System and method for non-destructive analysis for pre-cancer lesions and harvesting of cells from the same for culture

Abstract

Systems and methods are provided for identifying at least one of pre-cancerous or cancerous cells present in a whole organ or tissue of a subject. The method includes preparing the whole organ or tissue of the subject for imaging, exciting the whole organ or tissue of the subject with a beam of light, and acquiring imaging data from the whole organ or tissue of the subject. The method also includes processing the imaging data using a machine learning model to determine a presence of pre-cancerous or cancerous cells in the whole organ or tissue of the subject and generating a report identifying a location of pre-cancerous or cancerous cells in the whole organ or tissue of the subject.

Description

SYSTEM AND METHOD FOR NON-DESTRUCTIVE ANALYSIS FOR PRE-CANCERLESIONS AND HARVESTING OF CELLS FROM THE SAME FOR CULTURECROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit ofU.S. Provisional Patent Application 63 / 596,153, filed on November 3, 2023, and entitled “MACHINE LEARNING-GUIDED WHOLE-ORGAN OPTICAL IMAGING TECHNIQUE FOR NON-DESTRUCTIVE IDENTIFICATION OF EARLY PRE-CANCER LESIONS AND HARVESTING OF VIABLE CELLS FROM THE SAME FOR ORGANOID CULTURE,” the entire contents of which is hereby incorporated by reference, for any and all purposes.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0002] N / ABACKGROUND

[0003] The present disclosure relates generally to systems and methods for whole organ or tissue analysis and, more particularly, to systems and methods for non-destructive analysis of whole organ or tissue to determine a presence or absence of pre-cancer or cancerous cells and / or the harvesting of cells from the whole organ or tissue for cell culture.

[0004] In the new paradigm of Ovarian Cancer, most ovarian cancers are thought to have an origin on the Fallopian Tube (FT), resulting in efforts among pathologists to codify improved techniques for diagnosing and characterizing early precursor lesions on the fallopian tube. These early precursor lesions are termed as STIC (serous tubal intra-epithelial carcinoma). These efforts have resulted in the establishment of the SEE-FIM protocol (Sectioning and Extensively Examining the Fimbriated end), which is a multi-step, destructive analysis technique based on traditional histopathological methods. In this protocol, the FT specimens are FFPE processed (formalin fixation and paraffin embedding), followed by serial sectioning, and optional staining for IHC (immunohistochemistry) analyses for markers against p53 and Ki-67. This pathological analysis has become the standard practice for major hospitals, where it is generally performed on for all FT specimens excised from patients undergoing RRSO (risk reducing salpingo-oophorectomy)surgeries, as well as for those patients undergoing removal of their fallopian tubes in conjunction with other abdominal surgeries for removal of gynecological or other masses.

[0005] While the SEE-FIM protocol has been established as the “gold standard” for the identification of STICs, it is inefficient, tedious, and time-consuming. As a result, SEE-FIM has an inherently limited throughput and an inability to scale to high patient volumes. Moreover, the final pathological analysis is performed on a small, representative fraction (< 1%) of the whole FT tissue. As such, SEE-FIM has apparent detection rates of STICs of - 4% and - 1% in the FT tissue of high-risk and average-risk women, respectively. While the true prevalence of STICs is unknown in the literature, there is reason to believe it is significantly higher. This is suggested by a recent study (Visvanathan, K. et al. Fallopian Tube Lesions in Women at High Risk for Ovarian Cancer: A Multicenter Study. Cancer Prev. Res. (Phila. Pa.) 11, 697-706 (2018)) showing that, on average, the standard SEE-FIM analysis misses at least 50% of all STIC lesions which are present in the analyzed tissue fractions. Further still, SEE-FIM is a destructive analysis technique, in the sense that the FFPE processing and subsequent sectioning of the tissue precludes the use of many of the cutting-edge techniques, such as scRNA-seq and organoid culture that are necessary to study the underlying biology and unravel the molecular mechanisms at the onset of tumorigenesis and the transition from these early precancerous lesions to high grade serous carcinoma.

[0006] Thus, there is a need to develop new, efficient techniques as an alternative to the SEE-FIM protocol, for the rapid, high-throughput, non-destructive identification of STICs and other early precancer lesions, to improve patient care, as well as to harvest viable cells from these lesions to facilitate downstream analyses by the broader scientific community.SUMMARY

[0007] The present disclosure overcomes the aforementioned drawbacks by providing systems and methods for non-destructive, rapid, high-throughput identification of early precancer lesions, or STICs. The systems and methods provided herein are not limited to small tissue samples but can be applied to a whole-organ. In one non-limiting configuration, a machine learning-guided imaging system is provided.

