Device for hybrid ultrasound and optical imaging for early cancer detection

Abstract

A device for hybrid ultrasound and optical imaging for early cancer detection, comprising: • an ultrasound imaging module that can capture high-resolution structural images of soft tissues for real-time visualization of suspected lesions; • an optical imaging module comprising at least one of photoacoustic imaging (PAI) and diffuse optical tomography (DOT) to acquire functional and molecular imaging data, including hemoglobin concentration, oxygen saturation and metabolic activity of tissues; • a multimodal image fusion machine configured to integrate ultrasound and optical image data to provide comprehensive, high-contrast diagnostic visualization; • a data processing unit with an AI-powered machine learning module trained on large cancer datasets for automatic tumor detection, segmentation, classification, and malignancy prediction; • a communication interface for real-time data transmission that supports both wireless and wired connections to enable remote monitoring, data exchange, and cloud-based AI analytics; • An interactive user interface that provides real-time visualization of diagnostic imaging results, alerts, and decision support tools for clinicians, with the interface enabling medical staff to adjust imaging parameters and thresholds for optimized diagnosis.

Description

FIELD OF THE INVENTIONThis invention relates to medical imaging and diagnostic tools, and more particularly to hybrid imaging systems that integrate ultrasound and optical imaging to improve early detection of malignant lesions.BACKGROUND OF THE INVENTIONCancer is still a major global health problem which constitutes a considerable part of morbidity and mortality. Early detection is of critical importance to increase survival chances and allow rapid treatment before cancer transitions to a later stage. Classical imaging methods such as X-ray mammography, computed tomography (CT), magnetic resonance tomography (MRI) and positron emission tomography (PET) are frequently used for cancer detection and treatment. However, these conventional methods have numerous disadvantages, such as high prices, exposure to ionizing radiation, limited sensitivity to early-stage tumors, and lack of accessibility in resource-limited environments. While ultrasound images represent a non-invasive, real-time and affordable means for soft tissue diagnosis, they do not provide functional or molecular contrast, so benign and malignant lesions cannot be distinguished solely on the basis of structural images. Optical imaging methods such as photoacoustic imaging (PAI) and diffuse optical tomography (DOT) may also contribute to a deeper understanding of vascular and metabolic changes, but these methods have a limited penetration depth and poor anatomical resolution when used independently of each other.Current imaging strategies are bordering. Therefore, there is an urgent need for a multi-modal imaging approach that associates high resolution anatomical imaging with functional and molecular data to achieve improved cancer early detection and characterization. Existing imaging methods have specific deficiencies: conventional X-ray mammography and CT scans are associated with radiation exposure to patients and have low sensitivity to dense breast tissue, leading to false negative findings and unwarranted biopsies. MRI and PET scans are effective, but costly and time consuming and are cost prohibitive for many patients. Ultrasonic imaging with a stand-alone probe provides high resolution anatomical information, but does not provide functional contrast, making tumor aggression difficult to assess. In contrast, optical imaging methods (e.g., PAI and DOT) are very sensitive to tumor vascularization and metabolic activity, but have a limited depth of image.Current cancer diagnosis is often incoherent and relies on multiple imaging procedures, which delays diagnosis, increases burden on the patient, and increases cost. Thus, patients may first undergo an ultrasound study to analyze the location of the tumor, followed by an MRI or PET study to further characterize the tumor, resulting in long wait times between tests, which may delay the onset of treatment. Blood-based biomarker assays are more sensitive and easier to obtain and indicate malignant disease. However, they are distinguished by a low spatial accuracy and generally require a further imaging method in order to localize the tumor. Oncologists also often need more than one imaging method to confirm diagnosis and make treatment decisions, increasing the number of overlapping imaging studies and increasing the decision deadlines. This incoherent method may result in inefficient cancer diagnosis and treatment planning, suggesting the need for a single system that may integrate many imaging methods into a single device.The other major challenge in cancer imaging is dependent on the manual interpretation of the images by radiologists and oncologists, leading to inter-observer discrepancies and diagnostic discrepancies. Early stage detection of tumors is a debilitary challenge and these small and subtilous anatomical abnormalities are often overlooked with conventional imaging modalities. Many imaging systems use fixed threshold based alarming mechanisms that are reactive rather than predictive. In such practice, it may happen that slowly growing malignant tumors are not