imaging, can be taken during the procedure and form part of the new vessel data. The model can receive the new vessel data as additional input, and output a suggested procedure. In this example, the suggested procedure may be the same procedure as previously suggested, a different procedure, or a suggestion to stop the procedure because no further preparation is needed. For example, in the event of dissection, if a drug-coated balloon (DCB) is a primary recommendation, a secondary recommendation may be a drug-eluting stent (DES) or a drugeluting resorbable scaffold (DRS). The two instances of vessel data, e.g., before and during preparation, respectively, can be stored in a database as described herein and provided as a training example.

[0099] To that end, the temporal nature of this training data can cause the model to identify latent features or connections between features that vary with time. For example, the training data across different points of time of a procedure can reflect changes in the present stenosis, fractures in calcium, and / or other changes due to the suggested vessel procedure.

[0100] According to some examples, vessel data may be used when determining, a suggested procedure. The suggested procedure may be a procedure that should be performed by the user, e.g., the physician, prior to another procedure, e.g., stent implantation. In other examples, the suggested procedure may be performed automatically, e.g., by a surgical robot or other device configured to perform the procedure. The vessel data can include image data captured by extraluminal and / or intravascular imaging devices. Extraluminal imaging devices include, for example, computer tomography (CT) scanner, magnetic resonance imaging (MRI) scanner, ultrasound probe, fluoroscopy, and / or the like. Intravascular imaging devices may be, for example, an optical coherence tomography (“OCT”) probe, an intravascular ultrasound (“IVUS”) catheter, micro-OCT probe, near infrared spectroscopy (NIRS) sensor, optical frequency domain imaging (“OFDI”) probe, and / or any other device that can be used to image the blood vessel. In some examples, in addition to the vessel data, the suggested type of treatment may be determined based on other data obtained from any specific lesion by other means than imaging. As one example, a pullback of an OCT probe may capture vessel data, e.g., intravascular images. The vessel data captured by the OCT probe provides sharp delineations of calcium areas or zones as associated with vessel wall structure.

[0101] The system can use these intravascular images, e.g., vessel data, to determine characteristics of the lesion, e.g., calcification, formation, and / or inter-layer vessel data. The system can then use the characteristics of the lesion when determining a suggested type of vessel preparation procedure to use for a desired treatment of such a lesion within a vessel. For example, the characteristics associated with the lesion may be provided as input into an Al model trained to identify a suggested procedure, for example, for expansion, breaking up, and / or removal of the lesion within the vessel. As described herein, suggested procedures can include a suggestion to perform no vessel preparation, and / or to reassess the need to do a vessel preparation procedure later. For example, another round of imaging and vessel preparationsuggestion can be performed at a later period. The Al model trained to identify the suggested procedure may be the same or another model as compared to the Al model trained to identify the characteristics associated with the vessel and / or lesion. For example, the Al model may include a sub-model or module for generating OCT image data, which is fed as input to another sub-model or module for generating a suggested recommendation. In other examples, the Al model processes input in the form of OCT images end-to-end, to generate an output suggested recommendation. The Al model described herein can represent multiple models trained to perform these and other tasks.

[0102] As discussed above and herein, such images, whether alone or taken with other data of a specific lesion, can be utilized in determining a suggested procedure, e.g., lesion modification therapy or technique, associated with the lesion prior to another treatment, e.g., stent deployment, balloon angioplasty, scaffold placement, and / or any other procedure for performing removal of vessel blockages. By efficiently and accurately characterizing the vessel and / or lesion and using the characteristics of the vessel and / or lesion to identify a suggested procedure, further modifications and / or procedures associated with the lesion may be mitigated. Further, subsequent procedures, e.g., stent deployment, can be more successful due to the intervening suggested procedure associated with the lesion. For example, modifying a coronary calcium lesion prior to stent placement can ensure better stent performance and prevent stent failure. Stent underexpansion can occur if the calcium lesion is not first modified, such as by a balloon expansion or other modification therapy. Optimizing stent expansion performance and results is therefore of paramount importance, and calcium modification prior to percutaneous coronary intervention (PCI) is an important step in this process to ensure adequate stent expansion.

[0103] The model may be trained based on a database generated based on previously captured vessel data, patient data, procedures, and / or outcomes of procedures subsequently performed on the vessels in the captured vessel data. The data may be collected from cadaveric and / or living subjects. The model may be updated with procedures performed on the vessel based on the currently captured vessel data. In that regard, the model may include a feedback loop such that the model is continuously updated based on the most recent procedures performed on the vessel in the more recently captured vessel data, as well as user input indicating whether the procedure was correct, or whether other procedures are needed. As described herein, patient data can be input into the model, in addition to image data. The patient data can be stored in a database record, along with the intravascular image data corresponding to the respective patient. Feedback in the form of patient data can include measurements, such as heart rate, blood pressure, and / or blood flow rate, etc., as well as details provided by a physician performing a vessel preparation procedure. In some examples the blood flow rate may be determined based on intravascular image data. In addition, the patient data can include descriptions of if and / or how a suggested procedure was deviated from, when the procedure was performed. This and / or other data can be used as training examples for further training or fine-tuning the Al model.

[0104] Figure 1 illustrates a data collection system 100 for use in collecting intravascular data, calculating further data from the collected data, and for providing a graphical user interface enabling a user to view and interact with the collected and calculated intravascular data, as further described below. The systemmay include a data collection probe 104 that can be used to image a blood vessel 102. In some examples, the probe 104 may be an intravascular device, such as an OCT probe, an IVUS catheter, micro-OCT probe, near infrared spectroscopy (NIRS) sensor, optical frequency domain imaging (OFDI), and / or any other device that can be used to image the blood vessel 102. In some examples, the probe 104 may be a pressure wire or a flow meter, etc. The probe 104 may include a device tip, one or more radiopaque markers, an optical fiber, a torque wire, and / or the like. Additionally, the device tip may include one or more data collecting subsystems such as an optical beam director, an acoustic beam director, a pressure detector sensor, other transducers or detectors, and combinations of the foregoing.

[0105] A guide wire, not shown, may be used to introduce the probe 104 into the blood vessel 102. The probe 104 may be introduced and pulled back along a length of a blood vessel while collecting data. As the probe 104 is pulled back, or refracted, a plurality of scans or OCT and / or IVUS data sets may be collected. The data sets, or frames of image data, may be used to identify features, such as vessel dimensions and / or pressure and / or flow characteristics.

[0106] The probe 104 may be connected to a subsystem 108 via an optical fiber 106. The subsystem 108 may include a light source, such as a laser, an interferometer having a sample arm and a reference arm, various optical paths, a clock generator, photodiodes, and / or other OCT and / or IVUS components.

[0107] The probe 104 may be connected to an optical receiver 110. According to some examples, the optical receiver 110 may be a balanced photodiode based system. The optical receiver 110 may be configured to receive light collected by the probe 104.

[0108] The data collection system 100 may further include, or be configured to receive data from, a non- invasive imaging system 120. The non-invasive imaging system 120 may be, for example, an imaging system based on angiography, fluoroscopy, x-ray, nuclear magnetic resonance, and / or computer aided tomography, etc. Non-invasive imaging system 120 may be configured to non-invasively image the blood vessel 102. According to some examples, the non-invasive imaging system 120 may obtain one or more images before, during, and / or after a pullback of the data collection probe 104. Non-invasive imaging system 120 may be used to image a patient such that decisions can be made and various possible treatment options such as stent placement can be carried out. These and other imaging systems can be used to image a patient externally or internally to obtain raw data, which can include various types of image data.

[0109] The non-invasive imaging system 120 may be in communication with subsystem 108. According to some examples, the non-invasive imaging system 120 may be wirelessly coupled to subsystem 108 via network. For example, the non-invasive imaging system 120 may be wirelessly coupled to subsystem 108 via a communications interface, such as Wi-Fi and / or Bluetooth. In some examples, the non-invasive imaging system 120 may be in communication with subsystem 108 via a wire, such as an optical fiber. In yet another example, external imaging device 120 may be indirectly communicatively coupled to subsystem 108 or computing device 112. For example, the non-invasive imaging device 120 may be coupled to a separate computing device (not shown) that is in communication with computing device 112. As another example, data from the imaging system 120 may be transferred to the computing device 112 using a computer-readable storage medium, from a storage device via a network, and / or the like.

[0110] The subsystem 108 comprises a computing device 112. The computing device may include one or more processors 113, memory 114, instructions 115, data 116, and one or more modules 117.

[0111] The one or more processors 113 may be any conventional processors, such as commercially available microprocessors. Alternatively, the one or more processors may be a dedicated device such as an application specific integrated circuit (ASIC), a graphics processing unit (GPU), a field programmable gate array (FPGA), and / or other hardware-based processors. Although Figure 1 functionally illustrates the processor, memory, and other elements of device 112 as being within the same block, it will be understood by those of ordinary skill in the art that the processor, computing device, or memory may actually include multiple processors, computing devices, or memories that may or may not be stored within the same physical housing. Similarly, the memory may be a hard drive or other storage media located in a housing different from that of device 112. Accordingly, references to a processor or computing device will be understood to include references to a collection of processors or computing devices or memories that may or may not operate in parallel.

[0112] Memory 114 may store information that is accessible by the processors, including instructions 115 that may be executed by the processors 113, and data 116. The memory 114 may be a type of memory operative to store information accessible by the processors 113, including a non-transitory computer- readable medium, or other medium that stores data that may be read with the aid of an electronic device, such as a hard-drive, memory card, read-only memory (“ROM”), random access memory (“RAM”), and / or optical disks, as well as other write-capable and read-only memories. The subject matter disclosed herein may include different combinations of the foregoing, whereby different portions of the instructions 115 and data 116 are stored on different types of media.

[0113] Memory 114 may be retrieved, stored and / or modified by processors 113 in accordance with the instructions 115. For instance, although the present disclosure is not limited by a particular data structure, the data 116 may be stored in computer registers, in a relational database as a table having a plurality of different fields and records, XML documents, and / or flat files. The data 116 may also be formatted in a computer-readable format such as, but not limited to, binary values, ASCII or Unicode. By further way of example only, the data 116 may be stored as bitmaps comprised of pixels that are stored in compressed or uncompressed, or various image formats (e.g., JPEG), vector-based formats (e.g., SVG) or computer instructions for drawing graphics. Moreover, the data 116 may comprise information sufficient to identify the relevant information, such as numbers, descriptive text, proprietary codes, pointers, references to data stored in other memories (including other network locations) and / or information that is used by a function to calculate the relevant data.

