Multi-mode endoscope video transmission wire harness system and method based on high-speed SerDes chip

Abstract

The invention relates to the field of medical instruments, and discloses a multi-mode endoscope video transmission wiring harness system based on a high-speed SerDes chip, and the system comprises an endoscope front-end module which is used for collecting video data and front-end control signals, and comprises a first data collection unit which is used for collecting first physical data; the SerDes serializer module is used for serializing the video data and the first physical data into high-speed SerDes signals; and the multi-mode transmission wire harness is used for transmitting the high-speed SerDes signal, and the multi-mode transmission wire harness comprises a second data acquisition unit used for acquiring second physical data. The high-speed SerDes chip is adopted to serialize multi-channel video and sensor parallel data into a single pair or a small number of differential pairs for transmission, meanwhile, the internal structure of the wire harness is optimized, and the light weight and the flexibility of the wire harness are greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of medical devices, and in particular to a multimodal endoscope video transmission harness system and method based on a high-speed SerDes chip. Background Art

[0002] Minimally invasive surgery has become a key development in modern medicine, with endoscopic technology playing a central role. Despite continued advancements in optical imaging and instrument operation, the reliability of the data transmission harnesses that support these critical functions has become increasingly prominent, becoming a bottleneck restricting overall system performance and safety.

[0003] Existing endoscope transmission harnesses have significant limitations. They typically focus on single, high-speed video signal transmission, making them less adaptable to the physical stresses of complex operating environments. Harnesses frequently bend, twist, and stretch during surgery, and these constant mechanical stresses can easily cause fatigue damage to the internal conductor material. Over time, signal transmission quality inevitably degrades, manifesting as flickering video, artifacts, and even signal interruptions. This can have serious consequences in medical procedures requiring extremely high real-time performance.

[0004] Furthermore, existing technologies are severely lacking in their ability to monitor the health of the wiring harness itself. For example, critical physical deformation information, such as localized excessive bending or instantaneous stretching that the wiring harness may experience during operation, is often not perceived in real time. This is like walking on the edge of equipment performance degradation without an effective warning mechanism. This information blind spot makes it difficult for operators to accurately judge the stress conditions of the wiring harness, inadvertently accelerating the aging of the wiring harness material. Ultimately, the wiring harness fails due to cumulative damage long before its expected lifespan, which undoubtedly increases the operating costs of medical institutions and reduces the efficiency of equipment use. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a multi-modal endoscopic video transmission harness system and method based on a high-speed SerDes chip, which solves the technical problem that the existing endoscopic transmission harness lacks real-time perception, intelligent assessment and active warning capabilities of its own health status in complex operating environments, resulting in its reliability, service life and operational safety being difficult to guarantee.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a multimodal endoscope video transmission harness system based on a high-speed SerDes chip, comprising: an endoscope front-end module for collecting video data and front-end control signals, and comprising a first data collection unit for collecting first physical data; a SerDes serializer module for serializing the video data and the first physical data into a high-speed SerDes signal; a multimodal transmission harness for transmitting a high-speed SerDes signal, the multimodal transmission harness comprising a second data acquisition unit for acquiring second physical data; a SerDes deserializer module, configured to deserialize the high-speed SerDes signal to recover the video data and generate quality data of the SerDes signal; Host-side status monitoring module, used to collect connector status data; a harness health assessment module, configured to receive the second physical data, the SerDes signal quality data, and the connector status data, and to perform a health assessment on the multimodal transmission harness based on the second physical data, the SerDes signal quality data, and the connector status data; an operation behavior analysis module, configured to receive the second physical data and generate a real-time reminder signal according to a comparison result between the second physical data and a preset stress threshold; The user interface module is used to receive the health status assessment results and real-time reminder signals, and to display the health status assessment results and real-time reminder signals.

[0007] Preferably, both the first data acquisition unit and the second data acquisition unit include physical sensors, and the physical sensors include fiber Bragg grating sensors.

[0008] Preferably, the first physical data and the second physical data include instantaneous bending radius data and instantaneous tensile strain data.

[0009] Preferably, the SerDes signal quality data generated by the SerDes deserializer module includes bit error rate data, eye diagram margin data and jitter data.

[0010] Preferably, the host-side status monitoring module is used to collect the connector status data, and the connector status data includes connector contact resistance data and connector plug-in and plug-out times data.

[0011] Preferably, the harness health assessment module is used to perform a health status assessment on the multimodal transmission harness, wherein the health status assessment includes a prediction of the remaining service life of the harness.

[0012] Preferably, the preset stress threshold in the operation behavior analysis module includes a preset minimum safe bending radius threshold and a preset maximum safe tensile force threshold.

[0013] Preferably, the operation behavior analysis module is used to generate the real-time reminder signal when the second physical data exceeds the preset stress threshold and lasts for a preset time period.

[0014] Preferably, the user interface module is used to display the real-time reminder signal in a visual manner, and the visual manner includes displaying text prompts or color changes at the edge of the video screen.

[0015] A multi-modal endoscope video transmission harness method based on a high-speed SerDes chip includes the following steps: S1: Collect video data, front-end control signals and first physical data collected by a first data acquisition unit; S2: Serializing the video data and the first physical data into a high-speed SerDes signal; S3: Transmitting the high-speed SerDes signal through a multimodal transmission harness, wherein a second data acquisition unit acquires second physical data within the multimodal transmission harness; S4: Deserializing the high-speed SerDes signal to recover the video data, and generating quality data of the SerDes signal; S5: Collect connector status data; S6: performing a health status assessment on the multimodal transmission harness based on the second physical data, the SerDes signal quality data, and the connector status data; S7: generating a real-time reminder signal based on a comparison result of the second physical data and a preset stress threshold; S8: Displaying the health status assessment result and the real-time reminder signal.