alerted until the disease has already advanced, which leads to chances of premature intervention being missed. On the other hand, benign anomalies can result in false alarms, which can increase the number of subsequent tests and biopsies required. Medical KI / ML has exhibited its greatest effect in the area of diagnostic imaging; however, it has the potential to alter cancer diagnostics by improving detection accuracy, automated tumor classification processes, and improved clinical decision making. Intensive Imaging with AI Imaging analysis has the potential to learn from various types of imaging patterns, recognize early malignancy of the disease (e.g., via an image), and enable immediate risk stratification with advanced algorithms, all of which will contribute to early disease diagnosis, reduction in delays, and ultimately improvement in patient results. However, these applications of AI are still hindered by data quality, interoperability with hospital systems, and challenges for clinically taking new technologies into existing workflows.Therefore, a hybrid imaging system that combines the strengths of ultrasound and photography and simultaneously employs AI-based analyses for automatic tumor detection and classification is a potential response to the limitations of conventional cancer imaging. Hybrid ultrasound and visual imaging is an interesting solution to highly accurately detect cancer in a non-invasive and cost-effective manner, as it allows physicians to visualize tumor morphology, blood vessels, and metabolic activity in a single photo session. A hybrid imaging system that combines radio frequency ultrasound imaging with PAI and DOT can generate high resolution anatomy with functional and molecular data, thus providing detailed insights into the tumor characteristics. PAI utilizes tumor oxygenation, vascular density and metabolic activity to improve vascular variations, while DOT provides information on deep tissue irrigation and biochemical composition essential for distinguishing between malignant and benign tumors. Currently, there is no commercial solution that can combine ultrasound, CT, and direct CT into a single portable, AI-based system, which limits the extensive clinical application of hybrid imaging technology.The increasing demand for low cost, portable, and AI-assisted computerized cancer early detection has led to a significant demand for a compact, high resolution, real-time hybrid imaging system capable of integrating high resolution ultrasound (HRUS), photoacoustic imaging (PAI), and diffuse optical tomography (DOT) into one device. This system would have to be able to provide real-time multimodal imaging, AI-based tumor classification and seamless connectivity for remote diagnosis and telemedical applications. The apparatus must also be low in power consumption, portable, and capable of communicating with the existing hospital networks for use in a larger number of hospitals, ambulofens, and mobile healthcare stations.The present invention solves these problems by developing a hybrid ultrasonic and optical imaging apparatus for cancer premature detection that provides a combination of radio frequency ultrasound interaction with tissue and photoacoustic imaging (PAI) and diffuse optical tomography (DOT) over a single AI driven platform. This system involves multimodal imaging for whole body tumor assessment with simultaneous visualization of structural, functional and molecular processes. It employs AI image processing and performs in real time automatic tumor segmentation, classification and malignancy prediction. The format is intended to increase the efficiency of diagnosis, to save unnecessary imaging, to shorten patients and to optimize clinical procedures. It offers telemedicine in the cloud to allow specialists to provide patients even when they are spatially separated, and uses artificial intelligence to speed up and improve diagnosis while the patient is still on the table, and sends real-time data to electronic patient files (EHR). The portable low power system can be used in a variety of clinical environments, such as hospitals, ambulofens, and mobile screening units.This invention, which combines structural, functional and molecular imaging with AI-controlled analytics, significantly advances the state of the art in non-invasive cancer diagnostics and allows high diagnostic accuracy, low cost, and wide use in various healthcare areas. This hybrid imaging technology can break through the limits of imaging with only one modality and significantly increase the chances of cancer premature detection, shorten the clinical decision time, and eventually achieve the goal of improving patient results.SUMMARY OF THE INVENTIONThe invention encompasses a hybrid ultrasound and optical imaging system for cancer early detection that combines high resolution ultrasound imaging, photoacoustic imaging (PAI), and diffuse optical tomography (DOT). at least combined: FIR AR (structure)+MicroCT (structure)+RI (function), PET (molecule)(specific) The apparatus is intended to improve the accuracy of cancer diagnosis through structural, functional, and molecular imaging capabilities. It contains an ultrasonic imaging module that generates high-resolution images of soft tissue structures and thus enables early detection of tumor formations. It employs B-mode imaging, Doppler ultrasound, and elastography to assess the tissue stiffness