[0008] In accordance with one aspect of the disclosure, a method is provided for identifying at least one of pre-cancerous or cancerous cells present in a fallopian tube of a subject. The method includes preparing the fallopian tube of the subject for imaging, exciting the fallopian tube of thesubject with a beam of light, and acquiring imaging data from the fallopian tube of the subject. The method also includes processing the imaging data using a machine learning model to determine a presence of pre-cancerous or cancerous cells in the fallopian tube of the subject and generating a report identifying a location of pre-cancerous or cancerous cells in the fallopian tube of the subject.

[0009] In accordance with another aspect of the disclosure, a system is provided for imaging tissue of a subject. The system includes a tissue positioning assembly configured to receive the tissue of the subject, an excitation source configured to excite the tissue of the subject with a beam of light, and an imaging system configured to acquire imaging data from the tissue of the subject. The system also includes a processor configured to receive the imaging data and process the imaging data using a machine learning model to identify a presence of cancerous or pre-cancerous cells in the tissue of the subject and generate a report identifying a location of pre-cancerous or cancerous cells in the tissue of the subject.

[0010] In accordance with yet another aspect of the disclosure, a method is provided for determining a presence of pre-cancerous or cancerous cells in excised fallopian tubes. The method includes preparing the fallopian tube of the subject for imaging by applying a dye to the fallopian tube, exciting the fallopian tube a beam of light, and acquiring imaging data from the fallopian tube of the subject during excitation by the beam of light. The method also includes processing the imaging data using a machine learning model to determine a presence of pre-cancerous or cancerous cells in the fallopian tube of the subject and generating a report identifying a location of pre-cancerous or cancerous cells in the fallopian tube of the subject.

[0011] The foregoing and other aspects and advantages of the present disclosure will appear from the following description. In the description, reference is made to the accompanying drawings that form a part hereof, and in which there is shown by way of illustration one or more embodiment. These embodiments do not necessarily represent the full scope of the invention, however, and reference is therefore made to the claims and herein for interpreting the scope of the invention. Like reference numerals will be used to refer to like parts from Figure to Figure in the following description.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Various objects, features, and advantages of the disclosed subject matter can be more fully appreciated with reference to the following detailed description of the disclosed subject matter when considered in connection with the following drawings, in which like reference numerals identify like elements.

[0013] Fig. l is a schematic illustration of a system in accordance with the present disclosure.

[0014] Fig. 2 is a schematic illustration of a functionalized near-infrared fluorescent nanoparticle that can be used for targeted fluorescence imaging in accordance with the present disclosure.

[0015] Fig. 3 is a schematic illustration of a model for processing imaging data in accordance with the present disclosure.

[0016] Fig. 4 is a flow chart setting forth some non-limiting example steps of a process in accordance with the present disclosure.

[0017] Fig. 5A is a set of example hyperspectral images of a first side of a right side fallopian tube demonstrating whole organ imaging using the systems and methods of the present disclosure.

[0018] Fig. 5B is a set of example hyperspectral images of a second side of a right side fallopian tube demonstrating whole organ imaging using the systems and methods of the present disclosure.

[0019] Fig. 6A is a set of example fluorescence images of side 1 a right fallopian tube of a patient using the systems and methods of the present disclosure, including a 50mm NIR imaging lens.

[0020] Fig. 6B is a set of example fluorescence images of side 2 a right fallopian tube of a patient using the systems and methods of the present disclosure, including a 50mm NIR imaging lens.

[0021] Fig. 6C is a set of example fluorescence images of side 1 a right fallopian tube of a patient using the systems and methods of the present disclosure, including a 25mm NIR imaging lens.

[0022] Fig. 6D is a set of example fluorescence images of side 2 a right fallopian tube of a patient using the systems and methods of the present disclosure, including a 25mm NIR imaging lens.

[0023] Fig. 7 is a set of panels of fluorescence images acquired with different parameters in accordance with the present disclosure.

[0024] Fig. 8 is a set of generated mask samples of the model of Fig. 3 applied to fallopian tubes’ hyperspectral images in accordance with the present disclosure.DESCRIPTION

[0025] As will be described, the present disclosure provides systems and methods for nondestructive, rapid, high-throughput identification of early precancer lesions, or STICs. Thesystems and methods provided herein are not limited to small tissue samples but can be applied to a whole-organ. In one non-limiting configuration, a machine learning-guided imaging system is provided. The underlying system may be referred to as the “OVASEEK” system and can be configured for rapid scanning of whole FTs, and automatic detection and identification of lesions using artificial intelligence or machine learning to classify regions of abnormal foci relative to healthy FT epithelium. Once identified using this non-destructive imaging technique, viable cells can be harvested from pre-cancer lesions for downstream molecular analyses and growing organoids of these cells in in vitro cell culture.