[0016] The present invention provides a multi-modal endoscope video transmission harness system and method based on a high-speed SerDes chip. It has the following beneficial effects: 1. The present invention uses a high-speed SerDes chip to serialize multi-channel video and sensor parallel data into a single pair or a small number of differential pairs for transmission. At the same time, it optimizes the internal structure of the wiring harness, achieves a significant increase in the lightweight and flexibility of the wiring harness, and achieves the effect of reducing the complexity of the wiring harness, significantly reducing the wire diameter, and making it easier to flexibly operate in a narrow cavity.

[0017] 2. The present invention cleverly integrates micro physical sensors (such as fiber Bragg grating sensors) inside the wiring harness, monitors the SerDes signal quality and connector status, and combines algorithms to fuse and analyze multi-source data. This achieves real-time, comprehensive assessment of the wiring harness health status and accurate prediction of the remaining service life, thereby achieving wiring harness fault warnings, proactive maintenance, and significantly reducing unplanned downtime of medical equipment.

[0018] 3. The present invention monitors the instantaneous physical stress (such as bending radius and tensile strain) borne by the wiring harness in real time and compares it with the preset safety threshold, thereby generating non-invasive real-time reminder feedback to the operator, thereby achieving immediate correction of improper operating behavior, reducing human damage to the wiring harness from the source, extending the service life of the wiring harness, and assisting medical staff in standardizing operations.

[0019] 4. The present invention utilizes the data transmission capability of the high-speed SerDes chip, combines its advantages in signal integrity and anti-interference, and optimizes the wiring harness structure design to take into account mechanical performance, thereby achieving low latency and high stability of endoscopic high-definition video data over long distances (e.g., 10 meters without a relay), thus breaking through the traditional transmission distance limitations, ensuring the quality of real-time high-definition image transmission, and improving diagnostic and surgical efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a system architecture diagram of the present invention; Figure 2 Schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solution of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] Example 1: Please see the attached Figure 1 The embodiment of the present invention provides a multi-modal endoscope video transmission harness system based on a high-speed SerDes chip, comprising: The endoscope front-end module is used to collect video data and front-end control signals, and includes a first data acquisition unit for collecting first physical data; the first physical data and the second physical data include instantaneous bending radius data and instantaneous tensile strain data.

[0023] Specifically, the endoscope front-end module is the starting point of the system. Its primary function is to acquire high-quality video data and necessary front-end control signals. Furthermore, this module crucially integrates the first data acquisition unit. The design of this unit is crucial for real-time monitoring of the physical stress on the endoscope tip during operation.

[0024] The core of the first data acquisition unit lies in its fiber Bragg grating (FBG) sensor array. Specifically, multiple miniature FBG sensors are strategically embedded along the outer wall or internal structure of the tip and proximal end of the endoscope. These FBG sensors are distributed in an array, for example, at different locations on the tip (such as top, bottom, left, and right) and at multiple points along its length. When the tip of the endoscope is bent, twisted, or stretched within the body, the optical fiber where the FBG sensors are located will experience tiny strains. This strain causes the periodic structure of the fiber Bragg grating to change, resulting in a corresponding shift in the central wavelength of the reflected light.

[0025] To accurately determine these wavelength shifts, the first data acquisition unit is equipped with a compact optical spectrum interrogator. This interrogator typically uses a broadband light source (such as an LED or superluminescent diode) and a high-resolution spectrometer, or employs a tunable laser to scan the FBG's reflectance spectrum peak. By monitoring and analyzing the spectral signal returned from the FBG sensor in real time, the interrogator accurately captures the central wavelength shift of each FBG sensor. This raw wavelength drift data is then fed into a built-in microcontroller or dedicated signal processing unit. This unit then runs a pre-established strain-wavelength calibration model and geometry solution algorithm. This model, calibrated experimentally, correlates specific wavelength shifts with the precise strain experienced by the fiber.

[0026] Based on the spatial distribution of multiple FBG sensors and their respective strain data, a geometric solution algorithm can calculate the instantaneous bending radius and tensile strain of the endoscope tip in real time. For example, by measuring the wavelength variation of at least three non-collinear FBG sensors within a bending plane, the bending radius of that section can be accurately inferred. FBG sensors distributed along the axial direction are used to detect tensile or compressive strain.

[0027] Furthermore, to cope with complex surgical environments, the first data acquisition unit can integrate micro-temperature sensors, such as high-precision thermistors or thin-film thermocouples, within or immediately adjacent to the front-end module, as needed. These temperature sensors monitor the operating temperature of the endoscope's front end in real time, providing the system with environmental parameters that help correct for the temperature cross-sensitivity of the FBG sensors, ensure the accuracy of physical data acquisition, and provide a basis for system overheating protection. All collected video data, front-end control signals, and processed primary physical data (i.e., instantaneous bending radius and tensile strain) are transmitted via an internal interface to the SerDes serializer module for high-speed serialization.

[0028] A SerDes serializer module, the SerDes serializer module is communicatively connected to the endoscope front-end module and is used to serialize the video data and the first physical data into a high-speed SerDes signal; Specifically, the core component of a SerDes serializer module is a high-performance SerDes serializer chip. In practical applications, the appropriate chip model can be selected based on the required video resolution, frame rate, and sensor data bandwidth. For example, industrial-grade or medical-grade SerDes chips such as TI's DS90UB9x series or Analog Devices' ADV76xx series can be used. These chips typically feature multiple parallel input interfaces and one or more high-speed serial output interfaces. Key parameters of the serializer chip include the supported transmission rate (for example, the ability to handle data streams up to 10Gbps or higher to accommodate 4K or even 8K ultra-high-definition video transmission requirements), the number of channels (typically referring to the number of parallel input data bits, such as 10, 12, or 24 bits), and the packaging format (a compact package is required to fit within the integration space at the front end of the endoscope harness).

[0029] The workflow of the serializer module is as follows: First, it receives video data from the endoscope's front-end module. This video data is typically transmitted as parallel digital signals, such as pixel data in RGB or YUV format. Simultaneously, it also receives first physical data: instantaneous bending radius data and instantaneous tensile strain data generated by the front-end's first data acquisition unit. This physical data is typically digitized low-speed control or status information.

