CN-121971039-A - Cancerous wound management system and method based on AI evaluation and QFD

CN121971039ACN 121971039 ACN121971039 ACN 121971039ACN-121971039-A

Abstract

The invention discloses a cancerous wound management system and method based on AI evaluation and QFD, and relates to the technical field of medical health and artificial intelligence intersection. The system comprises a data acquisition module, a QFD requirement conversion module, an AI evaluation module, a multi-mode intervention module and a closed-loop adjustment module, wherein the method comprises the steps of requirement acquisition and sequencing, QFD requirement conversion, AI three-dimensional evaluation, multi-mode intervention and closed-loop adjustment. According to the invention, by combining QFD methodology with AI accurate assessment technology, accurate conversion of patient demands to quantifiable technical indexes is realized, and a full-period closed-loop management is formed by combining multi-mode physical intervention and multi-disciplinary collaborative scheme, so that the problems of demands and technology disconnection, rough assessment and single treatment pain point in traditional cancerous wound management are overcome, the treatment effect and patient experience are obviously improved, and the method is suitable for accurate and personalized management of cancerous wounds related to malignant tumors.

Inventors

ZHANG LEI
Tang Fengrong
ZHANG LIN

Assignees

上海市闵行区肿瘤医院

Dates

Publication Date: 20260505
Application Date: 20260108

Claims (10)

1. A cancerous wound management system based on AI assessment and QFD, comprising: The data acquisition module is configured to acquire clinical data and demand information of a malignant tumor related cancerous wound patient, wherein the clinical data comprises a wound image, a seepage related parameter, a pain score and a healing index, and the demand information comprises priority ordering of seepage control, pain relief and healing acceleration; The QFD demand conversion module is used for converting the demand information into quantifiable quality characteristics through a quality house matrix based on a quality function unfolding QFD methodology, wherein the quality characteristics comprise dressing imbibition related indexes, analgesia onset parameters, local oxygenation level and granulation tissue growth rate; the AI evaluation module is used for carrying out three-dimensional reconstruction on the wound image by adopting a three-dimensional scanning and deep learning algorithm and outputting three-dimensional morphological parameters of the wound and the concentration of the seepage protease; A multi-mode intervention module comprising a physical intervention unit configured to improve a local microenvironment of a wound or inhibit inflammatory factor release, a multidisciplinary MDT collaboration unit configured to generate a personalized treatment regimen and a psychological support plan; And the closed loop adjustment module is configured to dynamically adjust the target value of the quality characteristic according to the output result of the AI evaluation module and the execution feedback of the multi-mode intervention module to form a full-period management closed loop of 'demand-evaluation-intervention-feedback'.
2. The cancerous wound management system of claim 1 wherein the three-dimensional scan employed by the AI evaluation module is a binocular structured light scan, the deep learning algorithm is a boundary segmentation algorithm, and the wound stereomorphology parameters include wound area, depth, and irregular morphology parameters.
3. The cancerous wound management system of claim 1 wherein the QFD demand conversion module comprises: The demand ordering sub-module is used for scoring importance degree of the demands of the patient through an analytic hierarchy process and determining that the control of seepage, pain relief and healing acceleration are core demands; And the quality characteristic mapping submodule converts the core requirement into a specific technical index through a relation matrix.
4. The cancerous wound management system of claim 1 wherein the physical intervention unit of the multi-mode intervention module comprises a hyperbaric oxygen assist unit configured to improve local oxygenation of the wound and a pulsed blue light intervention unit configured to inhibit inflammatory factor release.
5. The cancerous wound management system of claim 1 wherein the multi-mode intervention module further comprises: The intelligent seepage management unit comprises a high-liquid-absorption foam dressing and a miniature pH sensor, wherein the pH sensor monitors the pH value of seepage in real time, and when the pH value is more than 7.5, dressing replacement early warning is triggered; The shape memory fixing unit adopts a nickel-titanium alloy wire braided supporting structure, and is automatically attached to the body surface shape of the wound part of the patient through body temperature induction.
6. The cancerous wound management system of claim 2 wherein the AI evaluation module comprises: The scanning device comprises a binocular structure light sensor and is used for acquiring three-dimensional point cloud data of a wound area; The image processing unit is used for carrying out boundary recognition on the three-dimensional point cloud data through a deep learning segmentation model and outputting wound area, depth and irregular morphological parameters; and the precision calibration unit is used for carrying out error correction on the three-dimensional reconstruction result based on the standard phascom model.
7. The cancerous wound management system of claim 1 wherein the closed loop adjustment module comprises: the data intercommunication interface is configured to perform data interaction with the hospital HIS system, the portable diagnostic instrument and the patient side APP, and acquire wound healing data and patient feedback in real time; The dynamic decision unit is based on a fuzzy PID algorithm, and automatically adjusts the operation parameters of the physical intervention unit according to the change rate of the seepage amount and pain score fluctuation.
8. A cancerous wound management method based on AI assessment and QFD employing a cancerous wound management system according to any of claims 1 to 7, comprising the steps of: S1, acquiring and sequencing requirements of cancerous wound patients through questionnaires and clinical interviews, and determining that the core requirements are seepage control, pain relief and healing acceleration; s2, QFD demand conversion, namely converting the core demand into the quality characteristic through a quality house matrix, and setting technical indexes that the liquid absorption amount of the dressing is more than or equal to 30mL/24h and the analgesic onset time is less than or equal to 15 min; s3, AI three-dimensional evaluation, namely reconstructing a three-dimensional shape of a wound by adopting a three-dimensional scanning and deep learning algorithm, and monitoring the concentration of the seepage protease; S4, multi-mode intervention, namely combining physical intervention and a multidisciplinary treatment scheme to execute personalized treatment; s5, closed-loop adjustment, namely dynamically correcting the quality characteristic target value according to the AI evaluation result, and optimizing the intervention scheme.
9. The method for managing cancerous wounds according to claim 8, wherein in step S3, the AI three-dimensional evaluation includes collecting three-dimensional point cloud data of a wound area through a binocular structured light sensor, performing boundary extraction on the point cloud data through a deep learning segmentation model, recognizing a gray threshold difference between a wound and normal tissues, performing three-dimensional morphological reduction on a segmentation result through a cloud server, and outputting a wound volume, a surface area and depth parameters.
10. The method of claim 8, wherein in step S4, the multidisciplinary regimen comprises a personalized analgesia regimen comprising administering a transdermal patch of fentanyl 30 minutes prior to surgery, irradiating the skin surrounding the wound with pulsed blue light 15 minutes prior to dressing change, continuously monitoring the pain VAS score after surgery, and triggering MDT remote consultation when the score is greater than or equal to 4.