[0026] Referring to Fig. 1, one non-limiting example of a system 100 in accordance with the present disclosure is provided. The system 100 may include data acquisition system 102 that may include an imaging system 104. The imaging system 104 may include a local processor and memory for storing data and / or processing data, or controlling the overall data acquisition system 102. The imaging system 104 may be separated from an excitation source 106 such that a tissue positioning assembly 108 is positioned between the excitation source 106 and the imaging system 104.

[0027] The imaging system 102 may be connected (via wire or wirelessly) to a computing system 110. The computing system 110 can include a processor or controller 112, a display 114, one or more inputs 116, one or more communication systems 118, and / or memory 120. In some configurations, the controller 112 can be any suitable hardware processor or combination of processors, such as a central processing unit (“CPU”), a graphics processing unit (“GPU”), and so on. In some configurations, the display 114 can include any suitable display devices, such as a liquid crystal display (“LCD”) screen, a light-emitting diode (“LED”) display, an organic LED (“OLED”) display, an electrophoretic display (e.g., an “e-ink” display), a computer monitor, a touchscreen, a television, a screen of a mobile device, such as a phone or tablet, and so on. In some configurations, the inputs 116 can include any suitable input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, and so on.

[0028] In some configurations, the communications systems 118 can include any suitable hardware, firmware, and / or software for communicating information. In this regard, as illustrated, the communications system 118 can connect to the data acquisition system 102. Additionally or alternatively, the computing system 110 and / or data acquisition system can communicate over acommunication network 122 and / or any other suitable communication networks. For example, communications systems 118 can include one or more transceivers, one or more communication chips and / or chip sets, and so on. In a more particular example, the communications systems 118 can include hardware, firmware, and / or software that can be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on.

[0029] In some configurations, the memory 120 can include any suitable storage device or devices that can be used to store instructions, values, data, or the like, that can be used, for example, by the controller 112 or data acquisition system 102 to present content using display 114, to communicate with a server 124 via communications system(s) 118, and so on. The memory 120 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof. For example, the memory 120 can include random-access memory (“RAM”), read-only memory (“ROM”), electrically programmable ROM (“EPROM”), electrically erasable ROM (“EEPROM”), other forms of volatile memory, other forms of nonvolatile memory, one or more forms of semi-volatile memory, one or more flash drives, one or more hard disks, one or more solid state drives, one or more optical drives, and so on. In some configurations, the memory 120 can have encoded thereon, or otherwise stored therein, a computer program for controlling operation of computing device 110. In such configurations, the processor / controller 112 (and / or a local processor of the data acquisition system 102) can execute at least a portion of the computer program. In doing so, information can be presented (e.g., images, user interfaces, graphics, tables), content can be received from the server 124, and so on. For example, the controller 112 and the memory 120, and / or local processors and memory of the data acquisition system 102 can be configured to perform the methods described herein.

[0030] The server 124 can include a communications system 126 for communicating, for example, via the communication network 122. The server 124 can also include a processor / controller 128, a display 130, one or more inputs 032, and / or memory 134. As described with respect to the computing system, the controller 128 can be any suitable hardware processor or combination of processors, such as a CPU, a GPU, and so on. The display 130 can include any suitable display devices, such as an LCD screen, LED display, OLED display, electrophoretic display, a computer monitor, a touchscreen, a television, and so on. The input 132 can include any suitable input devices and / or sensors that can be used to receive user input, such as a keyboard, a mouse, a touchscreen, a microphone, and so on. The communications systems 126 can include hardware,firmware, and / or software that can be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection, an Ethernet connection, and so on. Finally, the memory 134 can include any suitable volatile memory, non-volatile memory, storage, or any suitable combination thereof.

[0031] In one non-limiting example, the data acquisition system 102 was constructed using a vertical mount system to hold a NIRvana 640 camera (Teledyne Princeton Instruments, NJ), along with a fixed focus lens (choice of 25 mm, 50 mm or 100 mm SWIR C-mount lenses from Edmund Optics, depending on the field of view desired). The excitation source 106 included a near-infrared laser excitation source. Such as system, or similar systems with different components can be used to perform any of a variety of imaging techniques, including hyperspectral imaging or fluorescence imaging, as will be described.