[0030] The serializer chip integrates a data multiplexer and a high-speed encoder. The data multiplexer is responsible for time-multiplexing the parallel video data and the first physical data, integrating them into a single parallel data bus. The high-speed encoder then encodes the multiplexed data (for example, using encoding schemes such as 8b / 10b or 10b / 12b) to increase the signal's DC balance, reduce inter-symbol interference, and embed clock information. The encoded data is converted into a differential signal and transmitted through the SerDes output interface. This serialization process effectively converts multiple independent parallel signals into a high-speed serial data stream on one or a small number of differential pairs, significantly reducing the number of physical wires required in the wiring harness.

[0031] In addition to basic data serialization, some advanced SerDes chips also offer preliminary signal integrity monitoring capabilities. While the primary signal quality analysis occurs at the deserializer, the serializer can perform pre-emphasis or de-emphasis processing before data transmission to compensate for high-frequency losses during cable transmission, which helps improve overall signal quality. Additionally, some serializer chips may offer a simple line loopback test for preliminary verification of link connectivity.

[0032] Through the effective operation of the SerDes serializer module, synchronous, efficient, and high-bandwidth transmission of video data and physical sensor data is achieved, laying the foundation for subsequent long-distance, low-latency transmission.

[0033] A multimodal transmission harness is communicatively connected to the SerDes serializer module for transmitting high-speed SerDes signals. The multimodal transmission harness includes a second data acquisition unit for acquiring second physical data. The first data acquisition unit and the second data acquisition unit both include physical sensors, and the physical sensors include fiber Bragg grating sensors.

[0034] Specifically, the fabrication and integration of a multimodal transmission harness is a core step in building this endoscope system. In practice, this harness connects to the SerDes serializer module and implements high-speed SerDes signal transmission. It also contains a crucial component—the second data acquisition unit. Notably, both the "first data acquisition unit" at the front end of the endoscope and the "second data acquisition unit" within the harness itself use fiber Bragg grating (FBG) sensors as the primary physical sensing element.

[0035] The specific implementation of this wiring harness utilizes a hybrid cable solution. The first step is the deployment of high-speed differential copper cable pairs, using high-purity oxygen-free copper conductors. Typically, wire sizes range from AWG30 to AWG34. These wires are twisted in pairs to minimize common-mode noise during transmission and ensure signal purity. Each differential pair is insulated with a low-dielectric-constant material, such as expanded FEP or Teflon, to reduce signal attenuation and crosstalk, allowing high-speed data to travel farther and more stably.

[0036] Next comes the integration of embedded optical fibers, which constitute the second data acquisition unit. During actual harness manufacturing, multiple thin optical fibers are embedded within the harness, either along its longitudinal axis or in a spiral pattern. These fibers are not ordinary optical fibers; they are pre-programmed with multiple fiber Bragg grating (FBG) sensors before leaving the factory. The placement of these FBG sensors is carefully planned, evenly distributed along the length of the harness, typically spaced every 5 to 10 mm. This creates a continuous distributed sensor array. When the harness bends or stretches during use, the optical fibers housing these FBG sensors deform, causing the central wavelength of their reflected light to shift. This wavelength shift is monitored in real time by an accompanying demodulation device. Based on a pre-set calibration model, the instantaneous bend radius and tensile strain at different locations in the harness are accurately calculated. This real-time, high-precision acquisition of harness deformation data provides direct insight into the harness's "health," something that traditional harnesses cannot achieve.

[0037] In addition to the core components mentioned above, the wiring harness also integrates power cables and auxiliary data cables. The specifications and materials of these cables have also been optimized to ensure that they can provide power and transmit low-speed control signals without affecting the flexibility and overall performance of the entire wiring harness.

[0038] In terms of shielding design, a multi-layer shielding structure is adopted to achieve excellent EMI suppression capabilities. Specifically: each pair of high-speed differential copper cables will have its own independent shielding layer, usually aluminum foil or copper braided shielding layer, to reduce crosstalk between them. Then, all internal cables and optical fibers will be wrapped with a high-density copper braided shielding layer, which is the first line of external protection. On the outermost layer, another layer of aluminum foil or special conductive polymer will be added as the final electromagnetic barrier. All these shielding layers will be effectively grounded through low-impedance paths, so that external electromagnetic interference signals can be effectively guided to the ground end, thereby significantly reducing the impact on the SerDes signal and ensuring the clarity and stability of the transmission.

[0039] Finally, the choice of cable harness outer sheath material is crucial. Medical-grade polymers, such as medical-grade thermoplastic polyurethane (TPU) or polyetheretherketone (PEEK), are used. These materials not only offer excellent biocompatibility but, more importantly, are abrasion- and chemical-resistant (able to withstand repeated sterilization), and maintain high flexibility and resilience even after repeated bending. Thanks to these materials and the precise internal structure, the entire cable harness diameter can be kept to a strict limit of less than 3 mm. Compared to traditional solutions, this significantly reduces the size and weight of the cable harness, significantly improving the convenience and flexibility of endoscope operation in confined spaces.

[0040] The SerDes deserializer module is connected to the multimodal transmission harness for deserializing high-speed SerDes signals to recover video data and generate quality data of the SerDes signals. The SerDes signal quality data generated by the SerDes deserializer module includes bit error rate data, eye diagram margin data, and jitter data.

[0041] Specifically, when building this endoscope system, the integration of the SerDes (SerDes) deserializer module is a core component of the receiver. In practice, this module connects to the multimodal transmission harness. Its primary tasks are: first, deserializing the received high-speed SerDes signal to recover the original video data; second, simultaneously generating critical SerDes signal quality data. This data, including bit error rate, eye diagram margin, and jitter, is crucial for evaluating transmission link performance.

[0042] When deploying a SerDes deserializer module, its core is a high-performance SerDes deserializer chip. In actual projects, a model that matches the front-end SerDes serializer chip is typically selected, such as TI's DS90UB9x series or Analog Devices' ADV76xx series, to ensure optimal compatibility and performance. This chip is designed to directly receive differential high-speed SerDes signals from multimodal transmission harnesses.