[0114] The instructions 115 can be any set of instructions to be executed directly, such as machine code, or indirectly, such as scripts, by the processor 113. In that regard, the terms “instructions,” “application,” “steps,” and “programs” can be used interchangeably herein. The instructions can be stored in object code format for direct processing by the processor, or in any other computing device language including scripts or collections of independent source code modules that are interpreted on demand or compiled in advance. Functions, methods and routines of the instructions are explained in more detail below.

[0115] According to some examples, the computing device 112 may receive, either by a wired connection or via a wireless connection, data from the probe 104. The data may include, for example, intravascular data including intravascular imaging data, pressure data, temperature data, flow data, and / or the like. In some examples, where the probe 104 is an intravascular data collection device, the data received from the device may be used to determine characterizations associated with the lesion, plaque burden, fractional flow reserve (“FFR”) measurements at one or more locations along the vessel, calcium angles, external elastic lamina (“EEL”) detections, calcium detections, proximal frames, distal frames, EEL-based metrics, stent / no stent decisions, scores, recommendations for debulking and other procedures, evidence based recommendations informed by automatic detection of regions / features of interest, and / or stent planning, etc.

[0116] The modules 117 may include various modules configured to implement various functions described in more detail later in this document. Such modules may include a suggested procedure module and a display module. In some examples further types of modules may be included, such as modules for computing other vessel characteristics, and / or stent detection modules, etc. According to some examples, the modules may include an image data processing pipeline or component modules thereof. The image processing pipeline may be used to transform collected vessel data, e.g., image data, into two-dimensional (“2D”) and / or three-dimensional (“3D”) views and / or representations of blood vessels, stents, and / or detected regions.

[0117] According to some examples, the modules may additionally or alternatively include other types of image processing modules, such as a video processing software module, a preprocessing software module, an image file size reduction software module, a catheter removal software module, a shadow removal software module, a vessel enhancement software module, a blob enhancement software module, a Laplacian of Gaussian filter or transform software module, a guide wire detection software module, an anatomic feature detection software module, stationary marker detection software module, a background subtraction module, a Frangi vesselness software module, an image intensity sampling module, a moving marker software detection module, iterative centerline testing software module, a morphological close operation software module, a feature tracking software module, a catheter detection software module, a bottom hat filter software module, a path detection software module, a Dijkstra software module, a Viterbi software module, fast marching method based software modules, a vessel centerline generation software module, a vessel centerline tracking module software module, a Hessian software module, an intensity sampling software module, a superposition of image intensity software module and other suitable software modules as described herein. For example, the modules may include a Frangi vesselness software module, which may be highly effective for enhancing tubular structures, such as blood vessels, in intravascular imaging. This module may improve contrast and delineation of vessel boundaries, which is critical for accurate lumen segmentation and lesion characterization. This module may be particularly useful in preprocessing OCT and IVUS data to support downstream tasks, such as calcium detection and stent planning.

[0118] According to some examples, the modules may include software such as preprocessing software, transforms, matrices, and / or other software-based components that are used to process image data or respond to patient triggers to facilitate co-registration of different types of image data by other softwarebased components, and functions suitable for implementing various features of the disclosure. The modules can include lumen detection using a scan line based or image based approach, stent detection using a scan line based or image based approach, indicator generation, apposition bar generation for stent planning, guide wire shadow indicator to prevent confusion with dissention, side branches and missing data, and / or others. For example, the module may include lumen detection using a scan line based approach, which may provide robust and computationally efficient detection of a vessel lumen boundary. This module may be well-suited for real-time applications and may integrate seamlessly with OCT pullback data. Accurate lumen detection is foundational for calculating metrics, such as minimum lumen area or percent stenosis, and for identifying lesion locations relative to anatomical landmarks.

[0119] To facilitate co-registration of different types of image data, the modules may co-register two or more representations of image data from two or more pullbacks. As such, two or more representations of pullbacks along an overlapping portion of the vessel may be integrated together. The pullbacks may utilize one or more types of intravascular imaging devices, such as an optical coherence tomography (OCT) probe, an intravascular ultrasound (IVUS) catheter, micro-OCT probe, near infrared spectroscopy (NIRS) sensor, optical frequency domain imaging (OFDI) device, or any other device that can be used to image a blood vessel.

[0120] Registering the representations may include, for example, identifying a location in one of the representations and identifying a comparable, or equivalent, location in the other representation(s). As an example, if the location is identified in a first representation, then a comparable location would be identified in a second representation. Continuing with this example, if the location in the first representation is identified as position X along the vessel, the comparable location in the second representation would also be position X, or substantially position X, along the vessel. The location may be identified based on an input, or user instruction, received by the system. For example, the system is configured to receive the input, or user instruction, via the GUI or other component, corresponding to the selection of the location along a first or second representation of the vessel. The location may be, for example, an easily identifiable location along the vessel, such as a side branch, a bifurcation, a bend, a proximal or distal end of a previously implanted device, etc. The system is further configured to receive another input, or user instruction, corresponding to the selection of the comparable location along the other representation of the vessel. Based on the received inputs, the system co-registers the representations based on the understanding that the selected locations are at the same, or substantially the same, location along the vessel in each pullback.

[0121] In some examples, the locations may be automatically identified. For example, the system may automatically identify anatomical features or previously implanted medical devices, such as stents. The system may compare the anatomical features in the representations to determine which anatomical features are the same or substantially the same. In some examples, the system may determine that the anatomicalfeatures are the same or substantially the same based on a pattern, or sequence, of the anatomical features. For example, if the system identifies a side branch then a bend, then a region of vessel having increased calcium burden in the first representation, the system may identify a corresponding region in the second representation that includes the same sequence of a side branch, bend, then region having increased calcium burden. Based on the comparison, the system may register a given location in the first representation with the comparable location in the second representation. The system may similarly register the remaining portions of the representations.

[0122] In some examples, the modules may be configured to process the vessel data obtained by the probe 104 and / or imaging system 120 using Al models, ML algorithms, and / or the like.

[0123] The subsystem 108 may include a display 118 for outputting content to a user. As shown, the display 118 is separate from computing device 112. However, according to some examples, display 118 may be part of the computing device 112. The display 118 may output image data relating to one or more features detected in the blood vessel. For example, the output may include, without limitation, cross- sectional scan data, longitudinal scans, diameter graphs, and / or image masks, etc. The output may further include lesions and visual indicators of vessel characteristics or lesion characteristics, such as computed pressure values, vessel size and shape, and / or the like. The output may further include visual indicia for candidate stent placement, such as an overlay highlighting selected vessel regions for potential stent placement. In some examples, the output includes an indication of the suggested procedure. The display 118 may identify features with text, arrows, color coding, highlighting, contour lines, and / or other suitable human or machine readable indicia.

[0124] According to some examples the display 118 may be used to present a graphic user interface (“GUI”) to a user so that a user may interact with the computing device 112 and thereby cause particular content to be output on the display 118, typically using forms of input such as a mouse, keyboard, trackpad, microphone, gesture sensors, and / or any other type of user input device. One or more steps may be performed automatically or without user input to navigate images, input information, select and / or interact with an input, etc. The display 118 and input device, along with computing device 112, may allow for transition between different stages in a workflow, different viewing modes, etc. For example, the user may select a segment of vessel for analysis, enter data or commands in response to prompts when transitioning through different phases of a workflow, etc.

[0125] In some examples, information input by the user using the GUI may comprise annotations. For example, the system may be configured to receive annotations to one or more representations of the vessel displayed using the GUI. The annotations may be, in some examples, an indication of plaque burden, fractional flow reserve (“FFR”) measurements at one or more locations along the vessel, calcium angles, EEL detections, calcium detections, proximal frames, distal frames, EEL-based metrics, stent / no stent decisions, scores, recommendations for debulking and other procedures, evidence-based recommendations informed by automatic detection of regions / features of interest, and / or stent planning, etc. In some examples, the annotations may be related and / or associated with the suggested treatment. For example, the annotations may be a treatment device landing zone, balloon device zone, vessel preparation device zone,and / or lesion related zone. For example, the system may receive a user input corresponding to a proximal and distal location along the vessel corresponding to a proximal and distal location of a treatment device landing zone, balloon device zone, vessel preparation device zone, and / or lesion related zone. According to some examples, the system may receive an input corresponding to a proximal and distal location along the vessel selecting frames for the detection module. The lumen area of the selected frames may be corrected by the detection module by applying linear interpolation.

[0126] According to some examples, annotations may also or instead be automatically determined by the data collection system 100. For example, the system may, based on vessel data, determine one or more of plaque burden, FFR measurements at one or more locations along the vessel, calcium angles, EEL detections, calcium detections, proximal frames, distal frames, EEL-based metrics, stent / no stent decisions, scores, recommendations for debulking and other procedures, evidence-based recommendations informed by automatic detection of regions / features of interest, stent planning, a treatment device landing zone, balloon device landing zone, vessel preparation device zone, and / or lesion related zone, etc. The system may automatically provide the plaque burden, FFR measurements at one or more locations along the vessel, calcium angles, EEL detections, calcium detections, proximal frames, distal frames, EEL-based metrics, stent / no stent decisions, scores, recommendations for debulking and other procedures, evidence-based recommendations informed by automatic detection of regions / features of interest, stent planning, a treatment device landing zone, balloon device landing zone, vessel preparation device zone, and / or lesion related zone, etc. for output as one or more annotations on at least one of the vessel representations.

[0127] The display 118 alone or in combination with computing device 112 may allow for toggling between one or more viewing modes in response to user inputs. For example, a user may be able to toggle between different intravascular data, images, etc. recorded during each of a plurality of pullbacks of the probe 104 within the blood vessel 102. In some examples, the user may be able to toggle between different representations, such as a longitudinal representation, a cross-sectional representation, a three-dimensional representation, intravascular images, color images, black and white images, live images, and / or the like.

[0128] In some examples, the display 118, alone or in combination with computing device 112, may present one or more menus to the user (such as a physician), and the user may provide input in response by selecting an item from the one or more menus. For example, the menu may allow the user to show or hide various features. As another example, there may be a menu for selecting blood vessel features to display. Additionally, as another example, there may be a menu for selecting available treatments within a treatment facility, such as a hospital. For example, the menu may allow a user to select specifications of available treatments (such as a diameter and / or a length of available balloons) and / or brands of available treatments. According to some examples, the display may output a menu including one or more inputs for analyzing and / or processing the information associated with the vessel data. For example, the inputs may include adding and / or removing a side branch, recalculating virtual flow reserve (VFR), measuring selected regions of interest, correcting false positives (e.g., lumen bridge module), and / or the like.