[0032] In the non-limiting illustration of Fig. 1, the system is shown in a trans-illumination configuration, for imaging whole organ excised fallopian tubes. As illustrated, the system may yield images 136 of STIC cells labeled 138 with fluorescent nanoparticles which have been functionalized with specific anti-STIC antibodies (e.g., Laminin Cl). In this non-limiting example, a single cell 140 was be labeled with fluorescent rare earth nanocrystals 142.

[0033] As another example, referring to Fig. 2, a schematic diagram is provided showing a layer- by-layer functionalized near-infrared fluorescent nanoparticle that can be used for targeted fluorescence imaging of whole organs in accordance with the systems and methods provided herein. That is, to perform targeted fluorescence imaging of the abnormal foci, the systems and methods provided herein may use a synthesized layer-by-layer nanoparticle. In one, non-limiting example as provided in Fig. 2, gold nanorods (AuNR) may be used as a core that can be decorated with a small molecule fluorescent dye. The fluorescent dye may be NIR-emitting. In one nonlimiting example, such a dye may include IR-E1050, Nirmidas Biotech, such as may be useful for plasmonic enhancement of the emission intensity of the fluorescent dye. As illustrated in Fig. 2, the nanoparticle may be wrapped in a layer-by-layer wrapping. In one, non-limiting example, the wrapping may be formed with alternating polymer layers. For example, one layer may include poly styrene sulfonate (PSS). Another layer may include poly ethylene imine (PEI). Following the wrapping, an antibody conjugation may be included for targeting specific surface markers on the surface of STICs. In one non-limiting example illustrated in Fig. 2, anti-LAMCl antibodies may be conjugated the nanoparticles, for example to specifically target laminin yl, a proteinupregulated in high grade serous ovarian cancer. However, other targets or other targeting antibodies or non-antibody targeting mechanisms may also be used.

[0034] As described above, traditional methods for detecting and analyzing abnormalities in the fallopian tubes can be costly, and often lack the precision required for early-stage identification. These methods frequently depend on expert analysis for accurate segmentation and diagnosis. While there have been limited attempts at automated, expert-free detection in this area, the systems described above can implement a segmentation approach that empowers improved detection while reducing reliance on clinical personnel.

[0035] In one non-limiting example, hyperspectral imaging is used as input for a trailed analysis model. The model may be designed to deliver detailed and efficient segmentation of the fallopian tubes, facilitating timely and accurate diagnosis. In particular, the systems of Fig. 1 may store or access a deep neural network (DNN) or artificial intelligence (Al) or machine learning (ML) model or system. In one non-limiting example, a hybrid DNN, referred to as MobileUNETR3, may be used. MobileUNETR3 is based on the traditional U-Net architecture, but integrates both transformer blocks and convolutional layers. Regardless of the particular model, a model may be chosen that leverages convolutional operators for capturing fine-grained local details and selfattention mechanisms for extracting global abstract features. This combination can be particularly effective and robust in processing the noise-prone hyperspectral images of fallopian tubes, enhancing segmentation accuracy. Such a model can ensure that each fallopian tube within the detected samples is accurately classified and segmented.

[0036] One non-limiting example of an architecture or model for use with the systems and methods described herein is provided in Fig. 3. In the non-limiting example in Fig. 3, a hybrid DNN architecture is used that combines the strengths of convolutional neural networks (CNNs) and transformers to create a lightweight and effective model for medical image segmentation. As illustrated in this non-limiting example, the primary computation layers include a hybrid encoderdecoder structure that efficiently, precisely balance local and global feature extraction. This hybrid design enables the model to achieve significant improvements in segmentation accuracy while maintaining a minimal memory footprint and computation complexity.

[0037] Also, at the encoder stage, the architecture or model of Fig. 3 employs a CNN-based module for local feature extraction and downsampling, followed by a MobileViT block that integrates both CNN and transformer components. This sequence and combination allows theencoder to capture both fine-grained details and global contextual information. In the decoder stage, in turn, the model can be optimized to analyze this combined feature information and recover the global and local features extracted at the early stage of the encoding process. Together, these design elements allow the model to surpass current state-of-the-art models in terms of efficiency, with a reduced parameter count and computation complexity, achieving impressive accuracy performance across multiple medical segmentation datasets.

[0038] Referring now to Fig. 4, some steps of one, non-limiting method 400 in accordance with the present disclosure is provided. Referring now to Fig. 4, some steps of one, non-limiting method 400 in accordance with the present disclosure is provided. While these steps will be described in general terms to illustrate the substantial flexibility of the systems and methods provided herein and the fact that the systems and methods provided herein are not limited to a particular implementation, these steps will be further illustrated with respect to actual implementations and tests performed with the operation of the systems and methods provided herein relative to clinically-relevant information.