[0043] The chip integrates several key functional units. The first is the clock data recovery (CDR) circuit. Because SerDes signals don't have a dedicated clock line during transmission, the CDR circuit's task is to accurately extract clock information from the received serial data stream. It then uses this recovered clock to synchronize the data stream, ensuring that every bit is sampled correctly within the correct clock cycle.

[0044] Following closely behind is the adaptive equalizer. This equalizer operates dynamically. Factors such as the length, material properties, and connectors of the transmission cable bundle can cause high-speed signal attenuation and inter-symbol interference during transmission, blurring the signal waveform. The adaptive equalizer analyzes the received signal in real time and dynamically adjusts its equalization parameters to compensate for these transmission impairments. Even if signal quality degrades after long-distance transmission, the equalizer can restore the original waveform to the greatest extent possible, ensuring accurate reproduction of the HD video data. This compensation mechanism is a key technical guarantee for achieving long-distance, low-latency HD video transmission (for example, end-to-end latency can be controlled within 50 milliseconds).

[0045] In addition to the core deserialization and data recovery functions, the SerDes deserializer module also has a powerful signal quality monitoring function. This is not a simple bypass detection, but a real-time internal chip: The deserializer chip has an internal bit error rate tester. It continuously monitors the deserialized data stream, identifies the number of error bits through an internal algorithm, and calculates the bit error rate in real time. This bit error rate value is usually expressed as a negative power of 10 (for example, a typical value is 10 -12 ), which directly reflects the reliability and anti-interference ability of signal transmission.

[0046] The deserializer module internally generates and analyzes the received signal's "eye diagram." An eye diagram is a visual representation of high-speed digital signal quality. The module analyzes multiple eye diagram parameters in real time, generating key metrics such as "eye height" (reflecting the signal's noise margin) and "eye width" (reflecting the signal's jitter margin). A wider eye diagram opening indicates better signal quality and greater tolerance to noise and jitter. These metrics provide a quantitative indicator of the signal link's margin.

[0047] Jitter, the uncertainty of high-speed digital signals on the time axis, is a significant factor affecting signal integrity. The deserializer module analyzes and quantifies the jitter components of the received signal in real time, identifying types such as random jitter (RJ) and deterministic jitter (DJ). Excessive jitter can directly lead to increased bit error rates. The jitter data generated by the module provides an important basis for evaluating and optimizing the stability and reliability of the transmission link.

[0048] This signal quality data, generated in real time by the SerDes deserializer module, is transmitted to the subsequent cable health assessment module via an internal interface. This critical diagnostic information enables the system to continuously monitor the health of the transmission link and provide early warning of potential transmission performance issues, thereby ensuring video image clarity and overall system stability.

[0049] The host-side status monitoring module is used to collect connector status data; the host-side status monitoring module is used to collect connector status data, and the connector status data includes connector contact resistance data and connector plug-in and plug-out times data.

[0050] Specifically, when building this endoscope system, deploying a host-side status monitoring module is an essential component. In practice, this module focuses on collecting connector status data, with the core purpose of assessing the health of the connection point between the multimodal transmission harness and the host device. This critical connector status data includes connector contact resistance and connector plug-in and plug-out cycles.

[0051] When implementing the host-side status monitoring module, its design is not a simple data listing, but focuses on high precision and real-time performance.

[0052] First, connector contact resistance data is collected using the industry-recognized four-wire Kelvin measurement method. This method eliminates the influence of lead resistance on the measurement results, thereby obtaining the true, minute contact resistance value between connector pins. In practice, critical pins are periodically or continuously measured using independent current loops (injecting a precise constant current) and voltage measurement loops (directly measuring the voltage drop across the connector pins) while the connector is mated. For example, connector pins for high-speed SerDes signal paths, power paths, and critical control signal paths can be measured. By continuously monitoring minute changes in contact resistance, potential connector issues such as poor contact, oxidation, wear, or fretting corrosion can be detected early. An abnormal increase in contact resistance is a key indicator of connector performance degradation, which can ultimately lead to signal interruption or performance degradation. This precise measurement method provides a quantitative basis for the stability of the connector's electrical connection.

[0053] Secondly, a dedicated plug-in and unplug-out counting mechanism is integrated for collecting data on connector plug-in and unplugging times. There are multiple options for specific implementation methods, and they can be optimized according to actual needs: Mechanically triggered counting scheme: A common implementation is to trigger a counting signal using a microswitch or mechanical lever when the connector is fully inserted or removed. For example, a precision mechanical contact can be set on the connector housing or fixing mechanism. When the harness connector is fully inserted into the host connector, the contact closes or opens, thereby driving a digital counter.

[0054] Optical counting: A more advanced solution uses photoelectric beamforming or photoelectric reflective sensors. A pair of infrared photoelectric beamforming tubes or a photoelectric reflective sensor are placed along the connector insertion and removal path. When the connector is inserted or removed, it blocks or reflects the light beam, triggering a sensor signal that is then accumulated by the counter. This non-contact counting method reduces mechanical wear and improves the counter's lifespan and reliability.

[0055] Counting based on electrical contact state changes: Plug-in and unplug events can also be detected by monitoring the on / off state or impedance change of a specific pin inside the connector. For example, a low-speed detection pin can be set up. When the connector is fully inserted, this pin establishes a stable connection to ground or power, generating a clear voltage level transition that triggers counting. When the connector is unplugged, the opposite voltage level transition occurs.

[0056] Intelligent counting with identification chips: For more complex connectors, if an identification chip (such as an EEPROM or microcontroller) is integrated, specific identification information can be read or written via bus communication each time the connector is plugged in, and the plugging and unplugging event can be recorded on the host. This method can provide richer connector information, even including connection time.

[0057] Regardless of the counting method used, the module's internal counter accurately records the total number of times the connector has been inserted from the harness end to the host end. It's generally accepted in the industry that a connector's mechanical life is directly correlated to the number of plugging and unplugging cycles. By monitoring this data in real time, the system can accurately determine whether a connector is approaching its design life. Once the number of plugging and unplugging cycles reaches a preset threshold, the system issues an early warning of potential faults caused by connector wear, such as unstable signal, poor contact, or even complete failure.