Description

Cancerous wound management system and method based on AI evaluation and QFD Technical Field The invention relates to the technical field of medical health and artificial intelligence intersection, in particular to a cancerous wound management system and method based on AI evaluation and QFD. Background The cancerous wound is a refractory wound surface which appears in the progress process or after treatment of malignant tumors, is common to patients with advanced skin cancer, breast cancer, digestive tract tumor and the like, has the characteristics of large seepage volume, severe pain, difficult healing, easy infection and the like, seriously influences the life quality of the patients, and even aggravates the disease progress. The traditional wound assessment method mainly depends on subjective observation and experience judgment of medical staff, and has the core pain points that firstly, an assessment mode is extensive, traditional wound assessment mainly adopts two-dimensional photographing or visual observation, key parameters such as three-dimensional morphology of a wound, seepage components and the like cannot be accurately obtained, so that assessment result errors are large, personalized treatment is difficult to support, secondly, requirements and technology are disjointed, a clinical treatment scheme is mostly formulated based on general medical standards, individual requirements (such as seepage control priority, pain tolerance and the like) of a patient are not fully combined, treatment pertinence is insufficient, thirdly, a treatment mode is single, the traditional scheme mainly comprises dressing replacement and conventional analgesia, a multi-dimensional physical intervention and multi-disciplinary cooperative mechanism cannot be formed according to a wound healing dynamic adjustment scheme, fourthly, seepage management and wound fixing effects are poor, traditional dressing liquid absorption capability is limited, seepage state cannot be monitored in real time, a fixing device is difficult to adapt to complex body surface morphology, and problems such as dressing displacement and wound compression are easy to be caused. Quality Function Development (QFD) has been widely used in manufacturing and service industries as a systematic method for converting user demands into technical indexes, but effective landing has not been achieved in the field of cancerous wound management. Meanwhile, although the Artificial Intelligence (AI) three-dimensional reconstruction and deep learning technology is advanced in medical image evaluation, how to construct a full-period management system of 'demand-evaluation-intervention-feedback' by combining with QFD methodology is still a technical problem to be solved in the field of current cancerous wound management. Disclosure of Invention Aiming at the technical defects of rough assessment, disjointed requirements and technology, single treatment mode and lack of closed-loop adjustment mechanism in the conventional cancerous wound management, the invention provides a cancerous wound management system and method based on AI assessment and QFD, which realizes the accurate conversion of patient requirements to technical indexes through QFD methodology, combines an AI three-dimensional accurate assessment technology and a multi-mode intervention means, constructs a full-period closed-loop management system, and improves the cancerous wound treatment effect and the quality of life of patients. The technical scheme provided by the invention is as follows: the first aspect of the invention provides a cancerous wound management system based on AI evaluation and QFD, which is characterized by comprising: The data acquisition module is configured to acquire clinical data and demand information of a malignant tumor related cancerous wound patient, wherein the clinical data comprises a wound image, a seepage related parameter, a pain score and a healing index, and the demand information comprises priority ordering of seepage control, pain relief and healing acceleration; The QFD demand conversion module is used for converting the demand information into quantifiable quality characteristics through a quality house matrix based on a quality function unfolding QFD methodology, wherein the quality characteristics comprise dressing imbibition related indexes, analgesia onset parameters, local oxygenation level and granulation tissue growth rate; the AI evaluation module is used for carrying out three-dimensional reconstruction on the wound image by adopting a three-dimensional scanning and deep learning algorithm and outputting three-dimensional morphological parameters of the wound and the concentration of the seepage protease; A multi-mode intervention module comprising a physical intervention unit configured to improve a local microenvironment of a wound or inhibit inflammatory factor release, a multidisciplinary MDT collaboration unit configured to generate a personalized