[0039] That is, an example of the application of the OVASEEK imaging system was tested for whole-organ imaging of a human FT tissue specimen, from a patient undergoing salpingectomy. A right fallopian tube of a patient was tagged with the asset tracking bar code from Johns Hopkins. The fallopian tube tissue was successfully imaged and the model able to identify pre-cancerous cells.

[0040] Referring again to Fig. 4, at process block 402, the tissue is prepared for imaging. This may include excising the tissue and applying any imaging agent that is desired, if any. As one non-limiting example, imaging may include hyperspectral imaging. As an example, hyperspectral imaging may be performed without the addition of any exogenous contrast agent to bind or label the tissue and, as such, be “label free.” Additionally or alternatively, other imaging processes may be performed that are not “label free.” In this case, for example, after hyperspectral or other imaging without a label (or if no non-label imaging is to performed), the preparation at process block 402 may include applying a label, as will be described.

[0041] Regardless of the preparation process, at process block 404, an imaging process is performed to acquire imaging data. Referring to Figs. 5B and 5B, a hyperspectral data cube is provided for an entire set of hyperspectral images of the right FT of a patient. For hyperspectral imaging, two distinct light sources may be used. For example, a first light source may be a halogenbroadband source, to provide diffuse illumination for anatomical co-regi strati on, and a second light source may be a laser light source to provide near-infrared irradiation for studying the interaction of light with the tissue. The laser light source may have a rage of, for example, around or at 980 nm or 808 nm.

[0042] Previously, it has been reported that near-infrared imaging can identify the spectral and scattering signatures of tissues without a priori knowledge of the background or tissue autofluorescence. To achieve this goal, in one test, the above-described systems were used to perform NIR hyperspectral whole-organ imaging in the present OVASEEK imaging system using a 50 mm SWIR lens (Edmund Optics), along with the aforementioned light sources, and coupled with a band pass filter. A series of 11 BP (band pass) filters, ranging from 1050 nm to 1550 nm at 50 nm steps, each with an FWHM (full width at half maximum) of 50 nm, was used for this purpose. In Figs. 5A and 5B, three columns of data are shown for each band-pass filter: (i) the combined image with the background diffuse light source and the 980 nm laser; (ii) only background image with the diffuse illumination; and (iii) only laser 980 nm illumination. In particular, Fig. 5A and Fig. 5B show Side 1 and Side 2, respectively, of the right FT of the patient, in reflectance mode imaging configuration. Scale bar is 0-1 (a.u.) for each image.

[0043] Referring again to Fig. 4, as described, either as the only acquisition or as a subsequent past of data acquisition, a further imaging process may be performed. In one non-limiting example, the above-described hyperspectral data may be acquired at process block 404 and then the process 400 may loop back to process block 402 to prepare the tissue for a labeled imaging acquisition. That is, subsequent to a “label free” hyperspectral imaging, the above-described right FT was incubated with an aqueous suspension of the layer-by-layer nanoparticles, such as described with respect to Fig. 2, for specific targeting of the laminin yl protein. The right FT tissue was incubated in 2 ml of the nanoparticle solution, for a period of 2 hrs. After incubation, the tissue was washed 3 times with lx PBS to remove unbound nanoparticles, and imaged using OVASEEK. For fluorescence imaging, two distinct light sources were used. In one non-limiting example, the first source was a halogen broadband source, to provide diffuse illumination for anatomical coregistration, and the second source was a laser light source to provide NIR excitation which is close to the absorption maxima of the AuNR particles. In one non-limiting example, the laser source may be at or about the 808 nm wavelength. Figs. 6A-6D show the fluorescence imagesfor the entire set of images of the right FT of the patient, collected using the OVASEEK imaging system.

[0044] For the imaging process at process block 404 of Fig. 4, a 50 mm SWIR lens (Edmund Optics) was used to perform whole-organ, high throughput fluorescence imaging of the right FT, using a series of 5 BP (band pass) fdters, ranging from 1000 to 1200 nm at 50 nm steps, each with an FWHM of 50 nm. Figs. 6A and 6B show, respectively, the fluorescence images of Side 1 and Side 2 of the Right FT of the patient. Three columns of data are shown for each fdter: (i) the combined image with the background diffuse light source and the 808 nm laser; (ii) only background image with the diffuse illumination; and (iii) only laser 808 nm illumination.