[0058] The host-side condition monitoring module transmits this real-time contact resistance and plug-in / plug-out count data to the harness health assessment module via an internal digital interface. This data provides crucial supplemental information for predictive maintenance of the harness as a whole. Combined with physical stress data from the harness itself and signal quality data from the SerDes deserializer module, it provides a comprehensive, multi-dimensional health assessment of the endoscope harness, significantly improving system reliability and safety.

[0059] The harness health assessment module is communicatively connected to the second data acquisition unit, the SerDes deserializer module and the host-side status monitoring module, and is used to receive the second physical data, the SerDes signal quality data and the connector status data, and to perform a health status assessment on the multimodal transmission harness based on the second physical data, the SerDes signal quality data and the connector status data; the harness health assessment module is used to perform a health status assessment on the multimodal transmission harness, and the health status assessment includes a prediction of the remaining service life of the harness.

[0060] Specifically, the harness health assessment module is a core component for predictive maintenance. It closely communicates with multiple key units: a second data acquisition unit responsible for collecting harness deformation, a SerDes deserializer module providing link performance indicators, and a host-side status monitoring module monitoring interface conditions. Its core function is to receive and integrate various data from these units—namely, second physical data, SerDes signal quality data, and connector status data—and perform a comprehensive health assessment of the multimodal transmission harness based on this data. This assessment process is particularly critical, as it includes a prediction of the harness's remaining service life.

[0061] When deploying the harness health assessment module, its calculation and operation logic is implemented as follows: This module typically consists of a high-performance microcontroller (MCU) or an embedded processor, such as an ARM Cortex-M series MCU or Cortex-A series processor. It needs to be equipped with sufficient memory (RAM and non-volatile storage such as NAND Flash or EEPROM) to store evaluation models, historical data, and configuration parameters.

[0062] The module uses a dedicated communication interface (such as SPI, I 2 C, UART or higher-speed Ethernet / PCIe, etc.) to connect to various data sources.

[0063] Receiving Secondary Physical Data: This module receives real-time fiber Bragg grating (FBG) sensor data from the secondary data acquisition unit. This data has been demodulated and represents the instantaneous bend radius and tensile strain at different locations on the cable harness. The module performs preliminary timestamps and formatting on this data to ensure data consistency.

[0064] Receive SerDes signal quality data: Receive periodic or event-triggered SerDes signal quality data from the SerDes deserializer module, including real-time bit error rate (BER) data, eye diagram margin data (such as eye height and eye width), and jitter data (such as random jitter RJ and deterministic jitter DJ). This data is digitized and the module will verify and store it.

[0065] Receive connector status data: Receive connector contact resistance data and connector plug-in / plug-out count data from the host-side status monitoring module. Contact resistance data is typically sampled periodically, while plug-in / plug-out counts are cumulative. The module records this data and monitors its changing trends.

[0066] To train the model, we first need to construct a set of representative features that can reflect the degradation of the harness's health. These features include not only instantaneous values but also historical accumulations and changing trends.

[0067] Feature Set : : Instantaneous maximum strain ; : Instantaneous minimum bending radius ; :Instantaneous bit error rate ; : Momentary high eyes ; :Instantaneous eye width ; :Instantaneous connector contact resistance ; :Cumulative plug-in and pull-out times ; :Recently (e.g. within the past hour) Exceeding a preset threshold (e.g. ) cumulative duration; :Recent (e.g., within the past 24 hours) Less than a preset threshold (e.g. ) cumulative duration; :Recently (e.g. within the past 10 minutes) average growth rate; :Recently (e.g. within the past hour) The average rate of decline; : Recent (e.g., the past 100 plug-ins and unplugs) average growth rate.

[0068] The calculation of these features requires the module to maintain a historical data buffer and use support vector regression as the core health assessment algorithm.

[0069] Dataset: Collect a large amount of aging test data of wiring harnesses under different working conditions (such as bending, stretching, temperature cycling, plugging and unplugging, etc.). During the experiment, the above feature sets are continuously recorded. The data is collected and the actual health status of the wiring harness is recorded simultaneously (for example, the critical point where the wiring harness performance begins to deteriorate is determined through physical inspection, signal integrity testing, etc., and a health index is defined).

[0070] Label definition: The health status of the harness is defined as the health index (HI), which is a continuous value between 0 and 100. HI=100 indicates a brand new state, and HI=0 indicates a complete failure. , high-minded UI, or contact resistance> ) to determine the failure point.

[0071] SVR model training: using the collected feature set As input, HI as output, train the SVR model. The goal of SVR is to find a function , so that for all training samples, the feature To the function value The deviation does not exceed a preset threshold (error tolerance), while minimizing the complexity of the model. The radial basis function (RBF) kernel is usually used because of its strong nonlinear mapping ability.

[0072] Kernel function: ; SVR model: in, is the Lagrange multiplier, is the bias term. Model parameters (such as C value, Value and value) is optimized through cross-validation.

[0073] Model running stage (online): Real-time feature calculation: When the module is running, it will calculate and update the feature set in real time based on the current real-time sensor data and historical buffer data received All eigenvalues of .

[0074] Health index calculation: the current calculated eigenvector Input into the pre-trained SVR model and output the current health index This calculation is performed periodically, for example, updated every 1 second.

[0075] In the health index Based on the above, the Gaussian process regression (GPR) model is used to predict the remaining useful life of the wiring harness. GPR can provide the predicted mean and confidence interval, thus enhancing the reliability of the prediction.

[0076] RUL prediction model training phase (offline): Dataset: Collects health index decay trajectory data of multiple wiring harnesses from new to failure.

[0077] GPR model training: time Or cumulative number of operations As input, the health index As output, a GPR model is trained. GPR does not regress data points, but rather regresses a function distribution over them. It learns the trend and uncertainty of the health index over time and over the number of operations. The model estimates the mean and covariance functions (using a kernel function, such as the RBF kernel), which describe the evolution of the health index.