[0129] The content output on the display 118 may include one or more representations of the vessel 102. For example, the representations may include image data, longitudinal representations, three-dimensionalrepresentations, live representations, and / or the like. Image data can include image frames captured, for example, during a pullback of an intravascular imaging probe, such as OCT, NIRS, IVUS, and micro-OCT. Image data may include extraluminal images frames, such as angiography, computed tomography (CT), magnetic resonance imaging (MRI), and / or fluoroscopy, etc. Image data can include pixel data of each pixel representing an image. Image data can also include metadata associated with an image, such as the time it was taken, the equipment used to take the image, if any filters or settings were used to take the image, and so on. The longitudinal representation may include, for example, a representation of the blood vessel based on the lumen diameter that is symmetrical about the longest axis of the representation. In some examples, the representations may include graphical representations, such as a graphical representation of VFR, pressure values, flow values, and / or the like.

[0130] In some examples, the one or more visual representations of the images may include an indication of a lesion location, lesion severity, lesion length, and / or the like. Additionally or alternatively, the indication of the lesion may be color coded, where each color represents the severity, length, and / or other measurement related to the lesion.

[0131] In some examples, the content output on the display 118 may include candidate treatment zones. Such a candidate treatment zone may be, for example, a candidate stent landing zone. For example, the output may include an indication corresponding to a candidate proximal landing zone for a stent and a candidate distal landing zone for a stent. The candidate treatment zone may be determined based on the determined plaque burden, lesion locations, lesion length, and / or the like. The indications may be provided on any of the vessel representations, e.g., the three-dimensional representation, the longitudinal representation, the graphical representation, and / or image data such as the external images, etc.

[0132] In some examples, the candidate treatment zone may be associated with the suggested procedure. For example, the candidate treatment zone may include an indication corresponding to a location for a lesion modification treatment. The indication can be any graphical element, e.g., arrows, boxes, highlighting, colors, textual labels, and / or enlargement of the location, and so on.

[0133] According to some examples, the display 118 and / or computing device 112 may be configured to receive one or more inputs from a user corresponding to a selection made on one or more representations such as representations of the blood vessel. For example, an input may be received from the user corresponding to a selection of an image frame on a longitudinal representation of the blood vessel. In response, other representations being output may be updated to display a corresponding indication or image frame. For example, a displayed extraluminal image may be updated to have an indication along the vessel corresponding to the location of the image frame selected in the longitudinal representation, a circumferential indication may be provided on a three-dimensional representation corresponding to the location of the image frame selected in the longitudinal representation, and / or the cross-sectional image frame may be updated to correspond to the image frame selected in the longitudinal representation, etc. In some examples, the vessel data associated with the selected location may be updated and provided for display.

[0134] The data collection system 100 of Figure 1 may be used to collect vessel data, e.g., image data, that can be used to identify suggested procedures associated with a lesion within the vessel as part of the treatment of calcified coronary artery disease (CAD) or peripheral artery disease (PAD) . CAD affects coronary arteries, which supply blood to the heart. PAD affects peripheral arteries, which supply blood from the heart to extremities, including arms, hands, legs, and feet. The intravascular images may be IVUS and / or OCT images. The IVUS and / or OCT images provide images for assessment of coronary and peripheral lesions. The system can analyze the images to determine information associated with the vessel and / or lesion, such as a detailed morphological assessment of the lesion and assessment of the result of lesion modification techniques. The system can, in some examples, analyze and / or use the images when determining a suggested plan and / or guide for PCI with selection of proximal and distal landing zones, stent diameter, and / or length. According to some examples, the data collection system 100 of Figure 1 can use, in part, the vessel data, e.g., image data, the vessel information determined from the vessel data, and / or the characteristics of the lesions determined from the vessel data to identify a suggested procedure for modifying the lesion.

[0135] In practical terms, lesions such as coronary calcium can be subdivided into morphologic subtypes based on intra-coronary imaging findings. Eccentric calcification extends across two quadrants or less and thereby has an arc of < 180°, concentric calcification has an arc of > 180°, and nodular calcification presents as an eruptive calcium protrusion into the lumen. For example, depth and length of calcium are also important predictors of PCI result. A calcium arc > 180°, depth > 0.5 mm, and length of > 5 mm as determined by OCT had an increased risk of stent under-expansion. Although both OCT and IVUS can assess calcium length, OCT provides a better assessment of calcium depth due to the ability of light to penetrate calcium. Intravascular ultrasound, being unable to penetrate calcium, creates an acoustic shadow, thereby hindering depth assessment. However, surrogate markers can be used to determine the calcium thickness by IVUS with the presence of posterior reverberations being correlated with thinner calcium sheets (< 0.5 mm), while significant shadowing suggests thicker calcification (> 1 mm). Recently, an IVUS-specific scoring system was found to be useful in predicting stent under-expansion using four criteria: (1) a calcium arc > 270° for a length of 5 mm, (2) the presence of 360° calcium, (3) the presence of a calcified nodule, and (4) an adjacent vessel diameter < 3.5 mm. A score of 2 suggests that calcium modification should be undertaken.

[0136] Lesion modification techniques, including balloon-based technologies, atherectomy, ablative techniques, and more recently a lithotripsy-based technique, are at the disposal of the interventional cardiologists. The data collection system 100 of Figure 1 can be used to identify a suggested procedure, e.g., modification technique, in view of the characteristics associated with the lesion.

[0137] By training an Al model as described herein, the imaging system avoids hard-coded rules or determinations. For example, some determinative algorithms propose eccentric calcium with different degrees of arcs be treated with different forms of cutting or scoring balloons. Based on the labeled data used to train the model, the Al model can suggest procedures that are consistent with standard medical procedures at the time of output, enforced by the feedback loop that allows for the labels to be updated bytechnicians or physicians performing a vessel preparation procedure. In addition, the Al model can be updated to suggest procedures that are more recent or newer than when the Al model was initially created, reinforced by feedback of the operators. For example, as new medical procedures are developed or changed, those new medical procedures can be input into the Al model to output those new medical procedures as a suggested procedure.

[0138] The lesion modification may be determined based on the characteristics of the lesion. The lesion modification techniques include, for example, balloon-based therapies in eccentric calcification and ablative- or lithotripsy-based therapies in concentric and nodular calcification. Excimer laser coronary angioplasty (ELCA) has had variable results for lesion modification. Increasingly, lesion modification techniques are seen as being complementary, and combinations of techniques are often advocated when treating coronary calcium. Post-calcium modification imaging can be used to assess results and determine if further modification is required prior to stenting.

[0139] With reference to Figure 2, an IVL system 10 can comprise a console or power source 12 (in the form of an electrical generator, but alternatively in the form of a laser system), a handle 14 with therapy delivery control 15 and a catheter 20 with two lithotripsy emitters 22 (shown in the form of a pair of arcing electrodes, but alternatively they could comprise optical or laser emitters), and a fluid filled balloon 24. The IVL system 10 may, in some examples, be in communication with the data collection system 100 of Figure 1. In some examples, the data collection system 100 may be configured to track, in real time, the position of the balloon 24 during the procedure. As discussed above, the Al model can be used at different stages of vessel preparation to determine whether to stop the procedure or change the procedure, for example various types of orbital atherectomy for preparing vessels with concentric calcium, eccentric calcium, nodular calcium, heavily stenosed lesions (e.g., greater than 99% cross and less than 2% predilation), long and diffuse lesions (e.g., up to 60 mm in length), multi-vessel disease, and / or small diameter vessels (e.g., 2.5 mm). For example, the Al model may propose one or more modalities of vessel preparation and, upon completion and re-assessment, propose one or more additional modalities of vessel preparation that are the same or different from the previously proposed one or more modalities of vessel preparation. Once vessel preparation is complete, the Al model may propose one or more additional therapies.

[0140] In some examples, the system may dynamically adapt procedural recommendations based on realtime or sequential imaging data, enhancing safety and procedural success. One example of a change in procedure may occur as a mid-procedure adjustment. For example, the Al model may propose an initial OA for a concentric calcified lesion with >180° arc and moderate thickness. After the OA passes, an additional OCT pullback may show one or more indications that the initial OA was insufficient for full lesion modification. In some examples, those indications may include a partial calcium fracture, residual nodular calcification, and persistent lumen narrowing at >70%. Once the system reprocesses the updated vessel data, the Al model may suggest additional OA passes with a larger crown size and / or IVL to address deeper calcium not fractured by OA. This updated recommendation from the Al model may be a result of an identification that the initial OA was insufficient for full lesion modification. Thus, the Al model mayrecommend a complementary modality to optimize vessel compliance before stenting. Another example of change in procedure may occur as a post-procedure adjustment. For example, after completion of IVL and stent deployment, a post-procedural OCT pullback may show one or more indications of inadequate expansion. For example, those indications may include stent under-expansion (<80% expansion index) and residual calcium thickness >0.5 mm beneath stent struts. Once the system reprocesses the updated vessel data, the Al model may suggest mechanical optimization, such as high-pressure non-compliant balloon post-dilation and / or use of IVL post-stenting. The Al model may be trained on post-stent imaging and outcomes, so this updated recommendation from the Al model may be a result of identification that inadequate expansion is predictive of restenosis. Thus, the Al model may recommend further mechanical optimization. As another example, if a dissection is detected mid-procedure, the Al model may shift the recommendation from a drug-coated balloon (DCB) to a drug-eluting stent (DES) or drug-eluting resorbable scaffold (DRS). As yet another example, if thrombus is detected during imaging, the Al model may recommend embolic protection or defer lesion preparation until thrombus burden is reduced.

[0141] Optional marker bands B may be provided. The catheter 20 preferably includes a central tube defining a guide wire lumen through which a guide wire G passes for delivering the balloon 24 at the desired location along the guide wire G. A sheath can surround the central tube for defining a delivery lumen through which saline can be controllably delivered for balloon 24 inflation. The lumen can thus provide a concentric space around the central tube within which electrode wires (not shown) can be run from the control 15 to the emitters 22 among other components in accordance with the present invention. The sheath can be connected at a proximal end to a hub 17 that can include any number of ports allowing electrode wires to pass into the lumen along with saline for inflation, the guide wire G, and any number of other components as desired.

[0142] The balloon 24 may be placed in a deflated position so as to more readily pass through a patient’s vasculature to arrive at the site of calcification. In use, the balloon 24 will be inflated to a common pressure for angioplasty procedures (e.g. 4 atm) and the therapy actuated via the delivery control 15.

[0143] Figure 2 shows the balloon 24 inflated to a therapy delivery state where the lithotripsy emitters 22 may be “fired” to disrupt the vessel calcification. Optional indicator bands B may be provided to afford visualization and proper positioning by use of known imaging techniques. The balloon 24 is inflated to a typical angioplasty pressure (e.g. 4 atm) and therapy is delivered. The balloon 24 may naturally expand during or just after the therapy is delivered to clear the vessel for passage of blood. While two pairs of electrode pair emitters 22 are shown in Fig. 2, in alternative embodiments the number of electrode pair emitters can comprise only 1 pair or 2, 3, 4, 5, 6 or even more emitter pairs to address longer lesions, such as those encountered in the peripheral vasculature.