[0045] Continuing with the imaging at process block 404 of Fig. 4, a 25 mm SWIR lens (Edmund Optics) was used to perform focused imaging on the fimbriated end of the right FT, using a set of 5 filters (1000 and 1125 nm BP; and 1000, 1100 and 1200 nm LP filters). Figs. 6C and 6D show, respectively, the fluorescence images of the fimbriated end of Side 1 and Side 2 of the Right FT of the patient. Three columns of data are shown for each filter: (i) the combined image with the background diffuse light source and the 808 nm laser; (ii) only background image with the diffuse illumination; and (iii) only laser 808 nm illumination.

[0046] Referring again to Fig. 4, at process block 406, the acquired data is analyzed using a model, such as described above with respect to Fig. 3. As described, the model may be a trained learning or artificial intelligence model. In one non-limiting analysis process, an anti-EpCAM antibody (CD326, eBioscience, Thermo Fisher Scientific), labeled with ICG (indocyanine green), was used as a control to target the epithelial cell adhesion molecule protein. EpCAM is a glycoprotein with a basal level of expression in many epithelial cell surfaces, including in normal, non-diseased tissue. To establish a positive control for the OVASEEK imaging system, in one non-limiting example, EpCAM with ICG were labeled, and a 1300 nm BP filter was used in combination with an 808 nm laser excitation, to image using our OVASEEK imaging system by leveraging the nearinfrared emission of ICG. Subsequent to the fluorescence imaging protocol as depicted in Figs. 6A-6D, the right FT tissue of the patient was incubated with an aqueous suspension of anti- EpCAM-ICG for 1 hr. After incubation, the tissue was washed 3 times with l x PBS to remove unbound antibodies, and imaged using OVASEEK.

[0047] These control images are provided in Fig. 7 and illustrate that control data can be acquired that can be used for training of a model in accordance with the present disclosure. Three columnsof data are shown for each filter: (i) the combined image with the background diffuse light source and the 808 nm laser; (ii) only background image with the diffuse illumination; and (iii) only laser 808 nm illumination. The upper and second panel of Fig. 8, respectively, show whole organ images of Side 1 and Side 2 of the right FT of the patient, imaged using the 50 mm SWIR lens, following incubation with the anti-EpCAM- ICG. The third and last panel of Fig. 8, respectively, show the fimbriated end images of Side 1 and Side 2 of the right FT of the patient, imaged using the 25 mm SWIR lens, following incubation with the anti- EpCAM-ICG.

[0048] Referring again to Fig. 4, at process block 408 a report, which may include images, images with computer aided diagnosis (CAD) indicators, and or written or combined text and images forming annotated reports can be generated. As the above-described process is non-destructive, optionally, at process block 410, the tissue may be delivered for culture and further analysis, including pathology analysis.

[0049] In one non-limiting example, the post-imaging process may include tissue inking of abnormal foci identified using the above-described whole organ, non-destructive imaging. In one non-limiting example, a waterproof black ink (Higgins Black Magic Ink) was used to tattoo the locations of abnormal foci identified on the tissue. Based on the analyses of fluorescence imagery described above, there were 6 abnormal foci identified; 4 on Side 1 and 2 on Side 2 of the right FT.

[0050] Once the abnormal foci have been identified, a low-shear, reversible cell capture device can be used for harvesting viable cells from these foci. The use of these devices, aided by the OVASEEK-guided non-destructive identification of these early pre-cancer lesions, allows the generation of in vitro models, including organoids, and performance of molecular analyses to better understand the biology and study the mechanism of the development of these pre-cancer lesions, as well as the progression from the onset to high grade serous cancer. This greatly enhances the understanding of the biology of the STICs to HGSOC transition, as the current methods for identification of STICs using FFPE processing preclude the harvesting of viable cells.

[0051] The above-described systems and methods were also tested using animal samples of organs. The implementation, referred to herein as the OVASEEK whole organ imaging system, was tested using such sample organs. To simulate a fluorescently-tagged STIC lesion, small amounts (~ 0.5 pL of aqueous-dispersed SWNTs) were applied, either on the surface, or embedded in a pipette tip buried subcutaneously (~ 2-3 mm depth) below the surface of the tissue. With anoverlay of the 808 nm laser (used for excitation of the fluorophore, in this case, the SWNTs), a diffuse background illumination source, such as from a halogen lamp, for anatomical coregistration, it was possible to exactly pinpoint the location of the bound nanoparticles.

