Smart mirror for use in breast cancer diagnosis

Abstract

The invention relates to a smart mirror (A) that detects abnormal temperature changes or other pathological signs on the breast surface to provide early diagnosis of breast cancer.

Description

[0001] DESCRIPTION

[0002] SMART MIRROR FOR USE IN BREAST CANCER DIAGNOSIS

[0003] Technical Field

[0004] The invention relates to a smart mirror that detects abnormal temperature changes or other pathological signs on the breast surface to provide early diagnosis of breast cancer.

[0005] State of the Art

[0006] Currently, there are various applications for the diagnosis of breast cancer in the technical field. One of these applications is screening with radiation. This method, which uses X-ray radiation, is costly and can be applied in the healthcare center. Accordingly, the current method requires technical knowledge and professional help. This situation causes breast cancer to not be diagnosed early. In this sense, there is a need for new applications that are radiation-free, accessible, and safe without requiring technical knowledge and / or professional help.

[0007] The documents determined in the patent and literature search for the state of the art are summarized below.

[0008] Application numbered KR102543555 Bl relates to a device for self-diagnosing breast cancer with the support of artificial intelligence, without the help of a doctor. It also transmits the object image containing a feedback control command to the patient through augmented reality, encouraging the correct posture required for measurement and enabling breast cancer self-diagnosis through the screen panel of the smart mirror. It is noted that this method allows self-diagnose breast cancer by an Al neural network trained with thermal imaging. However, the document does not mention the use of structures to detect temperature and form change in that region.

[0009] Application numbered US2024008839A1 discloses a device for self-diagnosing breast cancer with artificial intelligence without the need for doctor help. This disclosure provides a thermographic ultrasound scanner device that guides an accurate posture correction of an ultrasound scanner using augmented reality and guides the patient with feedback control commands from a smart mirror. The document is not suitable for use in diagnosis and recommends use to correct the position of the patient.

[0010] As a result, due to the abovementioned disadvantages and the insufficiency of the current solutions regarding the subject matter, a development is required to be made in the relevant technical field.

[0011] Brief Description of the Invention

[0012] The present invention relates to a smart mirror for use in breast cancer diagnosis that meets the aforementioned needs, eliminates all disadvantages, and brings some additional advantages.

[0013] The object of the invention is to solve the abovementioned disadvantages by being inspired from the current conditions.

[0014] The main object of the invention is to develop a smart mirror to be used for early diagnosis of breast cancer. Thanks to its optical and / or thermal technologies, the smart mirror detects abnormal temperature changes or other pathological signs on the breast surface and enables users to quickly and non-invasively screen breast cancer at home or healthcare centers.

[0015] The smart mirror of the invention also has an artificial intelligence-based system that analyzes the detection results and informs the user.

[0016] The subject matter of the invention provides advantages in terms of smart mirror, radiation- free health screening, early diagnosis, cost, and time saving, ease of use and accessibility. The technical advantages of the invention according to the state of the art are summarized below:

[0017] • Radiation-Free Healthy Screening: While traditional methods use X-ray radiation such as mammography, the invention offers healthy screening without radiation exposure with the thermal imaging method. This is a safe alternative for frequent screenings.

[0018] • Increased Possibility of Early Diagnosis: While traditional methods are performed periodically and in health centers, the invention allows individuals to screen themselves regularly thanks to its ease of use and accessibility at home. This increases the rate of early diagnosis and ensures that patients are treated early.

[0019] • Cost and Time Savings: Traditional methods are costly and require making an appointment and going to the healthcare center. The invention is less costly and can be used at home or in easily accessible areas (shopping malls, medical centers, etc.). This enables healthcare services to reach a wider audience and reduces healthcare costs.

[0020] • Ease of Use and Accessibility: While traditional methods require technical knowledge and professional help, the invention can be easily used at home or in public places. This increases access to healthcare and allows individuals to better control their own health.

[0021] • Increasing Health Awareness: The invention raises awareness about healthy life and emphasizes the importance of early diagnosis. This improves the general health status by increasing the health awareness of the public.