[0144] The control 15 is used to produce one or a series of voltage pulses in accordance with a vessel wall or lesion treatment scheme. A high voltage pulse is provided to one of the emitters 22 comprising a pair of spaced electrodes and, in accordance with the illustrated embodiment, then in series to a second emitter 22 also comprising a pair of spaced electrodes. The high voltage pulse causes a spark across the first electrode pair then or also across the second electrode pair sequentially within the balloon 24. The somewhatconductive saline solution within the balloon 24 permits the high voltage spark across each electrode pair, thus creating an energy wave that propagates within the balloon toward the vessel calcification.

[0145] According to some examples, the identified procedure may be an RA procedure. For example, RA may be more useful for uncrossable and undilatable lesions or concentric calcification and can be combined with other modification techniques.

[0146] In another example, the identified suggested procedure may be an OA procedure. The vessel data, can be used to identify the suggested procedure as well as provide indicators to guide the OA procedure technique. In some examples, the vessel data may include information associated with the vessel, determined based on image data and / or patient history data, characterizations of the lesions within the vessel of interest, determined based on image data and / or patient history data, and / or the like. For example, the size and placement of the crown relative to a marker band or other feature of an OA catheter can be determined prior to the procedure. In such an example, the model may identify the size and placement of the crown relative to the OA catheter as part of the suggested procedure.

[0147] Figure 3 illustrates an example method for determining information associated with the vessel that was imaged, characteristics associated with the lesion within the imaged vessel, and identification of suggested procedures for modifying the lesion, using the imaging system of Figure 1. The following operations do not have to be performed in the precise order described below. Rather, various operations can be handled in a different order or simultaneously, and operations may be added or omitted.

[0148] The data collection system 100 captures vessel data via probe 104. The vessel data may be captured prior to treatment, e.g. a lesion modification procedure, vessel preparation, and / or the like. The vessel data may be, for example, intravascular image data. In the example of Figure 3, the vessel data is OCT image data. The data collection system 100 determines, based on the vessel data, information associated with the vessel and / or characteristics of the lesion. The information associated with the vessel can include, for example, plaque burden, fractional flow reserve (“FFR”) measurements at one or more locations along the vessel, calcium angles, external elastic lamina (“EEL”) detections, calcium detections, proximal frames, distal frames, EEL-based metrics, stent / no stent decisions, lumen diameter, mean diameter, percent stenosis, types of lesions, and / or the like. The characteristics of the lesion may include, for example, the presence of lesion, location of lesion, lesion type, and / or lesion arc / coverage, etc.

[0149] In some examples, the vessel data may be provided as input into an Al model trained to provide a prediction of the information associated with the vessel characteristics and / or characteristics of the lesion. The information associated with the vessel and / or the characteristics of the lesion may be stored in association with the vessel data. The vessel data may include, for example, patient data corresponding to the patient whose vessels are being imaged, image data, and / or the like. For example, the information associated with the vessel, the characteristics of the lesion, and / or the images may be stored in memory of the data collection system 100, a remote storage device, a repository, and / or the like.

[0150] Based on, at least in part, the characteristics of the lesion, a modification procedure may be determined. As shown in Figure 3, when the lesion is calcium, the modification procedure may be determined based on the degrees of the calcium arc, the thickness of the calcium, and the length of thecalcium in the region of interest of the vessel. In some examples, selection of a vessel modification procedure may be highly dependent on specific lesion characteristics, such as the degrees of the calcium arc, the thickness of the calcium, and / or the length of the calcium in the region of interest of the vessel. These parameters may directly influence mechanical resistance of a lesion to dilation and a likelihood of achieving optimal stent expansion, as a prediction of procedural success and long-term outcome. For instance, the presence of a calcium arc >270°, thickness >0.5 mm, and length >5 mm has been associated with a significantly increased risk of stent under-expansion. Therefore, tailoring the modification strategy to these parameters may optimize stent deployment and reduce restenosis risk. For example, orbital atherectomy is effective with severe calcification, lesion accessibility, and / or bidirectional modification. Specifically, OA can modify severe calcified lesions where other devices might struggle to cross. Moreover, OA is useful when the lesion is accessible but requires precise modification to facilitate stent placement. Further, OA can modify calcium in ether direction along a vessel, which is particularly advantageous with complex lesions. Thus, orbital atherectomy may be the determined modification procedure with higher degrees of the calcium arc, larger thickness of the calcium, and / or longer length of the calcium in the region of interest of the vessel. As another example, IVL is effective with balloon crossing and less severe calcification. Specifically, IVL is useful when balloons can easily cross the lesion but adequate calcium management is uncertain and is effective in less severely calcified lesions where balloon crossing is feasible. Further, IVL is simple to use and does not require exchanging guidewires, making it a convenient option in certain cases. Thus, IVL may be the determined modification procedure when lower degrees of the calcium arc, smaller thickness of the calcium, and / or shorter length of the calcium in the region of interest of the vessel are detected.

[0151] As an example of a modification procedure regarding calcium arc, a lesion with a calcium arc less than 180° may respond well to scoring and / or cutting balloons. These devices may create controlled fractures in the calcium, improving vessel compliance without requiring high-pressure dilation. As such, the modification procedure for a calcium arc of less than 180° may be determined to be scoring and / or cutting balloons. On the other hand, a lesion with a calcium arc greater than 180° may be more rigid and circumferentially constrain the vessel. In such cases, IVL and / or OA may be preferred to modify the lesion and facilitate stent expansion. As such, the modification procedure for a calcium arc of more than 180° may be determined to be OA and / or IVL.

[0152] As an example of a modification procedure regarding calcium thickness, a calcium thickness of less than 0.5 mm may be fractured with IVL or high-pressure balloon angioplasty. Posterior reverberation artifacts on IVUS often indicate thin calcium, which is more amenable to balloon-based therapies. As such, the modification procedure for a calcium thickness of less than 0.5 mm may be determined to be IVL and / or high-pressure balloon angioplasty. On the other hand, a calcium thickness of more than 0.5 mm may be more resistant to fracture. In such cases, OA is often more effective due to its ability to ablate and sand down dense calcium, especially when IVL fails to produce adequate fractures. As such, the modification procedure for a calcium thickness of more than 0.5 mm may be determined to be OA.

[0153] As an example of a modification procedure regarding calcium length, a calcium length of less than 5 mm may be treated with focal therapies, such as scoring balloons or short OA runs. As such, the modification procedure for a calcium length of less than 5 mm may be determined to be scoring balloons or short OA runs. As another example, a calcium length of more than 5 mm may require more aggressive and / or combination therapies. For example, OA may be used to debulk the lesion, followed by IVL to achieve deeper calcium modification. Long lesions also increase the risk of stent under-expansion and necessitate thorough lesion preparation. As such, the modification procedure for a calcium length of more than 5 mm may be determined to be OA and / or IVL.

[0154] The modification procedure may be performed, and post-procedural vessel data may be captured. The post-procedural vessel data can include, for example, intravascular imaging. The data collection system 100 may use the post-procedural vessel data to determine the effectiveness of the modification procedure. In some examples, the data collection system 100 may use vessel data taken during the procedure, for example when the procedure is paused and new image data is taken, such as through a pullback of an intravascular imaging probe. The intravascular imaging probe may be, for example, an OCT probe, IVUS probe, and / or the like. The system can identify a percent improvement of the target vessel between instances of imaging taken at different points before, during, and / or after the suggested procedure. The percent improvement can be determined, for example, by the same or different Al model. Example indicators for improvement can be based on changes in fractures in the calcium in the vessel, and / or changes in the level of stenosis.

[0155] Vessel data and / or patient data that can be included as features in training data examples can include one or more of the following: age, presence of hypertension, dyslipidemia, or diabetes, whether the patient is a current or former smoker, BMI, whether the patient had a prior myocardial infarction, renal insufficiency, and angina scoring from 0 to IV. Additional examples include characteristics of target vessels, including left anterior descending, circumflex, right coronary artery, or left main. Other examples can include the amount of severe calcification identified in the target vessel, as well as lesion location, including whether the lesion is proximal, mid, distal, or ostial. Other lesion characteristics include lesion length and calcification length. Other examples include patient historical data such as other conditions the patient may have, patient allergy data, patient medication data, or patient product-specific adverse event data including but not limited to hypersensitivity reactions, excess bleeding, and / or product performance failures. Patient historical data is stored securely and in compliance with any applicable privacy regulations. The patient historical data may be useful if a patient is receiving a subsequent procedure at the same facility or if the patient is receiving a procedure at a different facility from the facility in which the patient historical data was collected. The training data may be from cadaveric and / or living subjects. The training data examples can include a reference vessel diameter, a minimum lumen diameter, and / or a diameter stenosis. The training data may be readily available and may include, for example, pre-operative image data, historical patient, and / or indications of post-operative success or post-operative failures. In some examples, the training data may be publicly available, for example, with anonymized individual patient information for privacy. 1

[0156] If a percutaneous coronary intervention (PCI) was previously performed, other example features can include a reference vessel diameter, a lumen diameter, a diameter stenosis, a stent length, and / or acute gain. Post-PCI final complications can also be included as features, including severe dissection, slow flow or no reflow, abrupt closures, and / or perforations in the vessel.

[0157] Other examples of features that may be provided as part of the training data, include the following, for example in reference to calcified coronary lesions identified using OCT : lesion length, minimal lumen area, mean lumen area, area stenosis, calcium length, maximum continuous calcium arc, mean calcium arc, minimum calcium thickness, and / or calcium volume index. Other examples include characteristics related to a pre-IVL at minimum lumen area (MLA) sites, area stenosis percentage, lumen area, calcium angle, and / or max calcium thickness. Other examples include characteristics related to a post-stent at an MLA site, such as area stenosis, stent area, stent expansion, acute area gain, and calcium fracture. Other examples include characteristics related to pre-IVL at a maximum calcium site, such as area stenosis, lumen area, calcium angle, and / or maximum calcium thickness. Other examples include characteristics related to pre- IVL at the final minimal stent area (MSA) site, such as area stenosis, stent area, and / or stent expansion, acute area gain, and / or calcium fracture. Other examples include characteristics related to post-stent at the final MSA site, including area stenosis, stent area, stent expansion, acute area gain, and / or calcium fracture.

[0158] Other examples of features that may be provided as part of the training data include OCT characteristics for calcium fracture, such as the presence of any fracture, one fracture, two fractures, or more than three fractures, fracture length, fracture depth, maximum calcium arc at a calcium fracture, minimum calcium angle at the calcium fracture, calcium thickness at the calcium fracture, calcium fractures per lesion, calcium fracture per mm, lumen gain at fracture site, lumen area at fracture site, and / or stent expansion percentage. Other examples of features that may be provided as part of the training data include characteristics for OCT symmetry, eccentricity, and malapposition characteristics.