[0052] As described above, a system and method are provided for Al-powered lesion segmentation on human fallopian tubes. The system and method achieved automated and rapid early detection and segmentation of lesion areas, minimizing the need for manual intervention and expert analysis. The system and method also achieved high segmentation accuracy, potentially surpassing expertlevel performance. Further, the system and method showed that the Al-based approach can be adapted for other types of cancer or medical conditions, providing a versatile methodology for image-based diagnostic tasks.

[0053] As one further, non-limiting example, a dataset of 560 hyperspectral images was assembled that included with 448 images allocated to the training set and 112 for a testing set. The results demonstrated an Intersection over Union (loU) = 85.1%, which is comparable to state-of-the-art medical segmentation performance. This outcome indicates the effectiveness of the systems and method provided herein to achieve clinically viable segmentation accuracy using hyperspectral images. In particular, Fig. 8 is a set of generated mask samples using the model of Fig. 3 and the fallopian tubes hyperspectral images from this testing.

[0054] It is to be understood that the present disclosure is not limited in its application to the details of construction and the arrangement of components set forth in the following description or illustrated in the following drawings. The present disclosure is readily extended to other aspects and implementations and may be practiced or carried out in various ways. Also, it is to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting. The use of “including,” “comprising,” or “having” and variations thereof herein is meant to encompass the items listed thereafter and equivalents thereof as well as additional items. Unless specified or limited otherwise, the terms “mounted,” “connected,” “supported,” and “coupled” and variations thereof are used broadly and encompass both direct and indirect mountings, connections, supports, and couplings. Further, “connected” and “coupled” are not restricted to physical or mechanical connections or couplings.

[0055] As used herein in the context of computer implementation, unless otherwise specified or limited, the terms “component,” “system,” “module,” “controller,” “framework,” and the like are intended to encompass part or all of computer-related systems that include hardware, software, acombination of hardware and software, or software in execution. For example, a component may be, but is not limited to being, a processor device, a process being executed (or executable) by a processor device, an object, an executable, a thread of execution, a computer program, or a computer. By way of illustration, both an application running on a computer and the computer can be a component. One or more components (or system, module, and so on) may reside within a process or thread of execution, may be localized on one computer, may be distributed between two or more computers or other processor devices, or may be included within another component (or system, module, and so on).

[0056] In the methods described herein, the steps can be carried out in any order without departing from the principles of the disclosure, except when a temporal or operational sequence is explicitly recited. Recitation in a claim to the effect that first a step is performed, and then several other steps are subsequently performed, shall be taken to mean that the first step is performed before any of the other steps, but the other steps can be performed in any suitable sequence, unless a sequence is further recited within the other steps. For example, claim elements that recite “Step A, Step B, Step C, Step D, and Step E” shall be construed to mean step A is carried out first, step E is carried out last, and steps B, C, and D can be carried out in any sequence between steps A and E, and that the sequence still falls within the literal scope of the claimed process. A given step or sub-set of steps can also be repeated.

[0057] Furthermore, specified steps can be carried out concurrently unless explicit claim language recites that they be carried out separately. For example, a claimed step of doing X and a claimed step of doing Y can be conducted simultaneously within a single operation, and the resulting process will fall within the literal scope of the claimed process.

[0058] The term “substantially” or “about” as used herein refers to a majority of, or mostly, as in at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, at least about 99%, at least about 99.5%, at least about 99.9%, at least about 99.99%, or at least about 99.999% or more.

[0059] As used in the claims, the phrase “at least one of A, B, and C” means at least one of A, at least one of B, and / or at least one of C, or any one of A, B, or C or combination of A, B, or C. A, B, and C are elements of a list, and A, B, and C may be anything contained in the Specification.

[0060] The following discussion is presented to enable a person skilled in the art to make and use aspects of the disclosure. Various modifications to the illustrated configurations or processes will be readily apparent to those skilled in the art, and the generic principles herein can be applied to other aspects and applications within the scope of the present disclosure and the understanding of one of skill based thereon. Thus, the present disclosure is not intended to be limited to particular embodiments or aspects shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein. The following detailed description is to be read with reference to the figures, in which like components or elements fin different figures have like reference numerals. The figures, which are not necessarily to scale, depict selected aspects and configurations or processes and are not intended to limit the scope of the disclosure. Skilled artisans will recognize the examples provided herein have many useful alternatives and fall within the scope of the disclosure.