[0022] The invention plays an important role in the image analysis and data collection stages. The processor (Raspberry Pi) integration makes these processes more accessible and portable, while increasing the efficiency of the invention.

[0023] The structural and characteristic features and all the advantages of the invention will be understood more clearly by means of the figures and the detailed description with reference to these figures given below and therefore, the evaluation should be made by taking these figures and the detailed description into consideration.

[0024] Descriptions of the Figures

[0025] The present invention should be evaluated together with the figures explained below, in order that it will be constructed, and its advantages will be understood in the best way.

[0026] Figure 1 shows the general view of the smart mirror of the invention.

[0027] Figure 2 shows the internal view of the smart mirror device.

[0028] Description of References of the Parts A Smart mirror

[0029] 1 Thermal Camera

[0030] 2 Processor

[0031] 3 Screen

[0032] 4 Power supply

[0033] 5 Processor cooler

[0034] 6 USB Ports

[0035] 7 Power supply input

[0036] 8 Home button

[0037] Detailed Description of the Invention

[0038] The preferred embodiments of the invention are described in this detailed description only for a better understanding of the subject and in a way that does not have any limiting effect.

[0039] The invention relates to a smart mirror (A) that detects abnormal temperature changes or other pathological signs on the breast surface to provide early diagnosis of breast cancer. The smart mirror (A) of the invention includes at least one thermal camera (1) with at least one optical sensor to detect abnormal temperature changes or other pathological signs on the breast surface. Thanks to these sensors, it is ensured that temperature differences on the body surface are detected. Preferably, InfiRay P2 Pro is used as a thermal camera (1) in the invention. The connection of the thermal camera (1) is usually provided via USB, SPI or I2C.

[0040] At least one processor (2) for processing and analyzing the thermal images detected with the thermal camera (1) is located in the smart mirror (A) structure of the invention. Raspberry Pi 5 - 8 GB is preferably used as a processor (2) in the invention. The processor (2) is the operating system loaded with the necessary libraries (TensorFlow, OpenCV, etc.).

[0041] Thanks to the processor (2), anomalies can be determined and presented to the user with a visual and / or audible alert.

[0042] In the invention, the thermal camera (1) images are connected to the processor (2) with a USB or other appropriate connection for data transmission. The smart mirror (A) of the invention includes a screen (3) containing an interface suitable for the user to perform operations and see the results. The screen (3) can be documented and connected with the HDMI output in a preferred embodiment of the invention. The user interface can provide thermal images and analysis results, screening initiation by receiving data from the user, and displaying results, statistics, and general health status recommendations. Accordingly, the user can start the screening process and see the results instantly. Also, the results can be sent to mobile devices or displayed on the screen. The processor (2) communicates with a screen (3) to show the user interface and the analysis results.

[0043] The energy requirement of the processor (2) is provided from the power supply (4) during the operating period.

[0044] The smart mirror of the invention also has an artificial intelligence-based system that analyzes the detection results and informs the user.

[0045] It is suitable for use at home in a preferred embodiment of the invention. However, it is also possible to use it in healthcare centers and / or different places and places.

[0046] The smart mirror (A) has a power supply (4) that will provide power for the processor (2) and other components. The power supply (4) may preferably be a battery. If a battery is used, the appropriate adapter is used for charging.

[0047] At least one data storage unit in the form of a memory card is used for the data storage of the processor (2). Preferably, a microSD card and similar units can be used in the invention. In addition, at least one processor cooler (5) is included in the invention to provide cooling of the processor (2). According to this purpose, fans and heat sinks can be used, and the air flow can be directed.

[0048] A system used in the invention and operating on the basis of artificial intelligence includes the following modules.

[0049] Data Collection Module:

[0050] • Thermal Camera (1): It detects the temperature changes in the user's body and collects the data. • Graphical User Interface (GUI): It is the screen (3) on which the user interacts.

[0051] Data Processing Module

[0052] • Processor (2): It processes the collected data. Here, model training is carried out with TensorFlow or a similar library.

[0053] Data Analysis Algorithms: It analyzes thermal images and extracts health statistics.