and vascular properties required to distinguish between malignant and benign lesions.This integrated optical imaging module combines photoacoustic imaging (PAI) that recognizes hemoglobin concentration, oxygen supply to blood and vascular abnormalities with diffuse optical tomography (DOT) that provides metabolic and perfusion information about the tissue. The ultrasonic and optical sensor array is connected to a multi-mode image fusion and data processing unit that receives, processes and analyzes image data in substantially real time. A data storage module allows historical image records to be stored for longitudinal section studies and AI-based predictions. A real-time image processing module employs Kalman filtering and deep learning-based noise cancellation techniques to further improve the quality of the images passed to the machine learning module. Data processing module: This module is a machine learning module that is located in the data processing unit and automates tumor detection, segmentation, and classification using pre-trained AI models. It operates based on a supervised learning model and uses extensive annotated datasets of normal and tumorous tissue to improve diagnostic performance in real time.In addition, the module has an adaptive learning function that enables it to refine and improve its predictive algorithms in response to ongoing inputs from incoming imaging data and physician feedback, thereby achieving ever-higher precision in cancer detection. In addition, the device has a protocol-independent communication interface which enables uninterrupted use of existing hospital systems. The connection to mobile devices is via Bluetooth Low Energy (BLE) for local interaction, via Wi-Fi for remote analysis via the cloud, and via wired DICOM-compatible interfaces for input into hospital information systems, thus ensuring interoperability with standard medical imaging databases and telemedical applications.The user interface is optimized for interaction with the physician and has a high resolution multi-touch display that displays imaging overlays in real time, interactive diagrams, and adjustable alerts. Medical personnel can adjust diagnostic limits, review histological growth trends, and harmonize multimodal imaging representations on an intuitive dashboard tailored to the patient. The apparatus has an internal lithium ion rechargeable battery for more than 10 hours of continuous operation. It has a high efficiency sleep mode that helps to reduce effective power consumption in standby mode, allowing for longer useful life in clinical and mobile screening environments. Through the development of AI sound involving engine switching, multi-modal processing and real world processing. The overall object of this invention is to enable a multi-scale, ultrasensitive and specific optical imaging platform that can operate in conjunction with ultrasound and has the potential to detect malignancies at an early stage. With this approach involving the entire patient, health providers can diagnose tumors earlier than later if it becomes too late.It is innovative and equipped with a focus on the user experience. It provides a highly human-centered graphical surface over which clinicians can wirelessly navigate through the imaging results, alter the parameters used, and obtain AI-based diagnostic results. The system also provides interactive visualization tools and real-time alerts that aid the clinical decision process to be able to respond quickly to alerting findings. One of the primary objectives of this invention is to enable energy efficient operation while maintaining excellent diagnostic performance. Thanks to the power saving design and rechargeable lithium ion battery, the device can be operated over longer periods of time and has fewer down times, thus being suitable for a variety of healthcare facilities, from hospitals and ambulo lances to mobile screening programs. The apparatus that combines most advanced imaging modalities with sophisticated AI-based analysis tools is intended to alter cancer diagnostics to increase diagnostic accuracy, improve patient care, and allow for broad introduction of cancer premature detection.DETAILED DESCRIPTION OF THE APPARATUSReferring now to FIG. 1, a block diagram of a hybrid ultrasonic and optical imaging apparatus for detecting cancer premature detection is shown. The apparatus 400 consists of: 1. multi-modal imaging integration for improved cancer detection• Ultrasound imaging module (402): Uses radio frequency (5-15 MHz) probes to perform high resolution real time anatomical imaging of tissues and to detect tumors and structural abnormalities at early stages. The ultrasound module supports:◯ B-mode imaging to assess tissue morphology. Doppler ultrasound for assessing blood flow in and around tumors. Shear wave elastography for analysis of tissue stiffness that aids in cancer differentiation.• Optical imaging module (404): Integrated:◯ Photoacoustic Imaging (PAI): Uses pulsed laser excitation (700-1100 nm) to generate ultrasound signals from tissue chromophores that provide molecular contrast for detecting tumor angiogenesis and oxygen content. Diffuse Optical Tomography (DOT): Measures light absorption and scattering and provides insight into functional imaging, e.g. into oxygen saturation, hemoglobin concentration and metabolic activity.