[0078] RUL prediction operation phase (online): Historical trajectory input: The module takes the recent health index history trajectory of the current harness (for example, the past 24 hours) value sequence) as input and fed into the pre-trained GPR model.

[0079] Predicting future trends: The GPR model will use this historical trajectory and its learned health index decline pattern to predict the future The mean and variance of the values.

[0080] RUL calculation: setting a failure threshold (For example, ). The module will calculate the predicted When the mean is first reached or fallen below The difference between this time point and the current time is the remaining useful life (RUL). GPR also provides a confidence interval for the prediction, for example, "RUL is 100 hours to 20 hours," providing more comprehensive information for user decision-making. Predictions are typically made every 5 minutes.

[0081] When the calculated health index falls below a preset “warning threshold” (for example), the module generates a level one warning signal.

[0082] When the RUL prediction value falls below a preset “repair threshold” (e.g., “50 hours remaining” or “10 surgeries remaining”), the module generates a secondary warning signal.

[0083] When the health index falls below a preset “danger threshold” (for example), the module generates a three-level warning signal and may recommend immediate cessation of use.

[0084] When the predicted RUL falls below a preset threshold, or the harness health index reaches a warning level, the module immediately issues a warning signal. These warnings can be output in various forms, such as being sent to a host computer via a network interface for interactive display, or triggering indicator lights and buzzers. This allows operators or maintenance personnel to take preventive measures, such as arranging harness replacement, before a harness failure actually occurs, avoiding unplanned interruptions to critical medical operations. This significantly improves equipment reliability, safety, and maintainability, and optimizes spare parts management.

[0085] An operational behavior analysis module, communicatively connected to the harness health assessment module, is configured to receive the second physical data and generate a real-time alert signal based on a comparison result between the second physical data and a preset stress threshold. The preset stress thresholds in the operational behavior analysis module include a preset minimum safe bending radius threshold and a preset maximum safe tensile force threshold. The operational behavior analysis module is configured to generate a real-time alert signal when the second physical data exceeds the preset stress threshold for a preset duration.

[0086] Specifically, it is usually implemented by a standalone microcontroller (MCU) or as a dedicated processing thread or submodule within the harness health assessment module. It needs to have a processing speed fast enough to process input data in real time and be able to quickly respond to generate warning signals.

[0087] The operation behavior analysis module will use the communication interface (such as SPI or I 2 C) Continuously receives secondary physical data from the harness health assessment module. This data is updated in real time and represents the instantaneous bend radius and tensile strain at each point in the harness. The module performs preliminary low-pass filtering or moving average processing on this data to eliminate instantaneous noise or measurement fluctuations, ensuring the stability of subsequent judgments.

[0088] The module stores a series of preset stress thresholds. These thresholds are determined based on the harness's design specifications, material properties, fatigue life test results, and clinical experience. These thresholds are typically written to non-volatile memory at the factory, but can be calibrated or updated via a host computer when needed.

[0089] Preset minimum safe bend radius threshold: This is a critical parameter. For example, if the minimum bend radius allowed by the wiring harness design is R_min, then the threshold may be set to R_min or slightly above it. If the local bend radius of the wiring harness is less than this threshold, it means that the wiring harness is experiencing bending stress beyond the design range. Long-term exposure to this stress can lead to fatigue damage or even internal fracture.

[0090] The preset maximum safe tensile force threshold is another important parameter. It corresponds to the maximum tensile strain or equivalent tensile force that the cable harness can withstand. If the tensile strain on the cable harness exceeds this threshold, it indicates excessive strain or jamming, which can accelerate stress fatigue in the conductors within the harness and even cause breakage of optical fibers or copper wires.

[0091] The module will compare each received bending radius data with the preset minimum safe bending radius threshold in real time, and compare each tensile strain data with the preset maximum safe tensile force threshold.

[0092] More importantly, to avoid false alarms, the module introduces a "continuous preset duration" judgment mechanism. This means that even if the data at a certain moment exceeds the threshold, the module will not immediately alarm. Only when a certain stress data (whether the bending radius is too small or the tensile strain is too large) continues to exceed its corresponding preset threshold, and this duration reaches a specific preset duration (for example, it can be set to 50 milliseconds, 100 milliseconds or longer), the module will generate and issue a real-time reminder signal. This "preset duration" is a configurable parameter, the purpose of which is to filter out instantaneous, short-term stress peaks and only respond to persistent, potentially harmful abnormal operating behaviors.

[0093] Once the judgment condition is met (i.e., stress data exceeds the threshold and persists for a preset period of time), the operation behavior analysis module will immediately generate a real-time alert signal. This signal can be output in a variety of ways: Visual prompts: Send signals to the host computer to drive warning icons, pop-up windows on the display, or change the color of the user interface.

[0094] Auditory prompt: drives the buzzer or speaker to emit an alarm tone or voice prompt.

[0095] Haptic prompts: If supported by the system, the controller can even be driven to vibrate for feedback. These prompts are immediately fed back to the operating physician, guiding them to adjust their technique, such as "Do not bend the harness excessively" or "Pay attention to whether the harness is stuck."

[0096] Through this real-time, intelligent stress monitoring and early warning mechanism, the operation behavior analysis module achieves instant feedback on the doctor's operation behavior, effectively avoiding instantaneous or cumulative damage to the wiring harness caused by improper operation (such as sharp bending and excessive stretching), thereby significantly extending the actual service life of the multimodal transmission wiring harness, reducing equipment maintenance costs, and further improving the safety of medical operations and the reliability of equipment.

[0097] The user interface module is in communication with the harness health assessment module and the operational behavior analysis module, and is configured to receive and display health status assessment results and real-time alert signals. The user interface module is configured to visually display the real-time alert signals, including textual prompts or color changes on the edge of the video screen.