[0054] Data Storage Module:

[0055] • Database or File System: It stores user data and analysis results.

[0056] Result Presentation Module:

[0057] • Output System: It provides the user with health statistics and recommendations.

[0058] The algorithm for the development of the artificial intelligence used in the invention is summarized below.

[0059] 1. Data Set Preparation

[0060] The first step is the collection of the dataset that is required to train the Al model. In the present invention, the images obtained by the thermal camera will be used. Thermal cameras are the ideal data source for recognizing cancerous areas by detecting temperature changes on the skin surface. Labeling and pre-processing of thermal images (such as normalization, noise removal, data enhancement) is a critical step that directly affects the success of the model.

[0061] 2. Model Selection and Training

[0062] Convolutional Neural Networks (CNN) is preferred as the artificial intelligence algorithm. CNN gives very successful results in image processing tasks. At this stage, deep learning libraries such as TensorFlow or TensorFlow Lite will be used. The training of the model can be carried out with an advanced architecture such as Y0L0v8, as this architecture offers effective results in low latency and real-time detection.

[0063] • Training Process: During the training phase, the model will be trained using the thermal data set created on the computer. Training data should be labeled and processed to increase the accuracy of the model. During the training, the performance criteria of the model (accuracy, sensitivity, specificity) will be monitored and hyperparameter optimization will be made.

[0064] 3. Optimizing the Model Optimization is required for the efficiency of a model that will work on low-power hardware such as the Raspberry Pi or NVIDIA Jetson Nano. At this point, the model is reduced and optimized using TensorFlow Lite. In the optimization process, the size of the model is reduced, and the processing power requirements are reduced by using methods such as quantization and pruning.

[0065] 4. Integrating the Model into the Device

[0066] The data obtained by the thermal camera is analyzed by the artificial intelligence model that will work on the Raspberry Pi or Jetson Nano processor. The model regularly analyzes the temperature changes in the user's body every day. Thus, it detects possible anomalies and provides information about the health status of the user.

[0067] • Operation on the Device: The data received by the thermal camera sensor is transmitted to the artificial intelligence model in real time. By processing this data, the Raspberry Pi or Jetson Nano recognizes the abnormal temperature zones and provides feedback to the user's screen.

[0068] 5. Evaluation and Improvement of Results

[0069] The system updates the user's health statistics with each use and updates the training with new data to increase the accuracy of the model as the data set expands over time. Monitoring the performance of the algorithm, verifying the outputs, and improving them when necessary ensures that the system works correctly in the long term.

[0070] All of these steps will enable the invention to provide an effective artificial intelligence solution in the field of healthcare, while at the same time allowing the system to work portable and in real time.

[0071] 1. Data Monitoring and Analysis

[0072] While the artificial intelligence model detects diseases such as breast cancer with the data from the thermal camera, the Health Guide software constantly monitors this data and makes a deeper analysis by recording the user's historical health data. When the user is in front of the mirror every day, the thermal camera detects potential anomalies by detecting body temperatures. Using this data, the Health Guide provides services such as:

[0073] • Daily and weekly health reports: They generate reports on temperature changes related to the user's health status. • Visual feedback: It presents abnormal temperature changes with graphics and user- friendly images.

[0074] • Statistical analysis: By evaluating the data over time with statistical analysis, it provides early alerts about potential problems.

[0075] 2. Personalized Health Advice

[0076] In line with the outputs of the artificial intelligence model, the health guide provides personalized recommendations on the health status of the user. For example, if an unusual temperature difference is detected in the thermal analysis, the software will recommend:

[0077] • Doctor visit recommendations: It advises the user to consult a healthcare professional.

[0078] • Lifestyle advice: It provides suggestions for improvement in areas such as diet, exercise, and rest.

[0079] 3. User Interface

[0080] The Health Guide provides thermal camera results and health advice with a user-friendly interface. This interface has the following:

[0081] • Simple and intuitive design: It allows the user to get information quickly and easily about their health status.

[0082] • Interactive health tracking system: Users can see past screening results, monitor their progress, and review graphical summaries of their health status.