Claims

CLAIMS1. A method of identifying at least one of pre-cancerous or cancerous cells present in a whole organ or tissue of a subject, the method comprising: preparing the whole organ or tissue of the subject for imaging; exciting the whole organ or tissue of the subject with a beam of light; acquiring imaging data from the whole organ or tissue of the subject; processing the imaging data using a machine learning model to determine a presence of pre-cancerous or cancerous cells in the whole organ or tissue of the subject; and generating a report identifying a location of pre-cancerous or cancerous cells in the whole organ or tissue of the subject.

2. The method of claim 1, wherein acquiring imaging data includes performing at least one of a hyperspectral imaging process or a fluorescence imaging process.

3. The method of claim 1, wherein the whole organ forms a whole fallopian tube of the subject.

4. The method of claim 1, wherein preparing the whole organ or tissue includes applying a nanoparticle formulation to the whole organ or tissue.

5. The method of claim 4, wherein the nanoparticle includes a dye.

6. The method of claim 4, wherein the nanoparticle is coupled with an antibody conjugation configured to target specific surface markers of cells in the whole organ or tissue of interest being imaged.

7. The method of claim 6, wherein the antibody conjugation includes anti-LAMCl antibodies configured to target a protein upregulated in ovarian cancer.

8. The method of claim 1, wherein exciting the whole organ or tissue includes directing a beam of light over the entire surface of the whole organ or tissue from at least one of a diffuse background illumination source or a laser.

9. The method of claim 1, wherein preparing the whole organ or tissue includes applying a nanoparticle formulation to the whole or organ tissue and further comprising acquiring hyperspectral images from the whole or organ tissue, prior to applying the nanoparticle formulation.

10. The method of claim 1, wherein acquiring imaging data from the whole organ or tissue includes acquiring fluorescent emissions from fallopian tube.

11. The method of claim 1, further comprising delivering the whole organ or tissue for cell culture harvesting.

12. A system for imaging tissue of a subject, the system comprising: a tissue positioning assembly configured to receive the tissue of the subject; an excitation source configured to excite the tissue of the subject with a beam of light; an imaging system configured to acquire imaging data from the tissue of the subject; a processor configured to: receive the imaging data and process the imaging data using a machine learning model to identify a presence of cancerous or pre-cancerous cells in the tissue of the subject; and generate a report identifying a location of pre-cancerous or cancerous cells in the tissue of the subject.

13. The system of claim 12, wherein the processor is further configured to process the imaging data to produce at least one of hyperspectral images of the tissue or fluorescence images of the subject.

14. The system of claim 12, wherein the excitation source includes a near-infrared laser excitation source or diffuse background illumination source.

15. The system of claim 12, wherein the tissue includes excised whole fallopian tubes, and the pre-cancerous or cancerous cells include serous tubal intra-epithelial carcinoma.

16. The system of claim 12, wherein the machine learning model includes a hybrid deep neural network (DNN) architecture that combines convolutional neural networks (CNNs) and transformers to create a medical image segmentation model.

17. A method for determining a presence of pre-cancerous or cancerous cells in excised fallopian tubes, the method comprising: preparing the fallopian tube of the subject for imaging by applying a dye to the fallopian tube; exciting the fallopian tube a beam of light; acquiring imaging data from the fallopian tube of the subject during excitation by the beam of light; processing the imaging data using a machine learning model to determine a presence of pre-cancerous or cancerous cells in the fallopian tube of the subject; and generating a report identifying a location of pre-cancerous or cancerous cells in the fallopian tube of the subject.

18. The method of claim 17, further comprising delivering the fallopian tube for harvesting of viable cells for cell culture.

19. The method of claim 17, wherein the dye includes a nanoparticle coupled with an antibody conjugation configured to target specific surface markers of cells in the fallopian tubes.

20. The method of claim 19, wherein the antibody conjugation includes anti-LAMCl antibodies configured to target laminin yl.21 . The method of claim 17, wherein exciting the fallopian tube includes directing a beam of light to the fallopian tube from at least one of a diffuse background illumination source or a laser.

22. The method of claim 17, further comprising delivering the fallopian tube for harvesting viable cells for cell culture.

23. The method of claim 17, wherein exciting the fallopian tube includes using a near-infrared laser with a wavelength of about 808 nm or 980 nm.

24. The method of claim 17, wherein illuminating the fallopian tube includes using a diffuse background source formed by one of a broadband halogen lamp, or an LED source operating at 1050 nm.

25. The method of claim 17, wherein acquiring imaging data comprises using a series of bandpass filters ranging from about 1000 nm to about 1550 nm.

23. The method of claim 15, wherein processing the imaging data comprises using a hybrid deep neural network architecture combining convolutional neural networks and transformers to analyze the imaging data acquired through a series of bandpass filters.

Images