[0098] Example 2: Please see the attached Figure 2 A multi-modal endoscope video transmission harness method based on a high-speed SerDes chip includes the following steps: S1: Collect video data, front-end control signals and first physical data collected by a first data acquisition unit; S2: Serialize the video data and the first physical data into a high-speed SerDes signal; S3: Transmitting a high-speed SerDes signal through a multimodal transmission harness, wherein a second data acquisition unit acquires second physical data within the multimodal transmission harness; S4: deserializes the high-speed SerDes signal to recover the video data and generates quality data of the SerDes signal; S5: Collect connector status data; S6: performing a health status assessment on the multimodal transmission harness based on the second physical data, the SerDes signal quality data, and the connector status data; S7: generating a real-time reminder signal based on a comparison result of the second physical data and a preset stress threshold; S8: Display health status assessment results and real-time reminder signals.

[0099] When building a high-performance, highly reliable endoscopy system, a multimodal video transmission harness based on a high-speed SerDes chip was proposed. The core of this solution lies not only in its stable transmission of high-definition video but, more importantly, in its real-time and comprehensive "sensing" of the harness's health and providing early warning of potential risks, significantly improving both the safety of medical procedures and the durability of the equipment.

[0100] The first data acquisition unit is deployed here, where high-definition video signals are captured. Usually, high-resolution CMOS image sensors such as Sony IMX258 (13MP) or ON Semiconductor AR0821 (8MP) are used. They directly capture images of the lesion area and then process them into 4K UHD (3840x2160) @ 60fps digital video streams through the front-end image processor (such as ON Semiconductor AP1302). At the same time, some necessary front-end control signals, such as PWM commands for LED lighting and I / O for the focus motor, are also processed. 2 C or SPI control signals, as well as the GPIO control signals of the biopsy forceps, are also generated and summarized by the front-end STM32F4 series microcontroller. Even more cleverly, this unit also embeds a miniature fiber Bragg grating (FBG) sensor array, usually implanted every 5 mm in the bendable part of the distal end of the endoscope, for a total of 5 to 10. They can sense the instantaneous bending angle and local deformation strain of the endoscope tip in real time at a frequency of 200 Hz. These data are obtained through a customized miniature fiber demodulation module.

[0101] Next, within the endoscope's front-end module, adjacent to the data acquisition unit, a SerDes (SerDes) serializer module is integrated. It acts as an efficient "data packer," integrating all the information collected by the front-end. Here, 4K video data (up to 12 Gbps uncompressed), front-end control signals (approximately 10 Mbps), and the first physical data (approximately 1 Mbps) are time-division multiplexed and formatted. A high-performance SerDes serializer chip, such as the Texas Instruments DS90UB949 or Analog Devices' ADV7480, then efficiently serializes these parallel data streams (via MIPI CSI-2 or LVDS interfaces) into single or dual high-speed differential SerDes signals. The FPD-Link III protocol is typically used, with a single link transmission rate of 8 Gbps to 12.5 Gbps across two differential pairs. This bandwidth is sufficient to stably carry 4K@60 fps video and all auxiliary data. The transmission protocol itself includes an embedded clock, eliminating the need for additional clock lines.

[0102] These high-speed SerDes signals, along with the sensor data from the harness itself, are transmitted via a purpose-built multimodal transmission harness. This harness, the physical lifeblood of the entire system, consists of two or four high-purity oxygen-free copper differential pairs in sizes ranging from AWG30 to AWG34. Each pair is shielded with an individual aluminum foil or copper braid. These pairs are specifically designed to carry these high-speed SerDes signals, ensuring signal integrity even over harness lengths of 5 to 15 meters. The harness's multimodal nature is achieved by embedding two to four thin optical fibers within the same harness. Each fiber is pre-programmed with 200 to 500 fiber Bragg grating (FBG) sensors, evenly spaced every 10 mm along the length of the harness. These FBG sensors monitor the instantaneous bend radius (ranging from 2 mm to infinity) and tensile strain (ranging from -1000 microstrain to +5000 microstrain) of the entire harness in real time at a 100Hz frequency. When the harness is bent or stretched, the center wavelength of the FBG sensor drifts. This wavelength data is transmitted to the host-side fiber optic demodulator (such as the Luna Innovations ODiSI6000) and converted into specific deformation data, which is the second physical data. To ensure the reliability of the harness in complex electromagnetic environments, a multi-layer shielding structure is also adopted: each pair of differential lines is independently shielded, plus an overall copper braided shield with a high coverage rate (over 95%), and the outermost layer is covered with a conductive polymer shield. The outer sheath of the harness is made of medical-grade TPU material, and the diameter is strictly controlled within 2.8mm. This material can ensure that the harness has a bending life of more than 200,000 times and a tensile strength of more than 50N, fully meeting the stringent requirements of medical devices.

[0103] When the signal arrives at the host end, the SerDes deserializer module starts working. Usually, a deserializer chip that matches the front-end serializer is selected, such as Texas Instruments' DS90UB940 or Analog Devices' ADV7680. It directly receives the high-speed SerDes signal from the harness, accurately extracts the clock through the internal clock data recovery (CDR) circuit, and then uses a powerful adaptive equalizer to compensate for the signal attenuation and distortion generated during the harness transmission process in real time. The compensation gain can reach more than 30dB, ensuring the recovery of a clear parallel video data stream (such as MIPICSI-2 or HDMI format). More importantly, the deserializer chip also has its own signal quality monitoring function. It will output SerDes signal quality data in real time, including the bit error rate (BER), which is required to be less than 10 under normal operation. -12 Once it rises to 10 -9 It is considered abnormal; there are also eye diagram margin data, such as eye height should be greater than 70%UI, eye width should be greater than 60UI; and jitter data, total jitter (TJ) should be less than 0.3UI. These data are measured at a frequency of 10Hz through I 2 C or SPI interface to transmit to subsequent modules.

[0104] A host-side status monitoring module has also been deployed at the interface between the host and the multimodal transmission harness. This module is specifically responsible for collecting connector status data. Using the four-wire Kelvin measurement method, 10mA current is periodically injected into key pins at 500ms intervals to measure contact resistance. Normally, it should be below 20mΩ; a sustained resistance exceeding 50mΩ is considered abnormal. This accurately determines whether the connector is experiencing poor contact or oxidation. Furthermore, a microswitch or optical sensor is integrated into the connector's mating mechanism, triggering a count signal each time the harness is inserted or removed. A non-volatile counter within the module accumulates the number of insertions and removals, for example, recording every 5,000 insertions, up to a maximum of 100,000 insertions and removals. Because connectors are typically designed for a lifespan of 5,000 to 10,000 insertions and removals, this data directly reflects connector wear.