[0083] • Notification system: It informs the user about health emergencies or conditions that need to be followed.

[0084] 4. Al Integration

[0085] The data obtained by the artificial intelligence model is processed in the Health Guide software and converted into meaningful information. The data analyzed by the model are compared with the user's daily health status, disease risk profile and more general health statistics in the guide. The user can constantly monitor their health through this software and benefit from this data with the help of algorithms.

[0086] 5. Data Security and Privacy

[0087] The security of the user's health data is of great importance. In this regard:

[0088] • Data encryption: The user's personal health data is stored securely and protected by encryption methods. • Anonymous data analysis: It can be used to determine general health trends by anonymizing users' information.

[0089] 6. Additional Features

[0090] The Health Guide software can also provide users with general health information and guidance services. These features include:

[0091] • Symptom monitoring: The user can have the software analyze this information by entering their various symptoms on a daily basis.

[0092] • Personal health goals: It provides the user with the opportunity to set health goals and monitor their progress towards these goals.

[0093] 7. Portable Device Integration

[0094] Since the hardware part of the device is assembled in a mirror-sized structure, it is designed to work with mini computers such as the Raspberry Pi or Jetson. The Health Guide software integrated into these devices is optimized for energy efficiency while providing portability. The software can be used on the embedded screen of the mirror or through a mobile application.

[0095] Health Guide algorithm is a comprehensive and user-friendly software that will help users perform daily health screenings through artificial intelligence-powered thermal imaging. This software provides the interpretation of thermal analyses provided by the artificial intelligence model and informs users with personalized health recommendations.

[0096] In addition to the aforementioned algorithm, the development of a personal health application is summarized below.

[0097] 1. Data Monitoring and Analysis

[0098] While the artificial intelligence model detects diseases such as breast cancer with the data from the thermal camera, the Health Guide software constantly monitors this data and makes a deeper analysis by recording the user's historical health data. When the user is in front of the mirror every day, the thermal camera detects potential anomalies by detecting body temperatures. Using this data, the Health Guide provides services such as: • Daily and weekly health reports: They generate reports on temperature changes related to the user's health status.

[0099] • Visual feedback: It presents abnormal temperature changes with graphics and user- friendly images.

[0100] • Statistical analysis: By evaluating the data over time with statistical analysis, it provides early alerts about potential problems.

[0101] 2. Personalized Health Advice

[0102] In line with the outputs of the artificial intelligence model, the Health Guide provides personalized recommendations on the health status of the user. For example, if an unusual temperature difference is detected in the thermal analysis, the software will recommend:

[0103] • Doctor visit recommendations: It advises the user to consult a healthcare professional.

[0104] • Lifestyle advice: It provides suggestions for improvement in areas such as diet, exercise, and rest.

[0105] 3. User Interface

[0106] The Health Guide provides thermal camera results and health advice with a user-friendly interface. This interface has the following:

[0107] • Simple and intuitive design: It allows the user to get information quickly and easily about their health status.

[0108] • Interactive health tracking system: Users can see past screening results, monitor their progress, and review graphical summaries of their health status.

[0109] • Notification system: It informs the user about health emergencies or conditions that need to be followed.

[0110] 4. Al Integration

[0111] The data obtained by the artificial intelligence model is processed in the Health Guide software and converted into meaningful information. The data analyzed by the model are compared with the user's daily health status, disease risk profile and more general health statistics in the guide. The user can constantly monitor their health through this software and benefit from this data with the help of algorithms.

[0112] 5. Data Security and Privacy

[0113] The security of the user's health data is of great importance. In this regard: • Data encryption: The user's personal health data is stored securely and protected by encryption methods.

[0114] • Anonymous data analysis: It can be used to determine general health trends by anonymizing users' information.

[0115] 6. Additional Features

[0116] The Health Guide software can also provide users with general health information and guidance services. These features include:

[0117] • Symptom monitoring: The user can have the software analyze this information by entering their various symptoms on a daily basis.

[0118] • Personal health goals: It provides the user with the opportunity to set health goals and monitor their progress towards these goals.