• Multi-modal image fusion engine (406):Real-time synchronization of ultrasound and optical imaging data, thereby creating a comprehensive multiparametric image of tissue anomalies. Overlay of Structural and Functional Imaging Data to Improve the Accuracy of Cancer Early Detection.◯ 3D volume reconstruction for better visualization of tumor boundaries and depth.The hybrid approach provides both anatomical and functional imaging and is thus substantially more effective than stand-alone ultrasound or optical imaging systems.2. Advanced Data Processing and AI-Based Cancer ClassificationThe data processing unit (408) integrates real-time image reconstruction, signal processing, and machine learning algorithms to improve diagnostic accuracy.Functional key components:• Real-time signal enhancement engine:◯ employs Kalman filtering, wavelet denoised, and adaptive background subtraction to remove image artifacts and noise.◯ improves contrast in poorly visible tissue regions with deep learning-based super-resolution techniques.• AI-assisted tumor classification (machine learning module 410)◯ Trained on Thousands of Labeled Medical Records Using Monitored Deep Learning Techniques for Tumor Detection.◯ can segment lesions, classify tissue types, and predict the risk of malignancy based on multiparametric imaging data.◯ uses transfer learning and federated learning models to continuously improve diagnostic accuracy in new clinical cases.• Module for data storage and historical analysis:◯ stores high resolution imaging data and patient metadata for longitudinal tracking.◯ allows retrospective analysis that allows physicians to monitor tumor progression and response to treatment.◯ supports predictive modeling for early stage tumor detection, thus reducing the need for unnecessary biopsies.The AI-assisted diagnostic assistance system significantly improves the sensitivity and specificity of cancer detection, reduces false positive results and increases the trust of the physician.3. Real-time Connectivity and Remote Monitoring FunctionsThe communication interface (412) is designed for seamless data transmission and integration with the IT systems of hospitals and external healthcare providers.Supported Communication Protocols:• Wi-Fi connectivity: Allows for real-time uploading of imaging data to the hospital's PACS (Picture Archiving and Communication Systems) for central data management.• Bluetooth Low Energy (BLE): Allows wireless transfer of imaging results to mobile devices or tablets for immediate review by radiologists and oncologists.• Cloud-based telemedical support: Secure cloud storage of patient images to facilitate remote AI-controlled diagnosis.◯ enables teleconultations in which specialists from remote can analyze imaging data and make clinical recommendations.• DICOM Compatibility: Ensures standardized image formatting and interoperability with existing medical image databases.• Ethernet and USB ports: Allows wired data transfer for high speed image analysis in research and laboratory environments.Real-time Alert System:• Automatic alerts notify the physician if a suspect lesion exceeds predefined threshold malignant likelihood values.• AI-controlled risk scoring system prioritizes the cases requiring urgent intervention.This real-time connectivity provides efficient integration of clinical workflows and allows rapid diagnosis and intervention in high risk cancer patients.4. User Centric Interface and Energy Efficient DesignThe user interface (414) is configured for intuitive real-time interaction with the imaging data and provides a seamless experience for radiologists and oncologists.Touch screen based interactive imaging control:• High resolution multi-touch display for easy navigation through the multi-modal imaging results.• Interactive annotation tools that allow specialists to mark interesting regions and add diagnostic annotations.• Customizable display settings that enable contrast adjustments, 3D visualization, and overlay comparisons of ultrasound and optical image data.Customizable warnings and AI-based decision aids:• Clinicians allow adjustable diagnostic thresholds to be established based on tumor size, metabolic activity, and blood flow abnormalities.• An AI-controlled system to support clinical decisions (CDSS) provides automated secondary agreements and reduces inter-observer variability in cancer diagnosis.Energy Efficient and Portable Design:• With a Powerful Lithium-Ion Battery Offering: More than 10 hours of continuous imaging operation. energy efficient energy saving mode that is automatically activated upon inactivity. fast charging function which ensures minimum downtime. Ergonomic and Portable Form Factor:• Compact, light design optimized for point-of-care facilities, mobile clinics, and hospital bed cancer care.• It can be used in hospitals, ambuches and rural healthcare facilities and brings cancer early detection into underprovision areas.This user-oriented design and portability makes the hybrid ultrasound and optical imaging apparatus an ideal solution for decentralized, AI-controlled cancer early detection programs that improve early detection rates and patient results.This hybrid ultrasonic and optical imaging apparatus has been developed to improve the accuracy of cancer premature detection. It provides high resolution, multi-modal real-time imaging with AI-based diagnostics, thus ensuring more precise, non-invasive and efficient solutions for cancer provisioning and monitoring.