[0159] Other examples of features that may be provided include whether the patient has stable ischemic heart disease, acute coronary syndrome (non-ST-elevation myocardial infarction (NSTEMI) or unstable angina), or stabilized recent ST-elevation myocardial infarction (STEMI). Table 1, below, shows example features that can be provided as part of training examples in the training data.TABLE 1

[0160] The selected sub-set reflects a pragmatic and evidence-informed approach to feature selection, balancing clinical significance, technical feasibility, and data availability.

[0161] This sub-set is expected to be a strong foundation for training Al models because of clinically established predictors of procedural success. Lesion length, calcium arc, calcium thickness, and minimum lumen area have been established as predictors of stent under-expansion and procedural complexity. For example, a calcium arc of >180°, a calcium thickness of >0.5 mm, and a calcium length of >5 mm arefeatures associated with poor stent expansion. Additionally, minimum lumen area and diameter stenosis are routinely used to assess lesion severity and guide PCI decisions.

[0162] This sub-set is also expected to be a strong foundation for training Al models because of alignment with imaging modalities. The features in Table 1 are measurable using OCT and / or IVUS, which are the primary imaging modalities integrated into the system. As such, the Al model may be trained on data that is both accessible and reliable in real-world clinical workflows.

[0163] This sub-set is also expected to be a strong foundation for training Al models because of balanced representation in pre- and post-procedural data. The sub-set in Table 1 includes both pre-intervention (such as calcium angle and / or lesion length) and post-intervention (such as stent expansion and / or lumen gain) metrics. As such, the Al model may be trained from the full procedural context and may therefore better predict outcomes or suggest modifications.

[0164] This sub-set is also expected to be a strong foundation for training Al models because of practicality and interpretability. The features in Table 1 are interpretable by clinicians and align with existing scoring systems (such as OCT-based calcium scores and / or IVUS-based predictors). This facilitates clinical acceptance and validation of Al-generated recommendations.

[0165] This sub-set is also expected to be a strong foundation for training Al models because of data availability and consistency. The features in Table 1 are commonly recorded in clinical practice and registries, so these features are more likely to be available in sufficient quantity and quality for Al model training. Their inclusion supports generalizability and robustness of the Al model.

[0166] In these examples, the Al model can be trained according to a supervised learning problem for classifying changes in vessels from input image data and predicting a resulting percentage change or improvement. For example, the Al model can be trained with training examples labeled according to a percentage change, with model parameters updated using backpropagation with gradient descent and weight update. The objective function can be a type of distance-based loss, e.g., LI or L2 loss. The system can compare a predicted change from the Al model with a predetermined threshold for improvement, which can be set, for example, by a technician or physician performing the procedure.

[0167] According to some examples, the modification procedure performed on the vessel and / or postprocedural vessel data may be associated with the initial information, characteristics, and images. Collectively, the information, characteristics, images, performed procedure, and post-procedural vessel data may be used to form a database of training data for training a model to identify a suggested procedure. For example, the database may provide an indication that a given modification procedure was effective for a lesion of “x” degrees, “y” thickness, and “z” length while a different modification procedure was ineffective for a similar lesion.

[0168] During execution, the model trained to provide a suggested procedure, e.g., a suggested lesion modification procedure, can receive, as input, current vessel data. Current vessel data may be vessel data that is collected by data collection system 100 and is not yet part of the training data for the model. In some examples, in addition to or as an alternative, the model may receive, as input, the information associated with the vessel and / or the characteristics of the lesion within the vessel.

[0169] Machine learning (ML) can be used in order to provide preparation guidance. Specifically, it can be useful to bain a ML algorithm that includes labeled OCT data and the various vessel preparation techniques that could be used along with a predicted procedure outcome. The labeled OCT data can include calcium information from the Artificial Intelligence AI / ML algorithms that can be used to detect calcium and the external elastic lamina (EEL) of a vessel. An imaging and assessment system for detecting EEL based metrics is disclosed in published U.S. Pat. Pub. No. 2020 / 0294659.

[0170] The Al model can include a model or module for image processing. For example, the Al model can include a convolutional neural network or one or more convolutional layers, one or more pooling layers, and one or more fully-connected layers or other neural network layers. Other example architectures that may be at least partially implemented include other types of deep neural networks, recurrent neural networks (RNNs), such as long short-term memory (LSTMs), generative adversarial networks (GANs), and / or transformer architectures, including transformers with self-attention and / or cross-attention mechanisms.

[0171] Any of various different activation functions, model layer sizes, types, and / or other model hyperparameter values can be used. For example, the model can implement rectified linear unit (ReLU), Swish, or tanh, etc., as example activation functions.

[0172] The Al model can be trained, for example by the OCT system or another system, to perform one or more machine learning tasks. The Al model may be trained to perform separate tasks in a pipeline, or to perform one task end-to-end. For example, the Al model may be trained to receive image data, e.g., intravascular and / or extraluminal image data, and provide, as output, characteristics associated with the blood vessel, e.g., form or tissue nature, a classification of the type of lesion depicted in the image data, location of lesions, calcium, or other abnormalities within the vessel, and / or the size or severity of the lesion, and so on. The Al model can be trained according to a supervised learning approach to perform this task. For example, the Al model can receive training data including instances of image data, annotated with characteristics the Al model is trained to predict. The Al model can be initialized with random or semirandom model parameter values and receive and process the training data to generate a set of predictions. A loss between the model predictions and the annotations can be computed using one of several possible loss functions, e.g., LI or L2 loss, Jaccard loss, cross-entropy loss, or any of a variety of other loss functions for measuring the difference between predicted output and ground truth labels.

[0173] The total error across the individual losses computed by processing the training data can be used to perform backpropagation with model parameter update, for updating model parameter values of the Al model. The Al model can be trained across to one or more iterations of this processing / forward pass, loss computation, and backpropagation with model parameter updates. The Al model can be trained until one or more stopping criteria can be met. Stopping criteria can include a maximum wall clock time exceeded, a minimum loss improvement between iterations not being met, a minimum accuracy in the model output being met, and so on. Any of a variety of techniques can be used for training the model or improving model training and / or the performance of the trained model, e.g., stochastic gradient descent with model parameterupdate, batch gradient descent, gradient descent with momentum, regularization, drop-out, and / or schedules for changing model hyperparameters such as learning rate, batch size of training inputs, etc.

[0174] The Al models can be trained or fine-tuned through an offline training process, an online training process, or a combination of the two. For example, the Al model may be periodically retrained based on the contents of the updated database or fine-tuned using updated examples. In some examples, such as when the system receives feedback on the suggested procedure from an operator, the Al model can be finetuned using individual or batches of vessel data annotated with the feedback. This may occur, for example, when the model is already deployed on the system or another device in communication with the system, or updated offline and then used to update the model at a future point in time.

[0175] The Al models can further include components or models for generating recommendations based on received or generated OCT image data. For example, the Al model can perform an end-to-end process, in which OCT image data is generated by processing a received OCT image, and then generating a vessel preparation recommendation based on the learned OCT image data. As described herein with reference to Figure 5B, the Al model can be trained according to a multi-class classification task, in which the classes correspond to different vessel preparation options. The model can output an indication associated with a suggested vessel preparation process. The system implementing the model can further process the model output to provide, for example, data corresponding to the suggested vessel preparation process, including instructions, images, and / or other data that may be needed or beneficial to the user performing the vessel preparation.

[0176] The system can also prompt for user input to confirm or reject the recommended vessel preparation. The system can receive user input in the form of a confirmation / rejection indication, and possibly additional information explaining why the recommended vessel preparation was rejected or accepted. The suggested recommendation, along with the OCT image data and the user confirmation / rejection can be stored in the database 199, which can be accessed later for fine-tuning the Al model with updated training examples.

[0177] The system can also prompt for user input to limit the suggested procedures to available treatments within a treatment facility, such as a hospital. The system can receive user input regarding information about available treatments, including specifications of available treatments (such as a diameter and / or a length of available balloons) and / or brands of available treatments. The system can then output suggested procedures corresponding to the input available treatments and their compatibility with treating the specified lesion(s).

[0178] Figures 4 and 5 illustrate an example of using OCT as an imaging based modality for training an Al model, such as a convolutional neural network (CNN), whose output suggested procedures can be used for building a database, as described below and for suggested vessel preparation options and potential outcomes. As described above, the Al model can include one or more modules implementing a CNN or neural network layers of a CNN.

[0179] In Figure 4, an OCT procedure can be conducted including one or more pullbacks (as described above) after which a vessel preparation can be conducted, with input of the aspects and parameters of thepreparation into the Al model, along with outcome or results of the vessel preparation. These steps are primarily used in the building stage of the Al model.

[0180] Figure 5B is a flow diagram of an example process 500B for training the Al model to generate suggested vessel preparation options, according to aspects of the disclosure. A system of one or more processors, such as the system 100, can perform the process 500B. Some operations in the processes described herein can be omitted or performed multiple times, for example iteratively or in parallel. In some examples, other operations are added to the processes and / or performed in different orders.

[0181] The system receives training data including examples of OCT image data annotated with labels corresponding to a type of vessel preparation performed on the vessel characterized in the OCT image data, according to block 510B. The training data can include OCT image data as described herein, which may be stored in a database, for example the database 199, along with labels annotating the type of vessel preparation that was performed, for example on a patient from whom the OCT image data was taken. As described herein with reference to flowcharts 2-4, the database can be managed and updated by the system, which in turn can provide for additional training examples for further training or fine-tuning the Al model.

[0182] The training data can further include other data, such as historical data corresponding to the patient corresponding to the OCT image data. The historical data can include other medical information about the patient different than what is found in the OCT image data, such as other conditions the patient may have, patient allergy data, patient medication data, or patient product-specific adverse event data including but not limited to hypersensitivity reactions, excess bleeding, and / or product performance failures. Patient historical data is stored securely and in compliance with any applicable privacy regulations. The patient historical data may be useful if a patient is receiving a subsequent procedure at the same facility or if the patient is receiving a procedure at a different facility from the facility in which the patient historical data was collected. The training data can also include outcomes of performing the vessel preparation technique on the vessel indicated by the OCT image data.

[0183] The system processes the training data through the Al model to determine one or more losses, according to block 520B. For example, the Al model can generate an indication of a suggested vessel preparation process from a list of multiple classes. The system can generate a loss between the generated indication and a ground-truth label indicating the vessel preparation associated with the input training example. The system can compute any of various different loss functions for multi-class classification problems, such as categorical cross-entropy loss. For generating the indication corresponding to a suggested recommendation, the model can implement a softmax layer to generate output probabilities corresponding to the various different vessel preparation processes represented by the multiple classes. The system can receive the output probabilities to determine the suggested recommendation, for example, based on the suggested recommendation with the highest probability in the list of the output probabilities.