[0119] 7. Portable Device Integration

[0120] • Since the hardware part of the device is assembled in a mirror- sized structure, it is designed to work with mini computers such as the Raspberry Pi or Jetson. The Health Guide algorithm integrated into these devices is optimized for energy efficiency while providing portability. The software can be used on the embedded screen of the mirror or through a mobile application.

[0121] Health guide algorithm is a comprehensive and user-friendly software that will help users perform daily health screenings through artificial intelligence-powered thermal imaging. This algorithm provides the interpretation of thermal analyses provided by the artificial intelligence model and informs users with personalized health recommendations.

[0122] In the principle of using the smart mirror (A) of the invention, firstly, the patient is in front of the device and the screening is started. The screen (3) in the smart mirror (A) structure is activated when the user stands in front of it. A welcome message preferably appears on the screen (3) and the screening begins in a relaxing tone. Preferably, the system interacts with the user. The artificial intelligence first asks an introductory question to understand the general health status of the patient: "How are you feeling today?" The user can choose the appropriate one from the options (good, tired, uncomfortable, etc.) that appear on the screen (3). This response is recorded in the machine learning model of the system. In addition, the system asks about swelling, one of the common symptoms of breast cancer. The question, "Have you noticed any swelling in your chest area?" is asked to the user. The answer has an important role in the user-specific screening results and is processed in machine learning decision trees. Then, the question, "Have you noticed a color change or other abnormality in your chest area?" appears on the screen. This information is used to compare with the data obtained as a result of thermal imaging. The next question is about the last menstrual period. Because the menstrual cycle can be one of the breast cancer risk factors, the system asks, "When was your last period?" The user's response is recorded to assess the impact of hormonal changes. Because genetic predisposition is an important factor, the system asks, "Is there an individual in your family diagnosed with breast cancer?" The user's response plays an important role in risk assessment.

[0123] After the questions are completed, the system prompts the user to open the chest area and the thermal camera is activated. The thermal camera detects heat distributions in the chest area and records image data in high resolution. Image processing algorithms work to detect abnormal temperature differences.

[0124] The data received by the camera is processed by artificial intelligence-based image processing algorithms. Comparisons are made with the thermal images of healthy and cancerous individuals in the large database in the system. In this process, machine learning algorithms evaluate previous user data and image data.

[0125] The user's responses (swelling, discoloration, family history, etc.) and thermal images are combined and analyzed. Machine learning decision trees create a result based on this data and make a potential risk assessment.

[0126] After analysis is complete, the system presents the results to the user. If there is a negative situation, the results are explained in a gentle and soft language that will not cause the user to panic: "Based on your screening results, we have observed some symptoms. We recommend that you consult a specialist for further evaluation." The user is guided by individual recommendations and next steps.

[0127] If the user gives feedback on the results and recommendations, this data is added to the machine learning structure of the system. These feedback are used to improve the accuracy of the system for future screenings.

Claims

CLAIMS1. A smart mirror (A) to provide early diagnosis of breast cancer, characterized in that it comprises the following:• at least one thermal camera (1) with at least one optical sensor to detect temperature changes or pathological signs on the breast surface,• at least one processor (2) for processing and analyzing the thermal images detected by the thermal camera (1),• an artificial intelligence-based system interface that analyzes the detection results and informs the user,• a screen (3) suitable for the user to perform operations and see the results,• at least one power supply (4) to meet the energy requirement of the processor (2) during the operating period,• at least one data storage unit for data storage of the processor (2).

2. A smart mirror (A) according to claim 1, characterized in that it comprises at least one processor cooler (5) to provide cooling of the processor (2).

3. A smart mirror (A) according to claim 1, characterized in that it comprises an InfiRay P2 Pro as thermal camera (1).

4. A smart mirror (A) according to claim 1, characterized in that it comprises a Raspberry Pi 5 - 8 GB as a processor (2).

5. A smart mirror (A) according to claim 1, characterized in that it comprises a touch screen (3).

6. A smart mirror (A) according to claim 1, characterized in that it comprises a memory card as a data storage unit.

Images