Claims

An apparatus for hybrid ultrasound and optical imaging for cancer premature detection, comprising: • an ultrasound imaging module capable of capturing high resolution structural images of soft tissue for real-time visualization of suspected lesions; • an optical imaging module comprising at least one of photoacoustic imaging (PAI) and diffuse optical tomography (DOT) to acquire functional and molecular imaging data including hemoglobin concentration, oxygen saturation, and metabolic activity of tissues; • a multimodal image fusion machine configured to integrate ultrasound and optical image data to enable comprehensive, high contrast diagnostic visualization; • a computing device with an AI-based machine learning module trained on large cancer datasets for automatic tumor detection, segmentation, classification and malignancy prediction; • a communication interface for real-time data transmission supporting both wireless and wired connections to enable remote monitoring, data exchange and cloud-based AI analyses; • an interactive user interface providing real-time visualization of diagnostic imaging results, alerts and decision-assist tools to clinicians, the interface enabling medical personnel to adjust imaging parameters and thresholds for optimized diagnosis.The apparatus of claim 1, wherein the ultrasound imaging module comprises: • a radio frequency sonic head array (5-15 MHz) for anatomical imaging of soft tissue; • a Doppler ultrasound system for assessing blood flow and vascular abnormalities associated with tumors; • a shear wave elastography system for measuring tissue stiffness that helps to distinguish between benign and malignant growths.The apparatus of claim 1, wherein the optical imaging module comprises: • a pulsed laser excitation photoacoustic imaging (PAI) system (700-1100 nm) for measuring tumor associated oxygen supply and blood vessel density; • a diffuse tomographic optical system (DOT) that evaluates absorption and scattering of tissue with near infrared light to provide metabolic and blood flow data indicative of malignant disease.The apparatus of claim 1, wherein the computing device comprises: • a data storage module storing historical image data and metadata for longitudinal slice analysis and adaptive AI model training; • a real-time image processing engine employing Kalman filtering, deep learning based noise cancellation, and super resolution algorithms to improve image quality before the data is fed to the machine learning module; • a multi-mode AI-based diagnostic system incorporating ultrasound and optical data for real-time tumor classification, risk assessment, and decision assistance.The apparatus of claim 1, wherein the communication interface provides • Bluetooth Low Energy (BLE) connectivity for seamless integration with mobile devices and handhelds; • Wi-Fi connectivity for cloud-based AI diagnostics and expert remote consulting; • DICOM compatible wired connectivity that enables direct integration with hospital Picture Archiving and Communication Systems (PACS) and electronic health care records (EHRs) for standardized data exchange.The apparatus of claim 1, wherein the user interface comprises: • a high resolution multi-touch display capable of displaying image overlays in real time, interactive plots, and trend analyses; • customizable alert systems that enable healthcare providers to adjust automatic notifications for high risk findings to provide immediate intervention when the malignancy indicators exceed predefined thresholds.The apparatus of claim 1, wherein the apparatus of any: • rechargeable lithium-ion battery that allows for at least 10+ hours of continuous operation at full imaging power; • energy efficient standby mode that automatically reduces power consumption during inactive periods while maintaining standby for immediate operation.

Images