[0184] The system updates one or more model parameter values based on the determined one or more losses, according to block 530B. The system can perform, for example stochastic or batch gradient descent with backpropagation to update the model parameter values of the Al model. Any of various different techniques for machine learning training, including supervised learning approaches, can be used to trainthe model. The system can calculate a loss based on a cross-entropy loss between predicted classes and ground-truth labels in the training data.

[0185] In some examples, the model can be further trained according to one or more unsupervised learning techniques. For example, training in an unsupervised context can be based on clustering groups of unlabeled training data for finding similarities or differences between the examples. In other examples, dimensionality reduction techniques can be applied for training the model to reduce the overall amount of data needed by the model to make a suggested vessel preparation procedure.

[0186] The system determines whether stopping criteria have been met, according to block 540B. As described herein, stopping criteria can include a maximum wall clock time exceeded, a minimum loss improvement between iterations not being met, and / or a minimum accuracy in the model output being met, and so on. If the stopping criteria have been met (“YES”), the system outputs the trained model, according to block 550B. Otherwise (“NO”), the system can perform another training iteration of processing the training data and updating the model parameter values, according to blocks 520B and 530B, respectively. Figure 5A represents the Al model after sufficient OCT procedures (with pullbacks) are conducted and input into the Al model along with any other relevant patient data, such as biological information of the patient, wherein the trained Al model is ready to recommend one or more preparation methods or treatments with an optimal outcome, or with options as associated with probable outcomes possibly with risks or other options to consider. For example, the Al model can be trained on 250 or more examples.

[0187] For example, the trained Al model can be deployed in the system, so that when a new OCT pullback is conducted, the system can process the new data through the Al model for providing a recommendation on the most appropriate preparation method or technique, or a plurality of options from which to choose. In some examples, the OCT system can include a display device onto which any lesion type information can be displayed for operator review and / or any vessel modification or treatment regimen that is recommended or provided as an option.

[0188] Figures 6-8 illustrate example methods of building, deploying and using the above techniques in creating an ultimately usable database for analyzing a lesion, such as using an OCT system for imaging a lesion and for recommending and / or predicting the outcome of a particular treatment method or technique.

[0189] Figure 6 is an example logic flow and data to be provided by OCT to create a database of vessel preparation and outcomes. The system can generate image data corresponding to a pre-treated OCT area of interest, according to block 810. The system can generate the vessel data as described herein or receive the OCT image data from another source. Vessel data can include, for example, data characterizing a lesion arc, morphology, lesion length, and / or vessel diameter. The OCT image data can be used as part of a record stored in a database, such as the database 199 as shown and described with reference to FIG. 1. Additionally, the system may co-register two or more representations of image data as described herein. For example, the system may co-register two or more overlapping pullbacks of one or more intravascular imaging devices, such as an optical coherence tomography (OCT) probe, an intravascular ultrasound (IVUS) catheter, micro-OCT probe, near infrared spectroscopy (NIRS) sensor, optical frequency domainimaging (OFDI), any other device that can be used to image a blood vessel, or a combination thereof. The system can also receive patient historical data.

[0190] The system receives an input proposed vessel preparation process corresponding to the vessel data, according to block 820. Vessel preparation is performed, for example by an automated device or a user, such as a physician, according to block 830. Vessel data is generated or received for the area of interest after the procedure, according to block 840. The vessel preparation outcome can be recorded as part of the subject database record, according to block 850. The system can receive an indication as to whether an additional or different treatment is required, according to block 860. Vessel data with an indication of the corresponding vessel preparation technique to perform can be recorded as items in the database 199 and used to train the Al model to generate suggested recommendations on new data. Updates to the database, including updated vessel data, updated patient data, outcomes of performed procedures, and updates to procedure capabilities can be provided as data for additional training examples for training or fine-tuning the Al model.

[0191] Figure 7 is an example logic flow as the database of OCT information and vessel preparation outcomes is created and the model becomes suggestive of which vessel preparation to use. Ability for user to agree or disagree with vessel preparation suggestion to continue to enhance the database. The system generates or receives vessel data for an area of interest, according to block 910. The system outputs a suggested vessel preparation, according to block 920. The system can prompt for and receive user input indicating whether the suggested vessel preparation is correct or not, according to block 930. The system can update the database 199 based on the user input provided according to block 930, for example to add an additional database record including the vessel data and either the suggested vessel preparation (if the model was correct) or a different vessel preparation technique (if the suggested vessel preparation was incorrect). The vessel preparation can be performed, according to block 940, and updated vessel data can be received or generated on the area of interest following the procedure, according to block 950. The vessel preparation outcome can be inputted to the Al model, to determine whether additional or different vessel preparation is needed, according to block 970. If additional or different vessel preparation is required, the system can receive the vessel data generated as in block 960 and provide a new recommendation, according to block 920.

[0192] Figure 8 illustrates suggested procedures, e.g., vessel preparation modalities, based on type of lesion. The multiple classes used as possible classifications by the Al model can include these and possibly other modalities. For example, the model can be trained to output an indication that a percutaneous transluminal coronary angioplasty (PTCA) balloon preparation method is recommended. Example classes as shown in flowchart 4 include PTCA balloon, scoring balloon (for mixed plaque), scoring balloon (for fibrous plaque), IVL, OAS (for calcific lesions with more than a 180 degree arc), or IVL (for calcific lesions with less than a 180 degree arc). As described herein, the model can be trained to generate a list of output probabilities, with each probability representing a prediction by the model of how likely the respective suggested recommendation corresponds to the probability of being the correct vessel preparation to perform in view of the received OCT image data.

[0193] The suggested procedures can be limited to available treatments within a treatment facility, such as a hospital. The system can receive user input regarding information about available treatments, including specifications of available treatments (such as a diameter and / or a length of available balloons) and / or brands of available treatments. The system can then output suggested procedures corresponding to the input available treatments and their compatibility with treating the specified lesion(s).

[0194] The suggested procedure can include instructions for a method of treating a patient. For example, the instructions may include limitations to a specific procedure, such as whether there are limitations on how long a lesion can be to be treated with IVL (such as IVL suitability for <20 mm lesions). In addition to instructions regarding lesion length limitations, the system may provide a variety of other contextspecific instructions to guide the user. These instructions may be dynamically generated based on lesion characteristics, patient-specific data, and institutional constraints, and are intended to enhance procedural safety, efficacy, and compliance with best practices. These instructions may include device compatibility and sizing, procedural sequencing, lesion- specific warnings, imaging and verification, and / or institutional or operational constraints. Device compatibility and sizing instructions may include balloon-to-artery ratio guidance (such as, “Use a balloon with a diameter l.lx the reference vessel diameter”), stent sizing recommendations, which may be based on proximal and distal reference diameters, and / or device length selection, which may be based on lesion length plus margins (such as “Select a stent at least 5 mm longer than the lesion”). Procedural sequencing instructions may include a pre-dilation requirement (such as “Perform pre-dilation with a semi-compliant balloon before IVL”), an adjunctive therapy (such as “Consider scoring balloon post-RA to optimize lesion compliance”), and / or a post-dilation pressure range (such as “Post-dilate to 16-18 atm using a non-compliant balloon”). Lesion- specific warning instructions may include calcium morphology alerts (such as “Nodular calcium detected — consider OA over IVL”), thrombus presence (such as “Thrombus detected — consider embolic protection filter”), and / or dissection risk (such as “Avoid high-pressure dilation due to eccentric calcium and thin vessel wall”). Imaging and verification instructions may include imaging follow-up (such as “Repeat OCT pullback after lesion modification to assess calcium fracture”) and / or landing zone confirmation (such as “Ensure stent landing zones are free of calcium >180° arc”). Institutional or operator constraint instructions may include a device availability filter (such as “Recommended device not available — select alternative from approved list”) and / or instructions for use (IFU)-based constraints (such as “Device not indicated for vessels <2.5 mm in diameter”).

[0195] The suggested procedure can include updates to capabilities of each procedure. As each procedure is further developed, those additional capabilities can be added to a database, such as database 199, to output an updated suggested procedure.

[0196] The system may translate Al model outputs into actionable procedure and / or therapy options through a structured and interpretable pipeline. This process may include both internal model logic and optional post-processing steps. For example, the system may include structured post-processing and contextual filtering to ensure that the final output is clinically actionable, interpretable, and / or aligned with institutional and patient- specific constraints. In some examples, the Al model may be trained to output asingle class label (such as “IVL”, “OA”, “RA”, “scoring balloon”) and / or a ranked list of procedure options with associated confidence scores (such as IVL: 0.82, OA: 0.65, RA: 0.41). These outputs may be derived from a multi-class classification model, which may be trained on imaging data, annotations, and / or procedural outcomes. To translate the output into suggested procedures, the raw model output may be mapped to a procedure recommendation using a lookup table and / or rules engine that associates each class with a procedural name, device type and size ranges, indications and contraindications, and / or IFU-based constraints (such as lesion length, vessel diameter). This mapping may ensure that the output is a clinically meaningful recommendation. Optionally, additional post-processing may be applied to refine or filter the output. For example, post-processing may include device availability filtering (such as to remove options not stocked at the facility), IFU compliance check (such as to ensure the recommendation aligns with regulatory indications), conflict resolution logic (if multiple procedures are equally ranked, the system may prioritize a suggestion based on lesion morphology, operator preference, and / or historical success rates), and / or patient- specific contraindications (such as to filter out options based on allergy, comorbidities, and / or prior adverse events). The recommended procedure may be presented in a GUI, which may include a primary recommendation (optionally including a rationale), alternative options (optionally with confidence scores), visual overlays on imaging data (such as highlighting calcium arc or stent landing zones), and / or interactive elements for user feedback or override. A feedback loop may include logging user selections and outcomes to update the model over time, improving future recommendations and enabling personalization.

[0197] Figures 9-15 are example GUI displays that can be provided as output based on vessel data collected and / or processed using the data collection system 100 of Figure 1. The output includes at least one representation of the vessel data. As shown, the representations include a cross-sectional intravascular image, an angiography image including the vessel of interest, and a two-dimensional representation of the vessel. The two-dimensional representation is symmetrical about a longest axis of the axis. Each representation may include one or more indications corresponding to information associated with the vessel or characteristics of the lesion. The indications may include, for example, numerical values, e.g., a total angle of the lesion, a maximum thickness of the lesion, EEL values, lumen diameter values, and / or the like. The indication can be color coded, where each color represents a different type of tissue or lesion. In some examples, the indications can be color coded to indicate the severity of the lesion, and / or the like.