[0105] Next, all this data is aggregated into the cable harness health assessment module, typically comprised of a high-performance microcontroller or embedded processor. This module receives secondary physical data (100Hz) from the secondary data acquisition unit, signal quality data (10Hz) from the SerDes deserializer module, and connector status data (2Hz) from the host-side status monitoring module. It first synchronizes and normalizes this heterogeneous data source, then performs refined feature extraction. For example, it calculates the cumulative duration of bends less than 5mm in radius, the peak and duration of tensile strain exceeding 2000 microstrain, the BER growth rate within 1 minute, the percentage decrease in eye margin within 1 hour, and the average increase in contact resistance over 100 mating cycles. These features are then fed into a pre-trained machine learning model, which analyzes historical failure patterns to calculate a cable harness health index (0-100) in real time. For example, a health index below 30 is considered sub-healthy. Furthermore, the module uses Gaussian process regression or similarity prediction algorithms to predict the remaining useful life (RUL) of the wiring harness based on the decline trend of the health index, for example, "200 hours remaining" or "50 complete endoscopic procedures remaining." When the RUL falls below a preset threshold, the module immediately issues an alert.

[0106] Finally, there's a key operational behavior analysis module. Independently or as a subfunction of the health assessment module, it continuously receives secondary physical data (instantaneous bending radius and instantaneous tensile strain) at a frequency of 100Hz. This module stores preset stress thresholds: the minimum safe bending radius threshold is 3.0mm, and the maximum safe tensile force threshold is 40 Newtons (or the equivalent of 3000 microstrain). Its core logic is this: if the local bending radius of the harness is less than 3.0mm, or the tensile strain is greater than 3000 microstrain, and this abnormal state persists for more than 100 milliseconds, the module immediately generates a real-time alert signal. This "duration" judgment mechanism avoids instantaneous false alarms.

[0107] All of these health status assessment results and real-time warning signals are ultimately presented intuitively on the host's display interface. For example, there will be a dedicated "harness health" dashboard that displays the current health index as a percentage (such as 85% health) or level (excellent / good / warning / dangerous) and provides a prediction of the remaining life (such as "200 hours remaining"). At the same time, when the operational behavior analysis module triggers a real-time reminder, a red warning box (such as "Harness Excessively Bent!") will pop up on the display, accompanied by a flashing icon and a clear alarm tone (440Hz intermittent beeping). In extreme cases, the system can even temporarily restrict certain operational functions until the harness posture returns to safety.

[0108] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A multi-modal endoscope video transmission harness system based on a high-speed SerDes chip, characterized by: include: an endoscope front-end module for collecting video data and front-end control signals, and comprising a first data collection unit for collecting first physical data; a SerDes serializer module for serializing the video data and the first physical data into a high-speed SerDes signal; a multimodal transmission harness for transmitting a high-speed SerDes signal, the multimodal transmission harness comprising a second data acquisition unit for acquiring second physical data; a SerDes deserializer module, configured to deserialize the high-speed SerDes signal to recover the video data and generate quality data of the SerDes signal; Host-side status monitoring module, used to collect connector status data; a harness health assessment module, configured to receive the second physical data, the SerDes signal quality data, and the connector status data, and to perform a health assessment on the multimodal transmission harness based on the second physical data, the SerDes signal quality data, and the connector status data; an operation behavior analysis module, configured to receive the second physical data and generate a real-time reminder signal according to a comparison result between the second physical data and a preset stress threshold; The user interface module is used to receive the health status assessment results and real-time reminder signals, and to display the health status assessment results and real-time reminder signals.

2. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The first data acquisition unit and the second data acquisition unit both include physical sensors, and the physical sensors include fiber Bragg grating sensors.

3. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The first physical data and the second physical data include instantaneous bending radius data and instantaneous tensile strain data.

4. The multimodal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The SerDes signal quality data generated by the SerDes deserializer module includes bit error rate data, eye diagram margin data and jitter data.

5. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The host-side status monitoring module is used to collect the connector status data, and the connector status data includes connector contact resistance data and connector plug-in and plug-out number data.

6. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The harness health assessment module is used to perform a health status assessment on the multimodal transmission harness, wherein the health status assessment includes a prediction of the remaining service life of the harness.

7. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The preset stress threshold in the operation behavior analysis module includes a preset minimum safe bending radius threshold and a preset maximum safe tensile force threshold.

8. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The operation behavior analysis module is used to generate the real-time reminder signal when the second physical data exceeds the preset stress threshold and lasts for a preset time period.

9. The multi-modal endoscope video transmission harness system based on a high-speed SerDes chip according to claim 1, characterized in that: The user interface module is used to display the real-time reminder signal in a visual manner, and the visual manner includes displaying text prompts or color changes at the edge of the video screen.

10. A multi-modal endoscope video transmission harness method based on a high-speed SerDes chip, according to any one of claims 1 to 9, wherein the multi-modal endoscope video transmission harness system based on a high-speed SerDes chip is characterized in that: The following steps are involved: S1: Collect video data, front-end control signals and first physical data collected by a first data acquisition unit; S2: Serializing the video data and the first physical data into a high-speed SerDes signal; S3: Transmitting the high-speed SerDes signal through a multimodal transmission harness, wherein a second data acquisition unit acquires second physical data within the multimodal transmission harness; S4: Deserializing the high-speed SerDes signal to recover the video data, and generating quality data of the SerDes signal; S5: Collect connector status data; S6: performing a health status assessment on the multimodal transmission harness based on the second physical data, the SerDes signal quality data, and the connector status data; S7: generating a real-time reminder signal based on a comparison result of the second physical data and a preset stress threshold; S8: Displaying the health status assessment result and the real-time reminder signal.

Images