[0198] According to some examples, the display may include one or more inputs. The inputs can be configured to receive a user input corresponding to a selection, e.g., a selection from a drop down menu, or a typed input. As shown in Figure 9, the inputs may include an input for providing an indication of the vessel preparation that was performed and / or any post vessel preparation therapy performed. The vessel preparation and / or post vessel preparation may correspond to a lesion modification procedure. In some examples, a user, e.g., a physician, may provide this information such that the data collection system 100 stores this information in the database to be used for training the mode to provide a suggested procedure.

[0199] After the vessel preparation procedure is performed, post-procedure vessel data may be captured by data collection system 100. The post-procedural vessel data may be processed by the data collectionsystem 100 or another remote system to determine post-procedural information about the vessel and / or post-procedural characteristics of the lesion. Based on the post-procedural information and / or characteristics, the system may automatically determine a result of the vessel preparation, e.g., successful, needs additional modification, failure, and / or the like. As shown in Figure 10, in some examples, the display may include an input configured to receive a selection, e.g., a selection from a drop down menu, or a typed input associated with the vessel preparation result. The post-procedure vessel data may indicate that no lesions remain in the vessel. In some examples, when no lesions remain to be treated in the vessel, the outcome may be indicated as “good” or “successful.” In some examples, when the lesions within the vessel are below a first threshold value after the modification procedure, the outcome of the modification procedure may be determined to be good. The first threshold can correspond to a first threshold for length, thickness, and / or angle, etc. of the lesion(s). In examples where the lesions within the vessel are above the first threshold but below a second threshold after the modification procedure, the modification procedure may be determined to be “okay” or “needs additional modification.” The second threshold can correspond to a second threshold for length, thickness, and / or angle, etc. of the lesion(s). In yet another example, when the lesions within the vessel are above a third threshold after the modification procedure, the output of the modification procedure may be determined to be a failure. The third threshold can correspond to a third threshold for length, thickness, and / or angle, etc. of the lesion(s).

[0200] The information received via user inputs of Figures 9 and 10 may be saved in database 199 to be used as training data for the model(s).

[0201] Figures 11 and 12 illustrate example displays for providing suggested procedures based on the vessel data received via the data collection system 100 of Figure 1. As database 199 stores more information regarding vessel data and associated modification procedures, e.g., via user inputs of Figures 9 and 10, the data collection system 100 may execute one or more models to provide suggested procedures and / or subsequent procedures after an initial modification procedure.

[0202] For example, in Figure 11, the data collection system 100 may receive initial vessel data associated with at least a region of the vessel of interest. The data collection system 100, executing a model, may determine information associated with the vessel of interest and / or characteristics of the lesion(s) within the vessel of interest. The information associated with the vessel of interest and / or characteristics of the lesion(s) may be provided as input into a model which, when executed, provides a suggested procedure, e.g., “suggested vessel preparation procedure.” In some examples, the model, or another model, may provide a subsequent suggested procedure, e.g., “suggested post prep” procedure. The same Al model as described herein can be used for suggesting a post-preparation procedure, for example by providing updated vessel data to the Al model for generating suggested procedures to be performed as part of postpreparation. The subsequent suggested procedure may be a PCI procedure to be performed after the lesion modification procedure.

[0203] Figure 12 illustrates an example display provided for output by the data collection system 100 of Figure 1. The output may be provided based on vessel data received after the suggested procedure is performed and / or completed. For example, after the suggested procedure is performed, another pullbackof probe 104 may be completed such that the imaging subsystem 108 receives post-procedural vessel data. The data collection system 100 may determine, by executing a model, post-procedural information associated with the vessel and / or the lesion(s) within the vessel. As shown in Figure 12, the post-procedural vessel data indicates that a small amount of localized lesion, e.g., calcium, is still present. The model may identify, based on the remaining amount of lesion, an additional procedure, e.g., “suggested further prep” procedure, prior to a post vessel preparation procedure.

[0204] Figures 13 and 14 are similar to Figures 11 and 12, but show a suggested vessel preparation of “OAS - 1.25 mm Classic.” In Figure 13, the imaging system may provide a suggested procedure, including a suggested device for performing the procedure. Based on the post-procedural vessel data, the data collection system 100 may provide a suggested subsequent procedure, e.g., a subsequent vessel preparation procedure, as shown in Figure 13.

[0205] Figure 15 is similar to Figures 12 and 14, however Figure 15 shows a suggested treatment of “stent” and a total angle of zero degrees. Rather than provide for output a suggested subsequent procedure, e.g., a suggested further preparation procedure, the data collection system 100 may determine and provide for output a suggested treatment. Example suggested treatments include applying drug-coated balloons; stent placement, such as bare metal stents or drug eluting stents; and balloon angioplasty. The data collection system 100 can generate these suggested treatments, for example in addition to suggested vessel preparation procedures.

[0206] In another aspect the present invention may include imaging to detect the need for an embolism filter. The Al model may be trained to detect thrombus using intravascular imaging modalities. For example, in OCT images, thrombus may appear as low-signal, irregular intraluminal masses with signal- free shadowing. In IVUS images, thrombus may appear as echolucent or echogenic masses protruding into the lumen. In NIRS images, lipid-rich plaques prone to rupture and thrombus formation may be detected. The Al model may be trained to recognize these features in intravascular imaging modalities and to classify thrombus presence, type (such as red vs. white thrombus), and / or burden (such as length, volume, and / or degree of occlusion). The presence of thrombus is a known risk factor for distal embolization during interventional procedures such as atherectomy, balloon angioplasty, and / or stent deployment. Manipulation of thrombus-laden lesions can dislodge material, potentially leading to microvascular obstruction, noreflow phenomenon, or distal embolization and infarction. To mitigate these risks, the Al model may recommend embolic protection devices (such as filters or occlusion balloons). These devices may be suggested if the thrombus burden exceeds a defined threshold (such as >5 mm length or >50% lumen area) and / or if a procedure involves high-risk manipulation (such as orbital or rotational atherectomy). Upon detecting thrombus, the system may flag the lesion as high-risk, recommend embolic protection as part of the procedural plan, and / or suggest alternative strategies (such as deferred intervention and / or pharmacologic thrombus resolution). For example, thrombus information from a NIRS scan or an OCT scan may trigger a suggestion that an embolism filter be used. For example, the Al model can receive vessel data including image data from an IVUS, NIRS or OCT pullback, and generate a recommendation to use an embolism filter as a suggested vessel preparation procedure. An embolism filter may be utilized duringinterventions in vessels containing thrombus to prevent embolic debris from traveling downstream and causing blockages in smaller vessels. During some procedures, such as angioplasty or atherectomy, manipulating a thrombus may dislodge debris. An embolic filter may capture dislodged debris, preventing it from causing blockages in smaller vessels downstream.

[0207] It is also noted that the user could in all cases still add feedback to the database as to the outcome and machine learning could be used to continue to enhance the algorithm, which in turn keeps having more data backing the suggestions of vessel preparation and additional treatment.

[0208] Figure 16 depicts a flow diagram of an example process 1600 of providing a suggested procedure, for example using the data collection system of Figure 1, according to aspects of the disclosure. The following operations do not have to be performed in the precise order described below. Rather, various operations can be handled in a different order or simultaneously, and operations may be added or omitted.

[0209] The system receives, from an imaging modality, vessel data associated with at least a portion of a vessel, according to block 1610. For example, and as described with reference to Figures 4 and 5, the imaging modality may be OCT, although other modalities are possible, such as micro-OCT. Vessel data can include, for example, data characterizing a lesion arc, morphology, lesion length, and / or vessel diameter. Vessel data can also include patient data, for example, patient data corresponding to the patient whose vessels are being imaged. In some examples, the system can further determine whether the lesion is a medial lesion or an intimal lesion, and output, based on the determination, an indication as to whether the lesion includes the medial lesion or the intimal lesion.

[0210] The system determines, based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel, according to block 1620. As described above, initial characteristics can include, for example, numerical values, e.g., a total angle of the lesion, a maximum thickness of the lesion, EEL values, lumen diameter values, and / or the like. The indication can be color coded, where each color represents a different type of tissue or lesion. In some examples, the indications can be color coded to indicate the severity of the lesion, and / or the like. The lesion can be, for example, a calcium lesion.

[0211] The system provides, as input into an Al model, the determined initial characteristics of the lesion, according to block 1630. The Al model can be one model, or a composite of multiple models. For example, the system can determine the initial characteristics using a first Al model and provide the output of the first Al model to a second Al model. The Al model can be a CNN, a transformer, an RNN, such as an LSTM, and / or the like.

[0212] The system identifies, by executing the Al model, a suggested procedure for modifying the lesion within at least the portion of vessel, according to block 1640. For example, the Al model described with reference to Figure 8, above, can output a suggested procedure, such as a PTCA balloon, scoring balloon (for mixed plaque), scoring balloon (for fibrous plaque), IVL, OAS (for calcific lesions with more than a 180-degree arc), or IVL (for calcific lesions with less than a 180 degree arc).

[0213] The system provides, for output, the suggested procedure, according to block 1650. For example, the output can be provided to a GUI display and / or saved as part of a record of a database, e.g., the database 199, for updating the Al model.In some examples, the process 1600 can further include receiving, by the system, from the imaging modality after the suggested procedure is completed, post-procedural vessel data. The system can determine, based on the post-procedural vessel data, updated characteristics of the lesion.

[0214] Figure 17 depicts a flow diagram of an example process 1700 of providing a suggested procedure, for example using the data collection system of Figure 1, according to aspects of the disclosure. The following operations do not have to be performed in the precise order described below. Rather, various operations can be handled in a different order or simultaneously, and operations may be added or omitted.

[0215] The system receives, from an imaging modality, intravascular imaging data of a vessel, according to block 1710. For example, and as described with reference to Figure 1, the intravascular imaging data can be received from an image probe, such as for OCT, NIRS, IVUS, and / or micro-OCT.

[0216] The system determines, based on the intravascular imaging data, information associated with the vessel, according to block 1720. For example, and as described with reference to Figure 1, the system can determine a detailed morphological assessment of the lesion and assessment of the result of lesion modification techniques.

[0217] The system determines, based on the intravascular imaging data, a presence of a lesion within the vessel, according to block 1730. As described herein, the system can provide an indication of the presence, for example as a type of visual indicator.

[0218] The system determines, based on the intravascular imaging data, characteristics of the lesion, according to block 1740. The characteristics of the lesion may include, for example, the presence of lesion, location of lesion, lesion type, and / or lesion arc / coverage, etc.

[0219] The system provides, as input into an Al model, the characteristics of the lesion, according to block 1750. The Al model can be one model, or a composite of multiple models. For example, the system can determine the presence of the lesion and / or the characteristics of the lesion using a first Al model and provide the output of the first Al model to a second Al model. The Al model can be a CNN, a transformer, an RNN, such as an LSTM, and / or the like. The Al model can receive other vessel data as input, for example data characterizing the imaged vessel and / or patient data of a patient corresponding to the imaged vessel.

[0220] The system identifies, by executing the Al model, a suggested procedure for modifying the lesion within at least within at least the portion of the vessel, according to block 1760. The system provides, for output, the suggested procedure, according to block 1770.

[0221] The system stores, in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database, according to block 1780. For example, the database can be database 199 of Figure 1.

[0222] The system updates, based on the database, the Al model, according to block 1790. For example, the system can perform one or more training iterations to update model parameter values of the Al model, based on updated data in the database.

[0223] Unless otherwise stated, the foregoing alternative examples are not mutually exclusive but may be implemented in various combinations to achieve unique advantages. As these and other variations and combinations of the features discussed above can be utilized without departing from the subject matterdefined by the claims, the foregoing description of the embodiments should be taken by way of illustration rather than by way of limitation of the subject matter defined by the claims. In addition, the provision of the examples described herein, as well as clauses phrased as “such as,” “including” and the like, should not be interpreted as limiting the subject matter of the claims to the specific examples; rather, the examples are intended to illustrate only one of many possible embodiments. Further, the same reference numbers in different drawings can identify the same or similar elements.

Claims

Claims1. A system, comprising: one or more processors, the one or more processors configured to: receive, from an imaging modality, vessel data associated with at least a portion of a vessel; determine, based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel; provide as input into a model the determined initial characteristics of the lesion; identify, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; and provide for output the suggested procedure.

2. The system of claim 1, wherein the one or more processors are further configured to: receive, from the imaging modality after the suggested procedure is completed, post-procedural vessel data; and determine, based on the post-procedural vessel data, updated characteristics of the lesion.

3. The system of claims 1 or 2, wherein the one or more processors are further configured to: store, in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics; and update the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

4. The system according to any previous claim, wherein the imaging modality is optical coherence tomography or intravascular ultrasound.

5. The system according to any previous claim, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

6. The system according to any previous claim, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

7. The system according to any previous claim, wherein the lesion is a calcium lesion.

8. The system according to any previous claim, wherein the one or more processors are further configured to: determine, based on the vessel data, whether the lesion comprises a medial lesion or an intimallesion; and provide for output, based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

9. A method, comprising: receiving, by one or more processors from an imaging modality, vessel data associated with at least a portion of a vessel; determining, by the one or more processors based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel; providing, by the one or more processors, as input into a model the determined initial characteristics of the lesion; identifying, by the one or more processors by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; and providing for output, by the one or more processors, the suggested procedure.

10. The method of claim 9, further comprising: receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, post-procedural vessel data; and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

11. The method of claims 9 or 10, further comprising: storing, by the one or more processors in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics; and updating, by the one or more processors, the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

12. The method of any of claims 9 to 11, wherein the imaging modality is optical coherence tomography or intravascular ultrasound.

13. The method of any of claims 9 to 12, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

14. The method of any of claims 9 to 13, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

15. The method of any of claims 9 to 14, wherein the lesion is a calcium lesion.

16. The method of any of claims 9 to 15, further comprising: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

17. One or more non-transitory computer-readable storage media encoding instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: receiving, from an imaging modality, vessel data associated with at least a portion of a vessel; determining, based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel; providing as input into a model the determined initial characteristics of the lesion; identifying, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; and providing for output the suggested procedure.

18. The one or more non-transitory computer-readable storage media of claim 17, further comprising: receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, post-procedural vessel data; and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

19. The one or more non-transitory computer-readable storage media of claims 17 or 18, further comprising: storing, by the one or more processors in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics; and updating, by the one or more processors, the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

20. The one or more non-transitory computer-readable storage media of any of claims 17 to19, wherein the imaging modality is optical coherence tomography or intravascular ultrasound.

21. The one or more non-transitory computer-readable storage media of any of claims 17 to20, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

22. The one or more non-transitory computer-readable storage media of any of claims 17 to21, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

23. The one or more non-transitory computer-readable storage media of any of claims 17 to22, wherein the lesion is a calcium lesion.

24. The one or more non-transitory computer-readable storage media of any of claims 17 to23, further comprising: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

25. A system, comprising: one or more processors, the one or more processors configured to: receive, from an imaging modality, intravascular imaging data of a vessel; determine, based on the intravascular imaging data, information associated with the vessel; determine, based on the intravascular imaging data, a presence of a lesion within the vessel; determine, based on the intravascular imaging data, characteristics of the lesion, provide as input, into a model, the characteristics of the lesion, wherein the model is trained to identify a suggested procedure based at least in part on the information associated with the vessel and the determined characteristics of the lesion; identify, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; provide for output the suggested procedure; store, in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database; and update, based on the database, the model.

26. The system of claim 25, wherein the one or more processors are further configured to: receive, from the imaging modality after the suggested procedure is completed, post-procedural vessel data; and determine, based on the post-procedural vessel data, updated characteristics of the lesion.

27. The system of claim 25 or 26, wherein the imaging modality is optical coherencetomography or intravascular ultrasound.

28. The system of any of claims 25 to 27, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

29. The system of any of claims 25 to 28, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

30. The system of any of claims 25 to 29, wherein the lesion is a calcium lesion.

31. The system of any of claims 25 to 30, wherein the one or more processors are further configured to: determine, based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and provide for output, based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

32. A method, comprising: receiving, by one or more processors from an imaging modality, intravascular imaging data of a vessel; determining, by the one or more processors based on the intravascular imaging data, information associated with the vessel; determining, by the one or more processors based on the intravascular imaging data, a presence of a lesion within the vessel; determining, by the one or more processors based on the intravascular imaging data, characteristics of the lesion, providing as input, by the one or more processors, into a model, the characteristics of the lesion, wherein the model is trained to identify a suggested procedure based at least in part on the information associated with the vessel and the determined characteristics of the lesion; identifying, by the one or more processors executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; providing for output, by the one or more processors, the suggested procedure; storing, by the one or more processors in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database; and updating, by the one or more processors based on the database, the model.

33. The method of claim 32, further comprising:receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, post-procedural vessel data; and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

34. The method of claim 32 or 33, wherein the imaging modality is optical coherence tomography or intravascular ultrasound.

35. The method of any of claims 32 to 34, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

36. The method of any of claims 32 to 35, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

37. The method of any of claims 32 to 36, wherein the lesion is a calcium lesion.

38. The method of any of claims 32 to 37, further comprising: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

39. One or more non-transitory computer-readable storage media encoding instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: receiving, an imaging modality, intravascular imaging data of a vessel; determining, based on the intravascular imaging data, information associated with the vessel; determining, based on the intravascular imaging data, a presence of a lesion within the vessel; determining, based on the intravascular imaging data, characteristics of the lesion, providing as input into a model, the characteristics of the lesion, wherein the model is trained to identify a suggested procedure based at least in part on the information associated with the vessel and the determined characteristics of the lesion; identifying, by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; providing for output the suggested procedure; storing in memory, the information associated with the vessel, the determined characteristics of the lesion, and the suggested procedure in a database; andupdating, based on the database, the model.

40. The one or more non-transitory computer-readable storage media of claim 39, further comprising: receiving, by the one or more processors from the imaging modality after the suggested procedure is completed, post-procedural vessel data; and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

41. The one or more non-transitory computer-readable storage media of claim 39 or 40, wherein the imaging modality is optical coherence tomography or intravascular ultrasound.

42. The one or more non-transitory computer-readable storage media of any of claims 39 to41, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

43. The one or more non-transitory computer-readable storage media of any of claims 39 to42, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

44. The one or more non-transitory computer-readable storage media of any of claims 39 to43, wherein the lesion is a calcium lesion.

45. The one or more non-transitory computer-readable storage media of any of claims 39 to44, further comprising: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

46. A method, comprising: receiving, by one or more processors, vessel data associated with at least a portion of a vessel; determining, by the one or more processors based on the vessel data, initial characteristics of a lesion within at least the portion of the vessel; providing, by the one or more processors, as input into a model the determined initial characteristics of the lesion; identifying, by the one or more processors by executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; andproviding for output, by the one or more processors, the suggested procedure.

47. The method of claim 46, further comprising: receiving, by the one or more processors after the suggested procedure is completed, postprocedural vessel data; and determining, by the one or more processors based on the post-procedural vessel data, updated characteristics of the lesion.

48. The method of claims 46 or 47, further comprising: storing, by the one or more processors in memory as training data for the model, the initial characteristics, the suggested procedure, and the updated characteristics; and updating, by the one or more processors, the model based on the initial characteristics, the suggested procedure, and the updated characteristics.

49. The method of any of claims 46 to 48, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

50. The method of any of claims 46 to 49, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

51. The method of any of claims 46 to 50, wherein the lesion is a calcium lesion.

52. The method of any of claims 46 to 51, further comprising: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

53. A method, comprising: updating, by one or more processors based on post-procedural intravascular imaging data of a vessel and based on a database, a model, wherein the model is trained to identify a suggested procedure based at least in part on determined vessel characteristics and determined characteristics of the lesion, wherein updating the model includes: receiving, by the one or more processors, the post-procedural intravascular imaging data of the vessel; determining, by the one or more processors based on the post-procedural intravascular imaging data, vessel characteristics;determining, by the one or more processors based on the post-procedural intravascular imaging data, a presence of a lesion within the vessel; determining, by the one or more processors based on the post-procedural intravascular imaging data, characteristics of the lesion, providing as input, by the one or more processors, into the model, the characteristics of the lesion; identifying, by the one or more processors executing the model, a suggested procedure for modifying the lesion within at least the portion of the vessel; providing for output, by the one or more processors, the suggested procedure; and storing, by the one or more processors in memory, the determined vessel characteristics, the determined characteristics of the lesion, and the suggested procedure in the database.

54. The method of claim 53, wherein the characteristics associated with the lesion comprise at least one of a total angle, a thickness, or a length of the lesion.

55. The method of any of claims 53 to 54, wherein the suggested procedure is an orbital atherectomy, rotational atherectomy, intravascular lithotripsy, a scoring balloon, laser ablation, or an angioplasty balloon procedure.

56. The method of any of claims 53 to 55, wherein the lesion is a calcium lesion.

57. The method of any of claims 53 to 56, further comprising: determining, by the one or more processors based on the vessel data, whether the lesion comprises a medial lesion or an intimal lesion; and providing for output, by the one or more processors based on the determination, an indication as to whether the lesion comprises the medial lesion or the intimal lesion.

Images